70267 Economics of Adaptation to Climate Change SYNT HESIS REPORT EACC Publication and Reports 1. Economics of Adaptation to Climate Change: Synthesis Report 2. Economics of Adaptation to Climate Change: Social Synthesis Report 3. The Cost to Developing Countries of Adapting to Climate Change: New Methods and Estimates Country Case Studies: 1. Bangladesh: Economics of Adaptation to Climate Change 2. Bolivia: Adaptation to Climate Change: Vulnerability Assessment and Economic Aspects 3. Ethiopia : Economics of Adaptation to Climate Change 4. Ghana: Economics of Adaptation to Climate Change 5. Mozambique: Economics of Adaptation to Climate Change 6. Samoa: Economics of Adaptation to Climate Change 7. Vietnam: Economics of Adaptation to Climate Change Discussion Papers: 1. Economics of Adaptation to Extreme Weather Events in Developing Countries 2. The Costs of Adapting to Climate Change for Infrastructure 3. Adaptation of Forests to Climate Change 4. Costs of Agriculture Adaptation to Climate Change 5. Cost of Adapting Fisheries to Climate Change 6. Costs of Adaptation Related to Industrial and Municipal Water Supply and Riverine Flood Protection 7. Economics of Adaptation to Climate Change-Ecosystem Services 8. Modeling the Impact of Climate Change on Global Hydrology and Water Availability 9. Climate Change Scenarios and Climate Data 10. Economics of Coastal Zone Adaptation to Climate Change 11. Costs of Adapting to Climate Change for Human Health in Developing Countries 12. Social Dimensions of Adaptation to Climate Change in Bangladesh 13. Social Dimensions of Adaptation to Climate Change in Bolivia 14. Social Dimensions of Adaptation to Climate Change in Ethiopia 15. Social Dimensions of Adaptation to Climate Change in Ghana 16. Social Dimensions of Adaptation to Climate Change in Mozambique 17. Social Dimensions of Adaptation to Climate Change in Vietnam 18. Participatory Scenario Development Approaches for Identifying Pro-Poor Adaptation Options 19. Participatory Scenario Development Approaches for Pro-Poor Adaptation: Capacity Development Manual Economics of Adaptation to Climate Change SYNTHESIS REPORT ii ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT ©2010 The International Bank for Reconstruction and Development/The World Bank 1818 H Street NW Washington DC 20433 www.worldbank.org socialdevelopment@worldbank.org All rights reserved. This paper has not undergone the review accorded to official World Bank publications. The �ndings, interpretations, and conclusions expressed herein are those of the author(s) and do not necessarily reflect the views of the International Bank for Reconstruction and Development / The World Bank and its affiliated organizations, or those of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgement on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. RIGH TS A ND P ER MI S SI ONS The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org All images © The World Bank Photo Library. Contents Abbreviations and Acronyms ix Acknowledgments xi Executive Summary xv I. Introduction 1 Context 1 Objectives 1 Approaches: the two parallel tracks 2 The Synthesis Report 2 II. Concepts, methodology, and limitations 5 Concepts 5 Dealing with uncertainty 8 Methodology 10 Limitations 13 Future Work 16 III. Results from the global analyses 19 Putting a price tag on adaptation 19 Putting the �ndings in context 23 Lessons 25 IV. Results from the country analyses 29 Mozambique 30 Ethiopia 38 Ghana 47 Bangladesh 54 Bolivia 63 Vietnam 70 Samoa 80 iv ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT V. Lessons 89 The costs of adapting to climate change 89 Economic development and adaption to climate change 90 Climate uncertainty: the need for robust strategies 91 Current climate vulnerabilities 93 Hard vs. soft approaches to adaptation 94 Conclusion 96 References 97 Notes 100 Figures Figure ES-1 Pulling Together the Key Findings xvi Figure ES-2 A Two-Track Approach xvii Figure ES-3 Adaptation in Mozambique Entails Greater Climate Resilience xxiii Figure ES-4 Adaptation Signi�cantly Lowers Welfare Losses in Ethiopia xxiv Figure 1 Study Structure: Global and Country Tracks 2 Figure 2 Study Methodology: Global and Country Tracks 11 Figure 3 Total Annual Cost of Adaptation and Share of Costs, 20 NCAR and CSIRO Scenarios, by Region Figure 4 Total Annual Cost of Adaptation for the National Centre 21 For Atmospheric Research (NCAR) Scenario, by Region and Decade Figure 5 Total Annual Cost of Adaptation for the National Centre for 21 Atmospheric Research (NCAR) Scenario, by Region and Decade Figure 6 Development Lowers the Number of People Killed by Floods and 25 Affected by Floods and Droughts, 2000–50 Figure 7 Population Density and Coastal Area 32 Figure 8 Agriculture: Effects on Yield in 2050 Compared to Baseline 34 Figure 9 Decomposition of Impact Channels From a Macroeconomic Perspective 34 Figure 10 Reduction in CC Damages, 2003–50 35 Figure 11 Household Consumption: Coef�cient of Variation of 36 ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT v Year-To-Year Growth Rates Figure 12 Agroecological Zones in Ethiopia 39 Figure 13 Deviations of GDP From Base Scenario 40 Figure 14 Agricultural Year-to-Year Growth Rates: Standard Deviations 40 Figure 15 Regional GDP, Deviation From Base, Wet2 Standard Deviations 41 Figure 16 Net Present Value (NPV) of Welfare Differences 42 Figure 17 Standard Deviation of Year-To-Year Agriculture GDP Growth Rates, 43 with and without Adaptation Figure 18 Bene�t/Cost Ratio of Upgrading Road Standards 45 Figure 19 Surface Flow Average Difference from the No-Climate-Change Scenario, 2010–50 Figure 20 Deviation of Real Total and Agricultural GDP from Base, 49 Terminal Year Figure 21 Maps of Poverty, Flooding and Tidal Surges 55 Figure 22 Inundation Risk from Storm Surges in Coastal Areas With and 57 Without Climate Change in Relation to Current Population Density Figure 23 Inundation Risk From Storm Surges in Coastal Areas with and 58 without Climate Change In Relation to Current Poverty Rates Figure 24 Change In Inundation Depth (CM) From Monsoon Flooding 59 (with and without CC) Figure 25 Projected National Rice Production with and without Climate Change 60 Figure 26 Annual Percent Change in Agricultural GDP 64 from El Niño and La Niña Effects Figure 27 Projected Precipitation Changes to 2050 under 64 Different Climate Scenarios Figure 28 Changes In Runoff for the Red River and Mekong Delta 75 by Climate Scenario Figure 29 Regions of Samoa Used in the Climate-Economy Model 81 Tables Table ES-1 East Asia and Paci�c Will Shoulder the Biggest Burden xix Table ES-2 Infrastructure, Coastal Zones, and Water Capture xx the Bulk of Adaptation Costs vi ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT Table ES-3 Getting a Better Picture xxii Table ES-4 Key Adaptation Options in Mozambique xxiii Table ES-5 Better Road Design Standards Are a High Priority for Ghana xxiv Table ES-6 Polders and Cyclone Shelters are Key for Bangladesh’s xxv Cyclone-Induced Storm Surges Table ES-7 Agricultural Adaptation in Vietnam Reduces xxvi Inequalities and Helps Gdp Table ES-8 Better Design Standards Would Enhance Samoa’s Resilience xxvii Table 1 Recommendations for Future Work 17 Table 2 Total Annual Costs of Adaptation for All Sectors, by Region, 2010–50 19 Table 3 Comparison of Adaptation Cost Estimates by the Unfccc 24 and the Eacc, $ Billions Table 4 Sector Analyses Carried Out in Each Country Case Study 30 Table 5 Key Adaptation Options in Mozambique 37 Table 6 Adaptation Costs 42 Table 7 Adaptation Costs and Residual Damage 43 Table 8 Gcm Scenarios for the Ghana Case Study 48 Table 9 Real Welfare Impacts with No Adaptation Investments 51 Table 10 Deviations of Real Welfare from Baseline under Alternative 52 Adaptation Strategies Table 11 Cost of Adapting to Tropical Cyclones and Storm Surges by 2050 61 Table 12 Total Adaption Cost for Inland Flooding by 2050 61 Table 13 Cost-Bene�t Analysis of Adaptation Measures in the Agriculture 69 and Water Sectors Table 14 Regional Vulnerability to Climate Change 72 Table 15 Change In Crop Production In 2050 Due To Climate Change 75 with no Adaptation Table 16 Macroeconomic Effects of Climate Change 77 without/with Adaptation in 2050 Table 17 Deviations in Precipitation and Temperature in 2050 82 by Climate Scenario Table 18 Losses Due to Climate Change without and with Adaptation 83 Table 19 Average Adaptation Costs by Decade, High Scenario, 86 2005 $ Million/Year, No Discounting ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT vii Boxes Box 1 Cge Models—Partial and General Equilibrium Analysis 31 Box 2 Methodology and Field Work of Social Component in Mozambique 38 Box 3 Past Experience Adapting to Extreme Climate Events in Bangladesh 56 Box 4 Local Factors Influence Local Adaptation Preferences in Bolivia 71 ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT ix Abbreviations and Acronyms AR4 Fourth Assessment Report of the NGO Nongovernmental organization Intergovernmental Panel on Climate ODA Official development assistance Control OECD Organisation for Economic BAP Bali Action Plan Co-operation and Development CGE Computable general equilibrium PNC National Watershed Program (by its (model) Spanish acronym) CLIRUN Climate and runoff (model) PPM Parts per million CMIP3 Coupled Model Intercomparison PSD Participatory scenario development Project phase 3 SAR South Asia (World Bank region) CSIRO Commonwealth Scienti�c and Indus- SRES Special Report on Emissions Scenarios trial Research Organization (climate (of the IPCC) model) SSA Sub-Saharan Africa (World Bank DALY Disability-adjusted life years region) DIVA Dynamic and interactive vulnerability UN United Nations assessment (model) UNDP United Nation Development Program EACC Economics of Adaptation to Climate UNFCCC United Nations Framework Conven- Change tion on Climate Change EAP East Asia and Paci�c (World Bank WCRP World Climate Research Programme region) WHO World Health Organization ECA Europe and Central Asia (World Bank region) Note: Unless otherwise noted, all dollars are U.S. ENSO El Niño Southern Oscillation dollars; all tons are metric tons. GCM General circulation model GDP Gross domestic product IMPACT International model for policy analysis of agricultural commodities and trade IPCC Intergovernmental Panel on Climate Change LAC Latin America and Caribbean (World Bank Region) MIP Mixed integer programming (model) MIROC Model for interdisciplinary research on climate MNA Middle East and North Africa (World Bank Region) NAPA National Adaptation Plans of Action NCAR National Centre for Atmospheric Research (climate model) x ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT Acknowledgments ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT xi The Economics of Adaptation to Climate Change Hughes, David Johnson, Paul Kirshen, Gerald Nel- (EACC) study has been a large, multiyear undertaking son, James Neumann, Robert Nicholls, Jason Price, managed by a core team of the World Bank’s Envi- Adam Schlosser, Robert Schneider, Roger Sedjo, ronment Department led by Sergio Margulis (Task Kenneth Strzepek, Rashid Sumaila, Philip Ward, and Team Leader) and comprising Urvashi Narain, Kiran David Wheeler (consultants). Major contributions Pandey, Laurent Cretegny, Ana Bucher, Robert Sch- were made by Carina Bachofen, Brian Blankespoor, neider, Gordon Hughes, and Timothy Essam. Robin Ana Bucher, David Corderi, Susmita Dasgupta, Mearns, Anne Kuriakose, and Carina Bachofen coor- Timothy Essam, Robin Mearns, Siobhan Murray, dinated the social dimensions of the study. The study Hawanty Page, and Jasna Vukoje (World Bank), and would not have been possible without the exceptional Jeroen Aerts, Steve Commins, William Farmer, Eihab administrative support from Hawanty Page, Jasna Fathelrahman, Prodipto Ghosh, Dave Johnson, James Vukoje, Grace Aguilar, Hugo Mansilla, Arathi Sun- Juana, Tom Kemeny, Benoit Laplante, Larissa Leony, daravadanan, and Patricia Noel. Linh Phu, Mark Rosegrant, Klas Sanders, Arathi Sundaravadanan, Timothy Thomas, and Tingju Zhu Since the beginning, the EACC team has had (consultants). Sally Brown and Susan Hanson made intense interaction with the Environment Depart- important contributions to the coastal sector report; ment’s management, particularly Warren Evans and Miroslac Batka, Jawoo Koo, David Lee, Marilia Michele de Nevers, who are an integral part of the Magalhaes, Siwa Msangi, Amanda Palazzo, Claudia EACC team. The team is also immensely grateful to Ringler, Richard Robertson, and Timothy Sulser to Ravi Kanbur, Sam Fankhauser, and Joel Smith for the agriculture sector report; William Cheung to the serving on the advisory committee, providing overall �shery sector report; and Pieter Pauw and Luke M. guidance to the team, and resolving critical technical Brander to the water sector report (consultants). and strategic issues. Africa Country Case Studies. Overall coordination A very large number of people have directly worked by Raffaello Cervigni and Aziz Bouzaher (former on the various reports and background studies, and coordinator), with contributions by Stephen Mink an even larger number of people have provided direct and Stephen Ling (World Bank). and indirect support and advice to the teams. It will be hard to do justice individually to all people, so we Ethiopia. Raffaello Cervigni (coordinator), Aziz simply list the main teams of the EACC study and Bouzaher (former coordinator); Principal authors acknowledge indirect contributions in the end. were Kenneth Strzepek, Sherman Robinson, Len Wright and Raffaello Cervigni; additional experts on Synthesis Report. Sergio Margulis, Gordon Hughes, the team were Paul Block, Brent Boehlert, Paul Chi- Robert Schneider, Kiran Pandey, and Urvashi Narain nowsky, Chas Fant, William Farmer, Alyssa McClus- with editorial services from Bruce Ross-Larson, key, Michelle Minihane, Niko Strzepek, and Gete Laura Wallace, and Robert Livernash and adminis- Zeleke. The World Bank team comprised Marie Ber- trative support from Hugo Mansilla. Contributions nadette Darang, Susmita Dasgupta, Edward Dwum- to the country case studies were provided by the four, Achim Fock, Francesca Fusaro, Anne Kuriakose, respective country teams listed below. Stephen Ling, Sergio Margulis (team leader of the overall EACC study), Stephen Mink, Kiran Pandey, Global Track Report. Sergio Margulis and Urvashi Dawit Tadesse, and Fang Xu. Narain (coordinators), Laurent Cretegny, Anne Kuriakose, Glenn Marie Lange, and Kiran Pandey Mozambique. Jean-Christophe Carret (coordi- (World Bank), as well as Paul Chinowsky, Gordon nator), Kenneth Strzepek, Channing Arndt, Paul xii ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT Chinowsky, Anne Kuriakose (World Bank), James Alí, Victoria Parra Goitia, Gladys Tesoro Michel, Neumann, Robert Nicholls, James Thurlow, and Len Carmen Rosa Del Castillo Gutierrez, Consuelo Luna, Wright with support from Carina Bachofen (World Claudia Canedo, Carlos Cabrera, Alvaro Lambert, Bank), Sally Brown, Charles Fant, Adèle Faure, Susan Victor Vazquez, Erwin Kalvelagen, Miguel Morales, Hanson, Abiy Kebede, Jean-Marc Mayotte, Michelle Fernando Cossio, and Valeria Sanchez (consultants). Minihane, Isilda Nhantumbo, and Raphael Uaiene (all Consultants except as noted) authored the report. Samoa. Laurent Cretegny and Sergio Margu- Inputs were also provided by Aziz Bouzaher, Raf- lis (coordinators), Gordon Hughes (consultant) faello Cervigni, Sergio Margulis(team leader of the with contributions from Timothy Essam (World overall EACC study), Stephen Mink, and Antonio Bank); Simon Bannock, Michele Daly, Erwin Nucifora (all World Bank). Kalvelagen, Peter King, Thakoor Persaud, Isikuki Punivalu, Graeme Roberts, and Fernanda Ghana. Peter Kristensen (coordinator) and Robin Zermoglio (consultants). Mearns (World Bank); Channing Arndt, Dirk Willenbockel, Kenneth Strzepek, Eihab Fathelrah- Vietnam. Benoit Laplante (coordinator), David man, Robert Nicholls, Len Wright, Chas Fant, Paul Corderi, Laurent Cretegny, Douglas Graham, Steve Chinowsky, Alyssa McCluskey, Sherman Robinson, Jaffee, Robin Mearns, and Anne Kuriakose (World Michelle Minihane, William Farmer, Brent Boehlert, Bank), Gordon Hughes (consultant) with contribu- Jean-Marc Mayotte, Tony Dogbe, Joseph Yaro, David tions from Huynh Thuy (consultants); Tingju Zhu Pessey, Emilia Arthur, George Ahiable, Tia Yahaya, and Zhe Guo (IFPRI); Le Heng Nam, Nguyen Kamil Abdul Salam, Samantha Boardley, Simon Thuy Hang, Ha Le Thanh, and Thuy Dung (Insti- Mead, and Livia Bizikova (consultants). Other con- tute of Water Resources Planning); Do Duc Dung tributors include Aziz Bouzaher, Raffaello Cervigni, (Southern Institute of Water Resources planning); Anne Kuriakose, Sergio Margulis, Stephen Mink, Philip Adams, James Giesecke, Michael Jerie, and Urvashi Narain, Kiran Pandey, and John Fraser Stew- Nhi Hoang Tran (Centre of Policy Studies, Monash art (World Bank) as well as Saadia Bobtoya, John University); To Trung Nghia, Le Hung Nam, Le Asafu-Adjaye, Dyson Jumpah, Daniel Bruce Sar- Hong Tuan, Truong Trong Luat, and Vu Dinh Huu pong, Philip Acquah, Robert Schneider,and Victoria (Institute of Water Resources and Planning); Pham Bruce-Goga (consultants). Quang Ha, Mai Van Trinh, Tran Van The, and Vu Duong Quynh (Institute for Agricultural Envi- Bangladesh. Kiran Pandey (coordinator), Nilufar ronment); Bao Thanh, Luong Van Viet, Nguyen Ahmad, Susmita Dasgupta, Shakil Ferdausi, Anne Thi Phuong, and Bui Chi Nam (Sub-Institute of Kuriakose, Khawaja M. Minnatullah, Winston Yu Hydrometeorology and Environment of South (World Bank), Mainul Huq, Zahirul Huq Khan, Viet Nam); Tuyen Nghiem, Hue Le, and Huoung Manjur Murshed Zahid Ahmed, Ainun Nishat, Vu Dieu (Center for Natural Resources and Envi- Nandan Mukherjee, Malik Fida Khan, and Zul�qar ronmental Studies); Pamela McElwee (Arizona Iqbal (consultants). State University); Dang Thu Phuong (Challenge to Change); Nguyen Van Be, Le Canh Dung, Nyuyen The Plurinational State of Bolivia (hereafter Hieu Trung and Sinh Le Xuan(Can Tho University); referred to as Bolivia). Ana Bucher (coordina- Suan Pheng Kam, Marie Caroline Badjeck, Michael tor), Carina Bachofen, Robert Schneider, Laurent Phillips, and Robert Pomeroy(World Fish Center); Cretegny, David Corderi, Morten Blomqvist, Anne Louise Teh and Lydia The (University of British Kuriakose, and Ruth Llanos (World Bank);Magali Columbia); Be Nam Vo Thi (Sub-National Institute García Cárdenas, Jorge Cusicanqui, Bruno Condori for Agricultural Planning and Projection); and Hien ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT xiii Than Thi and Hue Nguyen Thu (Centre for Marine- life Conservation and Community Development). The team is also grateful for both general and speci�c comments received from the following colleagues: Julia Bucknall, Shanta Devarajan, Marianne Fay, Gherson Feder, Armin Fidler, Kirk Hamilton, Tamer Rabie, Peter Rogers, Jim Shortle, Michael Toman, Gary Yohe, Vahid Alavian, Jan Bojö, Henrike Brecht, Kenneth Chomitz, Vivian Foster, Alexander Lotsch, Kseniya Lvovsky, Chopp Margulis, Dominique van Der Mensbrughe, John Nash, Ian Noble, Giovanni Ruta, Apurva Sanghi, Robert Townsend, Walter Ver- gara, and Winston Yu. From outside the Bank, they include Marten van al Aast, Shailaja Annamraju, Roy Brouwer, Maureen Cropper, Anton Hilbert, Chris- tine Pirenne, Tamsin Vernon, Peter Wooders, and Vincent-van Zeijst. None of these colleagues and reviewers are in any way responsible for the contents and eventual errors of this report, which remain the sole responsibility of the study team. The EACC study was conducted by a partnership consisting of the World Bank (leading its techni- cal aspects); the governments of the Netherlands, the United Kingdom, and Switzerland (funding the study); and the participating case study countries. The team would like to thank the partnership that initiated, funded, and actively engaged with the study team through its multiyear journey. xiv ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT Executive Summary ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT xv As developing countries weigh how best to revitalize To address these objectives, the study was conducted their economies and craft a sustainable development on two parallel tracks: (1) a global track—a top-down path to boost living standards, they will have to factor approach, in which national databases were used to in the reality that the global annual average tempera- generate aggregate estimates at a global scale, draw- ture is expected to be 2º C above pre-industrial levels ing on a wide variety of sector studies; and (2) a by 2050. A 2º warmer world will experience more country level track—a bottom-up approach, in which intense rainfall and more frequent and more intense sub-national data were aggregated to generate esti- droughts, floods, heat waves, and other extreme mates at economywide, sectoral, and local levels. This weather events. As a result, it will have dramatic Synthesis Report integrates and summarizes the key implications for how countries manage their econo- �ndings of a global study report and seven country mies, care for their people and design their develop- case study reports—covering Bangladesh, Bolivia, ment paths. Countries will need to adopt measures to Ethiopia, Ghana, Mozambique, Samoa, and Vietnam adapt to climate change. These measures offer a way (Figure ES-1). By providing information on lessons to make the effects of climate change less disruptive learned and insights gained on adaptation to climate and spare the poor and the vulnerable from shoulder- change from global, country, and sector-level analyses, ing an unduly high burden. the hope is to help policymakers worldwide prioritize actions, along with developing a robust, integrated Against this backdrop, the global community adopted approach for greater resilience to climate risks. The the Bali Action Plan at the 2007 United Nations Cli- Report begins with the concepts and methodology mate Change Conference. The plan calls for devel- used for analyses in both the global and the country oped countries to allocate “adequate, predictable, and case studies, including a discussion of study limita- sustainable �nancial resources and new and additional tions. This is followed by a synthesis of key results resources, including official and concessional funding from the global and country tracks and a conclusion for developing country parties�1 to help them adapt with lessons learned. to climate change. It also underscores that interna- tional cooperation is essential for building capacity A Call to Change Course to integrate adaptation measures into sectoral and national development plans. What are the key �ndings of the Report? To begin with, economic development is perhaps the best hope for How high will the price tag be? Studies to date have adaptation to climate change. Development enables an provided only a wide range of estimates, from $4 bil- economy to diversify and become less reliant on sec- lion to $109 billion a year. That is why the Economics tors such as agriculture that are more vulnerable to of Adaptation to Climate Change (EACC) study was climate change effects. It also makes more resources initiated in early 2008 by the World Bank in partner- available for minimizing risk, and similar measures ship with the governments of Bangladesh, Plurina- often promote development and adaptation. tional State of Bolivia, Ethiopia, Ghana, Mozambique, Samoa, and Vietnam, and funded by the governments That said, it cannot be development as usual. Adaptation of The Netherlands, Switzerland, and The United will require a different kind of development—such as Kingdom. Its objectives are twofold: to develop a global breeding crops that are drought and flood tolerant, estimate of adaptation costs for informing the interna- climate-proo�ng infrastructure to make it resilient to tional community’s efforts in the climate negotiations, climate risks and accounting for the inherent uncer- and to help decision-makers in developing countries tainty in future climate projections in development assess the risks posed by climate change and design planning. And it will cost to adapt. Our global study national strategies for adaptation. estimates that the price tag between 2010 and 2050 for xvi ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT FIGURE ES-1 PULLING TOGETHER THE KEY FINDINGS EACC SYNTHESIS REPORT GLOBAL STUDY 7 COUNTRY CASE STUDIES BANGLADESH BOLIVIA ETHIOPIA GHANA MOZAMBIQUE SAMOA VIETNAM Agriculture Sector Sector Sector Sector Sector Sector Sector Water Sector 2 Sector 2 Sector 2 Sector 2 Sector 2 Sector 2 Sector 2 SECTORS SECTORS Health Sector 3 Sector 3 Sector 3 Sector 3 Sector 3 Sector 3 Sector 3 Sector 4 Sector 4 Sector 4 Sector 4 Sector 4 Sector 4 Sector 4 Coastal ... others ... others ... others ... others ... others ... others ... others Infrastructure ...Others Source: Revised estimates (World Bank 2010a) adapting to an approximately 2oC warmer world by farmers with the tools and resources to respond to 2050 will be in the range of $70 billion2 to $100 bil- climate change) (World Bank 2010a). lion a year (World Bank 2010a).Our country studies suggest that costs could be even higher, once cross- Finally, given the uncertainty surrounding both cli- sectoral impacts are taken into account. mate outcomes and longer-term projections of social and economic development, countries should try However, there are numerous “low-regret� actions— to delay adaptation decisions as much as possible and typically policies that would be priorities for devel- focus on low-regret actions. They should also build opment even without climate change—especially the resilience of vulnerable sectors. In agriculture, in water supply and flood protection. Economists for example, this would mean better management regularly urge policymakers to adopt mechanisms for of water resources and access to extension services managing water resources that recognize the scarcity to give policymakers greater flexibility in handling value of raw water, advice that is almost invariably either droughts or waterlogging caused by floods. ignored because of deeply embedded political inter- ests. The reality is that the costs of misallocation of How the Studies were Done water resources will escalate even without climate change and could be overwhelming with it. At the The intuitive approach to costing adaptation involves same time, steps must be taken to identify and help comparing a future world without climate change the poor and most vulnerable—including soliciting with a future world with climate change. The differ- their views on adaptation priorities and ensuring ence between the two worlds entails a series of actions an enabling environment (for example, providing to adapt to the new world conditions. And the costs ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT xvii of these additional actions are the costs of adapting to level of nutrition, and water supply availability). climate. Figure ES-2 summarizes the methodological The baselines used a consistent set of GDP and approaches of the two tracks. population forecasts for 2010–2050. Choosing climate projections. While there is con- For the global study, the following four steps were siderable consensus among climate scientists that taken: climate change is unequivocal, accelerating and Picking a baseline. For the timeframe, the world in human-induced (IPCC 2007), there is much 2050 was chosen, not beyond (forecasting climate less agreement on how climate change will affect change and its economic impacts becomes even natural and social systems. For that reason, two more uncertain beyond this period). Development climate scenarios were chosen to capture as large baselines were crafted for each sector, essentially as possible a range of model predictions—from establishing a growth path in the absence of extreme wet to extreme dry. climate change that determines sector-level per- Predicting impacts. An analysis was done to formance (such as stock of infrastructure assets, predict what the world would look like under FIGURE ES-2 A TWO-TRACK APPROACH Global Track Country Track Projections Projections Climate Climate Sub-national Global Data Sets Data Sets Water Run-off Water Run-off Baseline GDP/Population Baseline GDP/Population Economic, Social and Economic, Social and Environmental Impacts Environmental Impacts Participatory Scenarios w/vunerable Identi�cation of Adaptation Identi�cation of Adaptation groups Measures Measures Decision Rule Decision Rule Cost of Adaptation Cost of Adaptation Sectors Cross-Sectors Agriculture Coastal Zones National Macroeconomic Forestry Extreme Weather Events Analysis Fisheries Infrastructure Water Resources Sectors Cross-Sectors Health Agriculture Coastal Zones Ecosystem Services Extreme Weather Events Forestry Fisheries Infrastructure Water Resources Source: Revised estimates (World Bank 2010a) xviii ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT the new climate conditions. This meant trans- of welfare in the (future) world as they would lating the impacts of changes in climate on the have without climate change. This in principle various economic activities (agriculture, �sheries), overstates the costs of adaptation. Of course, on people’s behavior (consumptions, health), on governments can choose to not adapt at all, environmental conditions (water availability, for- incurring all damage from climate change, or ests), and on physical capital (infrastructure). adapt to the point where bene�ts from adapta- Identifying and costing adaptation alternatives. tion equal their costs, at the margin (“optimal� Adaptation actions were selected to offset the adaptation). predicted impacts and to restore welfare in each What exactly is “adaptation�? Countries face not of the major economic sectors analyzed—infra- only a de�cit in adapting to current climate varia- structure, coastal zones, water supply and flood tion, let alone future climate change, but also def- protection, agriculture, �sheries, human health, icits in providing education, housing, health, and and forestry and ecosystem services. The costs of other services. Thus, many countries face a more these actions together with the cost implications general “development de�cit,� of which the part of changes in the frequency of extreme weather related to climate events is termed the “adapta- events were also estimated. But a cross-sectoral tion de�cit.� This study makes the adaptation analysis of costs was not feasible. de�cit a part of the development baseline, so that adaptation costs cover only the additional costs to For the country studies, two additional steps were cope with future climate change. taken: Soft versus hard measures. “Hard� options (capital Evaluating economywide impacts. A macroeco- intensive) were favored over “soft� options (insti- nomic modeling framework—known as a Com- tutions and policies)—because they are easier to putable General Equilibrium (CGE) model—was quantify. used to facilitate the analysis of macroeconomic Public versus private adaptation. The focus was and cross-sectoral effects of the impacts and on planned adaptation (deliberate public deci- adaptation to climate change. sion) rather than autonomous or spontaneous Evaluating social impacts. A social component adaptation (households or communities acting was used to gather information on preferred on their own without public interventions but adaptation strategies and sequence strategies within an existing public policy framework). from a bottom-up, local–level perspective. It also How to include bene�ts. Some countries and some provided new evidence on how vulnerability is sectors may bene�t from changes in climate. socially differentiated, and on the importance The question is how to account for these gains. of social accountability and good governance A number of different approaches were used to for achieving pro-poor, climate-resilient devel- account for these gains. opment. It went beyond planned adaptation, How to handle uncertainty. Total adaptation weighing the potential of adaptation taken by costs for a speci�c climate projection assume households, collective action, nongovernmental that policymakers know with certainty that a organizations, and the private sector. particular climate projection will materialize. Thus, the use of two extreme scenarios, wettest For all of the studies, a number of concepts had to be and driest, provides a range of estimates for a agreed upon: world in which decision-makers have perfect How much to adapt. The studies assumed that foresight. If decision-makers end up having to countries would fully adapt—that is, adapt up hedge their bets and consider both scenarios at to the level at which they enjoy the same level the same time, costs will be higher. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT xix The Global Picture adaptation cost, and the Middle East and North Africa the lowest. Latin America and the Carib- Overall, the global study estimates that the cost bean and Sub-Saharan Africa follow East Asia and between 2010 and 2050 of adapting to an approxi- Paci�c in both scenarios. On a sector breakdown, mately 2oC warmer world by 2050 is in the range of the highest costs for East Asia and the Paci�c are $70 billion3 to $100 billion a year. This sum is the in infrastructure and coastal zones; for Sub-Saha- same order of magnitude as the foreign aid that devel- ran Africa, water supply and flood protection and oped countries now give developing countries each agriculture; for Latin America and the Caribbean, year. But it is still a very low percentage (0.17 percent) water supply and flood protection and coastal zones; of the income of countries (measured by their GDP, and for South Asia, infrastructure and agriculture. which was roughly $60 trillion in 2009). The costs vary by climate scenario and whether bene�ts from Not surprisingly, both climate scenarios show costs climate change are used to offset adaptation costs (see increasing over time, although falling as a percent- Table ES-1). age of GDP—suggesting that countries become less vulnerable to climate change as their economies The driest scenario (Commonwealth Scienti�c and grow. There are considerable regional variations, Industrial Research Organization, CSIRO) requires however. Adaptation costs as a percentage of GDP lower total adaptation costs than does the wettest are considerably higher in Sub-Saharan Africa scenario (National Centre for Atmospheric Research, than in any other region, in large part because of NCAR), largely because of the sharply lower costs the lower GDPs but also owing to higher costs of for infrastructure, which outweigh the higher costs adaptation for water resources, driven by changes for water and flood management. In both scenarios, in precipitation patterns. infrastructure, coastal zones, and water supply and flood protection account for the bulk of the costs. On the sectoral level, the EACC �ndings offer insights for policymakers who must make On a regional basis, for both climate scenarios, tough choices in the face of great uncertainty the East Asia and Paci�c Region bears the highest (see Table ES-2). TABLE ES-1 EAST ASIA AND PACIFIC WILL SHOULDER THE BIGGEST BURDEN (Global costs of adaptation by region) Aggregation type/ East Asia & Europe & Latin Middle South Asia Sub- Total Scenario Paci�c Centr.Asia America & East/ North Saharan Caribbean Africa Africa Gross-sum/ 25.7 12.6 21.3 3.6 17.1 17.1 97.5 Wet Scenario X-sum/ 17.9 6.9 14.8 2.5 15 14.1 71.2 Dry Scenario Note: Gross-sum is the sum of the positive costs of adaptation in all sectors and all countries, excluding gains (negative costs) from climate change. X-sum deducts these eventual gains in countries with overall positive costs of adaptation. Source: Revised estimates (World Bank 2010a). xx ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT TABLE ES-2 INFRASTRUCTURE, COASTAL ZONES, AND WATER CAPTURE THE BULK OF ADAPTATION COSTS (Global costs of adaptation by sector, X-Sum) Sector Wet Dry Infrastructure 27.5 13.0 Coastal zones 28.5 27.6 Water supply and flood protection 14.4 19.7 Agriculture, forestry, �sheries 2.5 3.0 Human health 2.0 1.5 Extreme weather events 6.7 6.4 Total 81.5 71.2 Source: Revised estimates (World Bank 2010a). Infrastructure.4 This sector has accounted for the Water supply.6 In some parts of the world, water avail- largest share of adaptation costs in past studies and ability has risen and will continue to do so, but in takes up a major share in the EACC study—in fact, others, it has fallen and will continue to do so—and the biggest share for the NCAR (wettest) scenario the frequency and magnitude of floods are expected because the adaptation costs for infrastructure are to rise. The EACC study shows that water supply especially sensitive to levels of annual and maximum and flood management ranks as one of the top three monthly precipitation. Urban infrastructure—drain- adaptation costs in both the wetter and drier sce- age, public buildings, and similar assets—accounts narios, with Sub-Saharan Africa footing by far the for about 54 percent of the infrastructure adapta- highest costs. Latin America and the Caribbean also tion costs, followed by railways at 18 percent, and sustain high costs under both models, and South Asia roads (mainly paved) at 16 percent. East Asia and sustains high costs under CSIRO. the Paci�c and South Asia face the highest costs, reflecting their larger populations. Sub-Saharan Agriculture.7 Climate change affects production by Africa experiences the greatest increase over time. altering yields and areas where crops can be grown. The EACC study shows that temperature and pre- Coastal zones.5 These zones are home to an ever- cipitation changes in both climate scenarios will growing concentration of people and economic signi�cantly lower crop yields and production— activity, yet they are also subject to a number of with irrigated and rainfed wheat and irrigated rice climate risks, including sea-level rise and storm affected the most. Developing countries fare worse surges, and possible increased intensity of tropi- for almost all crops compared to developed countries, cal storms and cyclones. The study shows that with South Asia shouldering the biggest production adaptation costs are signi�cant and vary with the declines. Moreover, the changes in trade flow patterns magnitude of sea-level rise, making it essential are dramatic—with exports for developed countries for policymakers to plan while accounting for the rising and South Asia becoming a much larger food uncertainty. One of the most striking results is that importer under both scenarios. That said, the total Latin America and the Caribbean and East Asia costs for the agricultural sector relative to other sec- and the Paci�c account for about two-thirds of the tors is lower than would have been expected, in part total adaptation costs. because welfare is restored through trade rather than ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT xxi by restoring yields within countries. Yet, it is impor- Putting Global Findings in Context tant to note that larger population shares, and among them, the poorer of the poor, are highly vulnerable How does this study compare with earlier studies? The because of their dependence on agriculture. EACC estimates are in the upper end of estimates by the UNFCCC (2007), the study closest in approach Human health.8 Climate change can trigger increases to this study, though not as high as suggested by a in the incidence of vector-born diseases, water-borne recent critique of the UNFCCC study by Parry and diseases, heat- and cold-related deaths, injuries and others (2009). A comparison of the studies is limited deaths from flooding, and the prevalence of malnu- by methodological differences—in particular, the use trition. The EACC study, which focuses on malaria of a consistent set of climate models to link impacts to and diarrhea, �nds adaptation costs falling in absolute adaptation costs and an explicit separation of costs of terms over time to less than half the 2010 estimates development from those of adaptation in the EACC of adaptation costs—thanks to improvements in basic study. But the major difference between them is the health conditions that accompany higher incomes nearly six-fold increase in the cost of coastal zone and development. While the declines are consistent management and defense under the EACC study. across regions, the rates of decline in South Asia and This difference reflects several improvements to the East Asia and Paci�c are faster than in Sub-Saharan earlier UNFCCC estimates under the EACC study: Africa—a continent that will shoulder more than 80 better unit cost estimates, including maintenance percent of the health sector adaptation costs by 2050. costs, and the inclusion of costs of port upgrading and risks from both sea-level rise and storm surges.9 Extreme weather events.23 Without reliable data on emergency management costs, the EACC study The bottom line: calculating the global cost of adap- tries to shed light on the role of socioeconomic tation remains a complex problem, requiring projec- development in increasing climate resilience. It tions of economic growth, structural change, climate asks: As climate change increases potential vulner- change, human behavior, and government invest- ability to extreme weather events, how many addi- ments 40 years in the future. The EACC study tried to tional young women would have to be educated to establish a new benchmark for research of this nature, neutralize this increased vulnerability? And how as it adopted a consistent approach across countries much would it cost? The �ndings show that by and sectors and over time. But in the process, it had 2050, neutralizing the impact of extreme weather to make important assumptions and simpli�cations, to events requires educating an additional 18 million some degree biasing the estimates. to 23 million young women at a cost of $12 billion to $15 billion a year. For 2000–50, the tab reaches Important shortcomings of this study relate to three about $300 billion in new outlays. This means that broad categories: uncertainty, institutions, and mod- in the developing world, neutralizing the impact eling limitations (Table ES-3). They are natural entry of worsening weather over the coming decades points for thinking about future work and knowledge will require educating a large new cohort of young needs. The highest priority in the immediate future women at a cost that will steadily escalate to sev- must be to reduce the range of uncertainty about eral billion dollars a year. But it will be enormously future climate impacts and to identify forms of adap- worthwhile on other margins to invest in education tation that are robust across the range of uncertainty for millions of young women who might otherwise that will remain. be denied its many bene�ts. xxii ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT TABLE ES-3 GETTING A BETTER PICTURE STUDY LIMITATION RECOMMENDATION Use of mathematical models Include institutional, social, cultural and political perspectives to identify and no ef�ciency criterion good policies. Find simpler rules for policymaking Climate uncertainty Consider more scenarios, Monte Carlo simulations and other probabilistic approaches Growth uncertainty Hard to improve other than through sensitivity analyses Technological uncertainty Incorporate better information from sector specialists and simulate the impact of potential advances. Non-consideration to institu- Context speci�c institutional capacity has to be assessed and considered to tional issues make recommendations realistic and feasible Limited focus on migration and Work with outside projections; limited current knowledge on cities and urbanization climate change Limited range of adaptation Include a broader range of strategies, including more local level No environmental services Pull better information and introduce more consistent estimates The Country Picture affected by high rainfall variability. The most vulnera- ble sectors from climate change are agriculture, which As for the EACC country studies, seven countries employs over 70 percent of the Mozambique popu- were selected based on overall vulnerability to major lation; energy, particularly hydropower generation, climate change impacts; differing environmental, which is dependent on water runoff; transport infra- social, and economic conditions and adequate data at structure, notably roads; and coastal areas. Findings the national level. Government interest at the highest from the social component suggest that livelihood level was also important. Mozambique, Ghana, and activities most sensitive to climate change continue to Ethiopia represent nearly the full range of agricultural take place in areas most exposed to climate change. systems in Africa. Vietnam and Bangladesh—Asian countries with most of their economic activity and On adaptation, the EACC Mozambique country population concentrated along the coast and in low- study shows that with small additional costs, sealing lying deltas—are among the world’s most vulnerable unpaved roads—a low-regret option—would restore to climate change, especially from extreme weather about one-�fth of the welfare loss owing to climate events and flooding. Bolivia is a poor Latin American change (see Figure ES-3). Remaining welfare losses country traditionally dependent on the Andean gla- could be regained with better agricultural produc- ciers to supply good portions of water demand, and it tivity or human capital accumulation (education). consists of a wide range of agro-ecosystems. Samoa Irrigation investments appear to be a poor alterna- represents a low-lying Paci�c island at increased risk tive. Investment costs are likely to be about US$400 to sea level rise and storm surge. million per year over 40 years. As part of the social component, participants in scenario development Mozambique is subject to frequent droughts, floods, workshops were asked to draft preferred adaption and tropical cyclones—events that threaten the coun- options for the government (planned) and non-gov- try’s economic performance, which is already highly ernment entities (autonomous) (see Table ES-4). For ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT xxiii FIGURE ES-3 the hard options, more resources will be needed, and for the soft options, an enabling economic and politi- ADAPTATION IN MOZAMBIQUE ENTAILS GREATER CLIMATE RESILIENCE cal environment. (Reduction in climate change damages, 2003-2050; 5 percent discount rate, constant 2003 US$) Ethiopia is heavily dependent on rain-fed agricul- 8 ture, and its geographical location and topography Discounted US $ billion (const.2003) 7 in combination with low adaptive capacity entail a 6 high vulnerability to the impacts of climate change. Historically the country has been prone to extreme 5 6.1 6.1 weather variability, resulting in seven major droughts 4 since the early 1980s—�ve of which led to famines. 3 The primary losses owing to climate change in Ethio- 2 0.6 pia arise from the effect of extreme weather events, 1 1.5 1.5 1.5 1.5 both droughts and flooding, on agricultural produc- 0 tion and infrastructure. Transport Expanding Agriculture Primary Infrastructure (3) Irrigation (4) R&D (5) Education (6) Adaptation strategies considered in Ethiopia build Agriculture R&D Irrigation Sealing Unpaved or Education Roads on current government programs—such as increas- ing the irrigated cropland area and investing in Note: In the worst scenario: NPV of damages without adaptation is $7.6 billion discounted at 5%) Source: World Bank 2010g. TABLE ES-4 KEY ADAPTATION OPTIONS IN MOZAMBIQUE Planned Adaptation Autonomous Adaptation Hard ■ Flood control dikes and levies ■ More robust buildings ■ Coastal flood control gates ■ Farm-scale water storage facilities ■ Dams and irrigation channels ■ Deep wells to provide drinking water for ■ Improved roadways people and animals ■ Improved communication infrastructure ■ Grain storage facilities ■ Improved hospitals and schools ■ Improved food processing equipment Soft ■ Improved early warning of climatic hazards, ■ Better utilization of short season, drought and of dam releases resistant crops to prepare for drought, floods, and ■ Better planning and management of forest, cyclones �sh, and other natural resources ■ Diversi�cation of flood and drought risk by ■ Resettlement of populations to lower risk maintaining �elds in both highland and lowland areas zones ■ Better household and community management ■ More credit and �nancial services for small busi- and use of natural resources, including wild fruits nesses and rural development ■ Practice of soil conservation agriculture ■ Better education and information for the rural ■ Migration to lower risk areas areas ■ Diversi�cation of livelihoods away from ■ Improved health care, social services, and agriculture social support for all people ■ Better planning of how much grain to save for personal consumption, and how much to sell for income generation Note: The options in plain text respond directly to climate hazards, while those in italics represent measures to increase the population’s adaptive capacity, or make them more resilient to shocks to their livelihoods. Source: World Bank 2010g. xxiv ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT FIGURE ES-4 from climate sensitive sectors, such as agriculture; upgrading road design; strengthening hydropower ADAPTATION SIGNIFICANTLY LOWERS WELFARE LOSSES IN ETHIOPIA development; and accelerating absorption of the rural (Net present value (NPV) of absorption differences) labor force into non-agriculture activities—including through skills-upgrading programs and encourage- Wet 2 Dry 2 Wet 1 Dry 1 ment of growth poles around medium-size munici- 0.0 Ratio (%) to NPV of Base GDP palities. Total adaptation costs range from US$1.22 -2.0 billion (wet) to $5.84 billion (dry) per year over 40 years, though it may be possible to reduce these costs -4.0 through a labor-upgrading program. -6.0 Ghana is highly vulnerable to climate change and -8.0 variability because it is heavily dependent on climate- -10.0 sensitive sectors such as agriculture—largely rain-fed with a low-level of irrigation development—forestry, No Adaptation Adaptation and hydropower. The country has a 565 kilometer long coastline that is inhabited by about a quarter of the population and is the location of signi�cant physical Note: NPV of absorption, difference from base (percent of NPV of infrastructure. The EACC Ghana country study esti- GDP). Absorption is de�ned as GDP, plus imports minus exports. mates that climate change causes a reduction in real Wet 1 and Dry 1 are the two secnarios used in the global analysis, and Wet 2 and Dry 2 are the wettest and driest scenarios in Ethiopia. household consumption of 5-10 percent in 2050, with Source: World Bank 2010e. rural households suffering greater reductions, primar- ily through its impact on agricultural production. agricultural research and development; boosting the share of paved and hardened roads; and altering the The study evaluated adaptation options focused on scale and timing of planned hydropower projects. roads, agriculture, hydropower, and coastal protection. The EACC Ethiopia country study shows that with- It found that changes in road design standards alone out adaptation Ethiopia’s GDP would be lower by 2 provide signi�cant reductions in welfare losses in most to 8 percent for the four different climate scenarios scenarios. The combination of better road design and analyzed. Adaptation reduces welfare losses by about investments in agriculture or agriculture and hydro- 50 percent (see Figure ES-4) and also lowers income power or education would minimize or even reverse variability. It also highlights the potential bene�ts of the losses caused by climate change under the four accelerating the diversi�cation of the economy away climate scenarios analyzed (see Table ES-5). TABLE ES-5 BETTER ROAD DESIGN STANDARDS ARE A HIGH PRIORITY FOR GHANA (NPV of deviations of real welfare from baseline under alternative adaptation strategies, US$ billions) Adaptation Investment in No Adaptation scenario Road Design Agriculture Hydro /Agric. Education Global Dry -13.118 -10.308 -0.121 -0.941 -2.090 Global Wet -10.095 -5.854 -2.973 2.116 0.584 Ghana Dry -2.709 -3.009 -1.193 -1.782 -1.308 Ghana Wet -4.050 -0.766 1.936 1.358 1.795 Note: The options in plain text respond directly to climate hazards, while those in italics represent measures to increase the population’s adaptive capacity, or make them more resilient to shocks to their livelihoods. Source: World Bank 2010f. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT xxv Bangladesh is one of the most vulnerable countries embankments to protect highly productive agricul- to climate risks. About two-thirds of the nation is less tural lands, drainage systems, and erosion control than 5 meters above sea level and is susceptible to measures. The study found that the total estimated river and rainwater flooding. Once every three to �ve cost would be $2.7 billion in initial investment years, up to two-thirds of Bangladesh is inundated and $54 million in annual recurrent cost—with 80 by floods. Cyclone-induced storm surges owing to percent of these costs stemming from road height climate change are expected to inundate an addi- enhancement. Given the large existing variability, tional 15 percent of the coastal area and increase the a prudent near–term strategy for Bangladesh is to inundation depth in these areas. The damages from address the large adaptation de�cit while investing a single typical severe cyclone with a return period to reduce uncertainties about future climate change of 10 years is expected to rise nearly �vefold to over which will de�ne where additional investments are $9 billion by 2050, accounting for 0.6 percent of required. By 2050, the number of people living in GDP. The burden is likely to fall disproportionately cities will triple while the rural population will fall on the rural poor in low-lying coastal areas who are by 30 percent. The long-term challenge is to move also affected by other climate-related hazards such as people and economic activity into less climate- saline water intrusion into aquifers and groundwater sensitive areas. and land submergence. Bolivia is exposed to hydro-meteorological extremes For storm surges induced by tropical cyclones, the and climate variability, particularly because of the EACC Bangladesh country study evaluated adapta- influence of the El Niño oscillation (ENSO), which, tion measures such as embankments, afforestation, regardless of climate change, occurs periodically in cyclone shelters, and early warning systems. It found different parts of the country. Floods, landslides, that the total estimated cost would be $2.4 billion in and droughts—which seriously affect food security initial investment and $50 million in annual recur- and the water supply—are also common. Given that rent costs (see Table ES-6). As for inland flooding, Bolivia’s economic mainstays are minerals and gas, it the focus was on infrastructure measures to avoid is relatively insensitive to climate change. Yet most further damage from additional inundation—road people are engaged in small-scale agriculture, a sec- network and railways, river embankments and tor that is quite vulnerable to climate changes. TABLE ES-6 POLDERS AND CYCLONE SHELTERS ARE KEY FOR BANGLADESH’S CYCLONE-INDUCED STORM SURGES (Total adaption cost for inland flooding by 2050, US$ Million) Baseline Scenario (additional risk CC Scenario (existing risks) (1) due to CC) (2) (total risk = (1) + (2)) Adaptation Option IC AMC IC AMC IC AMC Polders 2,462 49 893 18 3,355 67 Afforestation 75 75 Cyclone shelters 628 13 1,219 24 1,847 37 Resistant housing 200 200 Early warning system 39 8 39 8 Total 3,090 62 2,426 50 5,516 112 CC = climate change; IC = investment cost; AMC = annual maintenance cost Source: World Bank 2010c. xxvi ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT The EACC Bolivia country study focused on agricul- In agriculture, adaptation measures ranged from ture and water resources. Even in the more optimistic autonomous actions by farmers (such as sowing dates, scenario of wetter conditions, agricultural productiv- switching to drought-tolerant crops, and adoption of ity can only increase if the capacity to store and use salinity-tolerant varieties of rice) to planned public the needed additional water is available for farmers actions (such as greater spending on research, devel- and poor peasants. Thus, “no-regrets� measures would opment, and extension; and extending the areas of include better water resources management and build- irrigated land).The study shows that the impacts of ing water storage and irrigation infrastructure. These climate change on agriculture and related sectors, even types of measure are already envisaged by Bolivia in with no adaptation, appear to be relatively modest— its development agenda, but the study shows that the given that farmers are expected to change, without development agenda must now be accelerated. government interventions, the crops and crop vari- eties that they grow and their methods of farming. Vietnam’s exposure to weather-related events and The CGE macroeconomic analysis suggests that the disasters ranks among the highest among all developing drop in real GDP and real consumption is much less countries. Storms and floods occasionally resulting from severe with adaptation—and adaptation offsets most tropical cyclones have caused extensive and repeated of the disproportionate impact of climate change on damages to buildings and infrastructure, agriculture and the poor (see Table ES-7). In fact, GDP would not �sheries sectors, and resulted in a large number of fatali- only fall less but possibly also even increase. That said, ties. Climate change may well bring an increase in the a major concern is how much poorer households will frequency, intensity, and duration of floods, and greater suffer from lower agricultural incomes and higher drought problems in the dry season. food prices relative to the general cost of living. The EACC Vietnam country study focused on Samoa is a country at extreme risk from a variety agriculture, aquaculture, forestry, and coastal ports. of natural disasters including tropical cyclones and TABLE ES-7 AGRICULTURAL ADAPTATION IN VIETNAM REDUCES INEQUALITIES AND HELPS GDP (Percentage deviations in 2050 from baseline with no climate change) No adaptation With adaptation Adaptation bene�ts Dry Wet MoNRE Dry Wet MoNRE Dry Wet MoNRE GDP -2.4% -2.3% -0.7% -1.1% -0.7% 0.7% 1.3% 1.6% 1.3% Aggregate consumption -2.5% -2.5% -0.7% -1.4% -0.8% 0.6% 1.1% 1.7% 1.3% Agricultural value-added -13.9% -13.5% -5.8% -3.8% -3.4% 5.4% 10.0% 10.1% 11.2% Regional GDP North Central Coast -6.6% -6.1% -2.6% 0.5% -0.3% 4.8% 7.1% 5.8% 7.4% South East 1.1% 0.8% 1.0% 0.0% 1.1% 0.2% -1.1% 0.3% -0.9% Rural household consumption Bottom quintile -6.5% -6.3% -2.6% -1.9% -1.4% 2.4% 4.7% 4.9% 5.0% Top quintile -1.6% -1.7% -0.4% -1.5% -1.0% 0.0% 0.1% 0.7% 0.4% Note: MoNRE is Vietnam’s Ministry of Environment, which has established an “of�cial� climate change scenario that is roughly similar to the Hadley Center (UK) projections. Source: World Bank 2010i. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT xxvii TABLE ES-8 BETTER DESIGN STANDARDS WOULD ENHANCE SAMOA’S RESILIENCE (Impact of climate change with and without adaptation) No adaptation With adaptation Bene�t of adaptation NCAR CSIRO NCAR CSIRO NCAR CSIRO Design standards for 10-year return period Present value @ 5%, $ million 103.9 212.4 34.8 24.5 69.1 187.9 Annualized equivalent, $ million per year 5.9 12.1 2.0 1.4 3.9 10.7 Loss/bene�t as % of baseline GDP 0.6 1.3 0.2 0.2 0.4 1.2 Loss/bene�t as % of baseline consumption 0.9 1.9 0.3 0.2 0.6 1.7 Design standards for 50-year return period Present value @ 5 %, $ million 19.9 37.0 4.5 5.4 15.4 31.6 Annualized equivalent, $ million per year 1.1 2.1 0.3 0.3 0.9 1.8 Loss/bene�t as % of baseline GDP 0.1 0.2 0.0 0.0 0.1 0.2 Loss/bene�t as % of baseline consumption 0.2 0.3 0.0 0.0 0.1 0.3 Note: NCAR is wettest scenario; CSIRO is driest scenario. Source: World Bank 2010h. tsunamis caused by earthquakes. It is also subject to that extreme weather variability in the coastal zone inter-annual climate fluctuations associated with El will involve signi�cant costs for either investments in Nino (ENSO), which affect precipitations as well as coastal protection or the relocation of assets. In the air and sea temperatures. Many scientists believe that longer term, the relocation of assets—or even whole climate change will lead to some increase in the inten- villages—may be the best option as it would shift sity of tropical cyclones accompanied by greater vari- economy activity such as tourism, crops, and other ability of rainfall with more frequent episodes of heavy businesses away from the coast. rainfall and drought. A major concern is that about 70 percent of the population lives in low-lying areas that Lessons and Recommendations would be vulnerable to inundation as a result of the combined effects of sea level rise, more severe storm Lesson 1: The cost of developing countries to adapt to surges, and flooding caused by heavier rainfall. climate change between 2010 and 2050 is estimated at US$70 billion to US$100 billion a year at 2005 prices. The EACC Samoa country study focuses on the This amounts to about “only� 0.2 percent of the projected implementation of design standards to ensure that GDP of all developing countries in the current decade buildings and other assets can cope with higher and at the same time to as much as 80 percent of total winds and more intense precipitation without dam- disbursement of ODA. age. It found that the adoption of more stringent design standards today would reduce the impact of The averages across all developing countries hide a very the climate change in future and the residual dam- uneven distribution of the burden of adaptation across age after adaptation (see Table ES-8). It also found regions as well as decades. Our estimates of the overall xxviii ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT cost of adaptation are 0.6-0.7 percent of GDP for Countries that reach the middle of the 21st century the Sub-Saharan Africa region in 2010-19 falling to with large shares of their populations engaged in sub- about 0.5 percent of GDP in 2040-49. In contrast, the sistence agriculture—with substantial illiteracy and equivalent �gures for the East Asia and Paci�c region lethargic or inept institutions—will be particularly are 0.13-0.19 percent in 2010-19 and about 0.07 per- vulnerable to the effects of climate change. Rapid cent in 2040-49. Apart from Sub-Saharan Africa, the development leads to a more flexible and resilient regions facing high relative costs of adaptation are the society, so that building human and social capital— Latin America and Caribbean region and (under the including education, social protection and health, and dry climate scenario) the South Asia region. In addi- skills training—are crucial to adaptation. tion, the absolute costs of adaptation increase over time and will certainly continue to increase after 2050. In all of our country studies, the burden of existing Our projections suggest that real GDP will increase climate variability is especially heavy in areas that have more rapidly than the costs of adaptation over the next high concentrations of poor and socially vulnerable four decades, even on quite conservative assumptions populations. Climate change exacerbates this pattern. about growth in GDP per person. However, it would In the Southern region of Bangladesh, for example, be unsafe to assume that this trend will continue into the rural poor are expected to face the largest declines the second half of the current century. in per capita consumption, declining productivity of the subsistence crops, and land losses owing to greater Lesson 2: Economic development is a central element of adap- salinity brought forth by sea level rise. tation to climate change, but it should not be business as usual. Lesson 4: Do not rush into making long-lived investments Economic development is the most basic and cost in adaptation unless these are robust to a wide range of effective method of adaptation, provided that it is climate outcomes or until the range of uncertainty about properly managed. It generates the resources and future weather variability and climate has narrowed. opportunities to adapt to climate change at a relatively Start with low-regret options. low cost by ensuring that the design and location of new infrastructure, buildings, and other assets take account For public policymakers, the fundamental prob- of the effects of climate change on their performance. lem is one of uncertainty regarding both climate Our country studies show that a failure to adapt to outcomes and longer–term projections of social climate change may lead to very large weather-related and economic development (such as anticipated losses—both in terms of the destruction of infrastruc- migration of people from rural areas to the cities). ture and foregone opportunities for future growth. This uncertainty is particularly large for patterns of In Ethiopia, robust growth based on infrastructure precipitation. Some country studies highlight cru- investment is the �rst line of defense against climate cial differences between alternative wet and dry sce- change impacts. In Bolivia, development measures narios and their effects on agricultural production, are not only robust to changes in climate impacts but water resources, and transport infrastructure. Other also help reduce them by increasing local resilience. countries show large variation in the magnitude of increased precipitation. As a result, countries should Lesson 3: Invest in human capital, develop competent try to delay adaptation decisions as much as possible and flexible institutions, focus on weather resilience and and focus on low-regret actions—those actions that adaptive capacity, and tackle the root causes of poverty. are robust under most climate scenarios. These are Eliminating poverty is central to both development and typically policies or investments that can be iden- adaptation, since poverty exacerbates vulnerability to ti�ed as priorities for development even without weather variability as well as climate change. climate change. For Africa, our studies show that ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT xxix expanding the road system and increasing the share Economic development has been accompanied by of paved roads would yield high returns by lowering a tendency for more rapid urban growth in coastal transport costs and expanding markets. They would areas than in inland cities. This may reflect relative also lessen flood impacts and enhance farmers’ abil- differences in transport costs as well as government ity to respond to changes in agricultural compara- policies or individual preferences. There will be many tive advantage. opportunities to reduce weather risks and associated costs by intelligent urban and land-use planning. Lesson 5: Adaptation to climate change should start with Whether in rural or urban areas, the rule of thumb the adoption of measures that tackle the weather risks is simple: whenever possible, ensure that growth and that countries already face, for example, more investment infrastructure take place in locations that are less in water storage in drought-prone basins or protection exposed to weather risks. The right incentives must against storms and flooding in coastal zones and/or urban also be adopted to discourage accumulating physical areas. Climate change will exacerbate these risks. capital in the shadow of dykes considered to be “safe.� As the New Orleans tragedy illustrated, a sufficiently Climate change will always hide beneath climate extreme event will breach a dyke. variability. Systems that effectively cope with existing climate variability will be more successful in adapt- Lesson 7: Hard and soft approaches to adaptation are two ing to future climate change than those that cannot. sides of the same coin. Good policies, planning, and insti- The short–term priority is to better prepare for the tutions are essential to ensure that more capital-intensive weather risks that countries are already facing. One measures are used in the right circumstances and yield the clear example concerns the impact of storms, espe- expected bene�ts. cially in coastal areas. Despite the uncertainty over future rainfall, there is relative certainty that warmer The distinction between “hard� (capital-intensive) climate will lead to rising sea levels and increased and “soft� (institutions and policies) adaptation is intensity of storms. With the inevitable increase easily exaggerated. There is no point in building the in urban populations, the costs of failing to protect best type of road in the wrong place, while the best coastal cities against major storms will rise rapidly. institutions will provide no protection against a storm At the same time, the de�ciencies of storm water that destroys buildings or power lines. The challenge drainage in coastal or inland cities already lead to is to get the balance between hard and soft adaptation avoidable—and sometimes large—losses caused by right. In some �eld sites in Vietnam, afforestation of urban flooding that have disproportionate effects on mangroves ranks above the infrastructure options the health and welfare of the poor. The Vietnam study such as sea dike repair, given the lower costs of man- suggests that it is important to enhance the capacities grove planting and the potential for this activity to be of agricultural and water systems to cope with cur- more pro-poor. In Ghana, a number of soft measures rent climate variability and build resilience into such are given priority over hard measures—including an systems from now on. Samoa similarly shows that upgrade of peri-urban slums and controlled develop- improving building codes to cope with current vari- ment of new ones, and the protection, management, ability makes infrastructure more climate-resilient. and sustainable use of coastal wetlands. Lesson 6: Beware of creating incentives that encourage development in locations exposed to severe weather risks. Where possible build future cities out of harm’s way— flood plains or coastal zones that are exposed to sea level rise and storm surges. One ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 1 Introduction governments of Bangladesh, Plurinational State of Context Bolivia, Ethiopia, Ghana, Mozambique, Samoa, and Vietnam—initiated the EACC study to esti- Under the Bali Action Plan adopted at the 2007 mate the cost of adapting to climate change. The United Nations Climate Change Conference, study, funded by the governments of the Nether- developed countries agreed to allocate “adequate, lands, Switzerland, and the United Kingdom, also predictable, and sustainable financial resources aims to help countries develop plans that incorpo- and new and additional resources, including offi- rate measures to adapt to climate change. cial and concessional funding for developing coun- try parties� to help them adapt to climate change Objectives (UNFCCC 2007) . The plan views international cooperation as essential for building capacity to The EACC study has two broad objectives: to integrate adaptation measures into sectoral and develop a global estimate of adaptation costs for national development plans. Yet studies on the informing the international community’s efforts in costs of adaptation offer a wide range of estimates, the climate negotiations, and to help decision mak- from $4 billion to $109 billion a year. A recent ers in developing countries assess the risks posed critique of estimates suggests that these may be by climate change and design national strategies substantial underestimates (Parry et al. 2009). for adapting to climate change. Similarly, National Adaptation Programmes of Action, developed by the least-developed coun- These two objectives complement each other. To tries under Article 4.9 of the United Nations some extent, however, they are also at odds with Framework Convention on Climate Change each other, and cannot be fully consistent: support- (UNFCCC), identify and estimate costs for only ing developing country efforts to design adaptation urgent and immediate adaptation measures. They strategies requires incorporating country-speci�c do not incorporate the measures into long-term characteristics and sociocultural and economic development plans. conditions into the analyses. Identifying the global costs of adaptation to climate change to support The Economics of Adaptation to Climate Change international negotiations requires analysis at a (EACC) study is intended to �ll this knowledge more aggregate level. Reconciling the two involves gap. Soon after the Bali Conference of Parties, a a tradeoff between the speci�cs of individual coun- partnership—comprising the World Bank and the tries and a global picture. 2 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT Approaches: the two parallel tracks conducted at the national level, and include analysis of three to �ve sectors, depending on the country. Figure To address the two objectives, the EACC was con- 1 below depicts the various EACC study components ducted on two parallel tracks: a global track, where and their links. national databases were used to generate aggregate estimates at a global scale; and a series of country- Given that climate change is a relatively new subject, level studies, where national data were disaggregated the numerous reports produced as part of the EACC to more local and sector levels, helping to understand global and country tracks, including this report, cover adaptation from the bottom-up perspective. The top- many technical areas—from climate science to social down and bottom-up approaches were compared and economic areas, as well as a number of sectors, and to the extent possible integrated. Some elements including agriculture, energy, water resources, infra- had to be analyzed separately, or solely, under each structure, and coastal zone management. Given the perspective. political importance of climate change, the �ndings of this and similar studies are highly relevant for policy The Synthesis Report making in both developed and developing countries. While aimed at a very broad audience, it is primar- This report is a synthesis of the global study report ily written for policy makers in developing coun- and seven country case study reports. The global study tries. Given their different objectives, related EACC consists of a number of sector studies, which were reports may be of interest to a diverse audience. These commissioned by the EACC project. Country case reports and background papers are available at www. study reports present �ndings from sector analyses worldbank.org/eacc. FIGURE 1 STUDY STRUCTURE: GLOBAL AND COUNTRY TRACKS EACC SYNTHESIS REPORT GLOBAL STUDY 7 COUNTRY CASE STUDIES BANGLADESH BOLIVIA ETHIOPIA GHANA MOZAMBIQUE SAMOA VIETNAM Agriculture Sector Sector Sector Sector Sector Sector Sector Water Sector 2 Sector 2 Sector 2 Sector 2 Sector 2 Sector 2 Sector 2 SECTORS SECTORS Health Sector 3 Sector 3 Sector 3 Sector 3 Sector 3 Sector 3 Sector 3 Sector 4 Sector 4 Sector 4 Sector 4 Sector 4 Sector 4 Sector 4 Coastal ... others ... others ... others ... others ... others ... others ... others Infrastructure ...Others ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 3 This report presents a synthesis of the methodology The remainder of the report comprises of four chap- and results derived from research conducted for the ters. The next chapter presents the concepts and meth- EACC global and country study tracks in Bolivia, odology used for analyses in both the global and the Bangladesh, Ethiopia, Ghana, Mozambique, Viet- country case studies. This chapter also describes some nam, and Samoa. It provides information on lessons of the study’s limitations. Chapter III introduces the learned and insights gained on adaptation to climate results from the global analysis, while chapter IV change from global, country, and sector-level analyses. focuses on results from the country analyses. Chapter It develops a robust, integrated approach for increas- V presents lessons learned and recommendations for ing resilience to climate risks, and presents recom- policy makers. mendations to help policy makers set priorities. 4 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT Two ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 5 CONCEPTS, METHODOLOGY, AND LIMITATIONS sectors and countries offset, if at all, adaptation costs Concepts in another sector or country. This multiyear study is based on a number of crucial How much to adapt concepts, including the de�nition of adaptation costs; understanding the links between adaptation and Adaptation is clearly not a rigid set of actions, and development; and dealing with the inherent uncer- governments can choose the amount of, or level of, tainty in climate projections and climate impacts. adaptation. One possibility is to adapt completely, so that society is at least as well off as it was before cli- Adaptation costs mate change. At the other extreme, countries could choose to do nothing, experiencing the full impact One of the biggest challenges of the study has been of climate change. In the intermediate cases, coun- to develop a usable de�nition of adaptation costs. The tries invest in adaptation using the same criteria as concept is intuitively understood as the costs societ- for other development projects—until the marginal ies incur to adapt to changes in climate. The IPCC bene�ts of the adaptation measure just exceed the de�nes adaptation costs as the costs of planning, pre- costs. This leads to a portfolio of adaptation actions paring for, facilitating, and implementing adaptation that either restores social welfare relative to a baseline measures, including transaction costs. without climate change, or leaves some amount of residual damage from climate change. This de�nition is difficult to implement in practice, however. For one, “development as usual� needs to How much to adapt is consequently an economic be conceptually separated from adaptation. That problem—how to allocate resources to adapt to cli- requires deciding whether the costs of development mate change while also meeting other needs. And initiatives that enhance climate resilience ought to be therein lies the challenge. Poor urban workers who counted as part of adaptation costs. It also requires live in a fragile slum dwelling might �nd it difficult deciding how to incorporate in those costs the adap- to decide whether to spend money to make their liv- tation de�cit, de�ned as countries’ inability to deal ing quarters less vulnerable to more intense rainfall, with current and future climate variability. It requires or to buy school books or �rst-aid equipment for de�ning how to deal with uncertainty about climate their family—or how to allocate between the two. projections and impacts. And it requires specifying Poor rural peasants might �nd it difficult to choose how potential bene�ts from climate change in some between meeting these basic education and health 6 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT needs and some simple form of irrigation to compen- This study mostly uses the de�nition of adaptation sate for increased temperatures and their impact on costs where the objective is to restore welfare. agricultural productivity. These examples suggest that desirable and feasible levels of adaptation depend on Links between adaptation and development both available income and other resources. Economic development is perhaps the best hope for The de�nition in practice adaptation to climate change: development enables an economy to diversify and become less reliant on Corresponding to a chosen level of adaptation is sectors such as agriculture that are more vulnerable an operational de�nition of adaptation costs. If the to the effects of climate change. Development also policy objective is to adapt fully, the cost of adapta- makes more resources available for abating risk. And tion can be de�ned as the minimum cost of adapta- often the same measures promote development and tion initiatives to restore welfare to levels prevailing adaptation. For example, progress in eradicating before climate change. Restoring welfare may be pro- malaria helps countries develop and also helps societ- hibitively costly, however, and policy makers may opt ies adapt to the rising incidence of malaria that may for an efficient level of adaptation instead. Adapta- accompany climate change. tion costs would then be de�ned as the cost of actions that satisfy the criterion that their marginal bene�ts At the same time, adaptation to climate change is exceed their marginal costs. Because welfare would essential for development: unless agricultural societ- not be fully restored, there would be residual damage ies adapt to changes in temperature and precipitation from climate change after allowing for adaptation. (through changes in cropping patterns, for example), ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 7 development will be delayed. Finally, adaptation less capacity to adapt to change, whether induced by requires a new type of climate-smart development climate change or other factors, because of their lower that makes countries more resilient to the effects of stage of development. A country’s adaptive capacity climate change. Urban development without atten- is thus expected to increase with development. This tion to drainage, for example, will exacerbate the meaning is perhaps better captured by the term devel- flooding caused by heavy rains. opment de�cit. These links suggest that adaptation measures range The adaptation de�cit is important in this study for from discrete adaptations, or interventions for which establishing the development baseline from which to adaptation to climate change is the primary objec- measure the independent additional effects of climate tive (WRI 2007); to climate-smart development, or change. Because the adaptation de�cit deals with cur- interventions to achieve development objectives that rent climate variability, the cost of closing the de�- also enhance climate resilience; to development that cit is part of the baseline and not of the adaptation is not business-as-usual, as compared to interventions costs. In practice, the distinction is difficult to apply that can exacerbate the impacts of climate change because the costs of addressing current climate vari- and that therefore should not be undertaken. The Bali ability and future climate change are often intermin- Action Plan calls for “new and additional� resources gled. Hence, most parts of this study do not attempt to meet adaptation costs. This study therefore de�nes to estimate the costs of adaptation by reference to a adaptation costs as additional to the costs of develop- baseline under which the adaptation de�cit has been ment. So, the costs of measures that would have been closed. Two exceptions are adaptation to sea level rise undertaken even in the absence of climate change are in Bangladesh and to changes in extreme weather not included in adaptation costs, while the costs of events in Samoa. doing more, doing different things, and doing things differently are included. It is not simple to assess how the use of alternative baselines in which the adaptation de�cit has or has The adaptation de�cit not been closed will affect estimates of the costs of adaptation. For infrastructure, for example, clos- The separation of adaptation from development costs ing the adaptation de�cit will usually mean that a is linked to the concept of the adaptation de�cit, larger stock of infrastructure assets has to be climate- which captures the notion that countries are under- proofed, thus implying an increase in adaptation prepared for current climate conditions, much less costs. In contrast, closing the adaptation de�cit in for future climate change. Presumably, these short- agriculture might imply a higher reliance on irrigated falls occur because people are under-informed about agriculture, which may reduce the extent of adapta- climate uncertainty and therefore do not rationally tion required as a consequence of changes in rainfall allocate resources to adapt to current climate events. patterns due to climate change. Adaptation costs are The shortfall is not the result of low levels of develop- likely to be reduced in the agricultural sector as a ment but of less than optimal allocations of limited result. The practical difficulty of separating the costs resources resulting in, say, insufficient urban drainage of closing the adaptation de�cit from the costs of infrastructure. The cost of closing this shortfall and adaptation for carefully de�ned baseline means that bringing countries up to an “acceptable� standard for the estimates reported in this study tend to over-state dealing with current climate conditions given their the “true� costs of adaptation costs, though by differ- level of development is one de�nition of the adapta- ent degrees for different sectors and countries. As an tion de�cit. The second use of the term, perhaps more illustration, adaptation measures in some of the coun- common, captures the notion that poor countries have try studies actually generate bene�ts for some climate 8 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT scenarios that increase total welfare rather than just monsoon or the El Niño Southern Oscillation are restoring it. This is because they combine adaptation even more uncertain. Consequently, decisions about with development measures that would be justi�ed investments in assets having a useful life of 20, 30, or even without climate change. even 40 years—such as dams, dikes, urban drainage, bridges, and other infrastructure—have to be based A typology of adaptation measures on incomplete information with a large variance in projections of future climate conditions. The second Proactive and reactive measures: Reactive mea- major uncertainty concerns economic growth. Faster sures will be the dominant response until threats economic growth puts more assets at risk in absolute become better understood. But countries can terms, but higher levels of investment and technical become more proactive in disaster preparedness. change mean that countries have greater flexibility The frequent cyclones and extreme coastal events to absorb and respond to climate-induced changes in Bangladesh, for example, have led the country in productivity and other climate shocks. Recent to greatly improve its early warning systems, and experience shows that predicting economic growth is the number of deaths from such events has sig- a fragile science, while projecting how technological ni�cantly decreased. change may affect adaptation over the next 40 years is nearly impossible. Soft and hard measures: Many soft measures— such as water and energy pricing, strengthening Uncertainty about climate outcomes property rights, and flood plain and landslide area zoning—have robust adaptation and development The EACC study—both the country and the global results. But they take time and require strong tracks—calculates adaptation costs as if decision institutions to put in place. The timing may be makers know with certainty what the future climate consistent with the time frame of global warm- will be. This must be complemented by considering ing, however, if concerted action begins now. how to maximize the flexibility of investment pro- grams to take advantage of new climate knowledge as Public and private adaptation: Adaptation mea- it becomes available. For most countries and sectors, sures can be classi�ed by the types of economic the study was able to identify policies and investments agent initiating the measure—public or private. that generate good outcomes over the range of the The literature distinguishes between autonomous wettest and driest climate scenarios considered. But or spontaneous adaptation (by households and these scenarios could not encompass the full range communities acting on their own without public of possible outcomes. Of the 26 climate projections interventions but within an existing public policy available for the A2 SRES, an assessment of adapta- framework) and planned adaptation (from a tion costs was feasible only for two projections for the deliberate public policy decision). global track and for two-to-four projections for the country studies. Further, climate models are evolving Dealing with uncertainty all the time, so it is inevitable that projections made in 2012 will differ from those made in 2008. Uncertainty complicates the analysis of adaptation to climate change in three different ways. First, for A good faith effort has been made to examine the most countries there is no consensus whether future wettest and driest scenarios available for each situ- climate will be wetter or drier, or how the frequency ation. This range is simply a snapshot of the state and severity of major storms will change. Changes of climate science when this study was undertaken in regional climate phenomena such as the Asian and does not reflect any view of the distribution of ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 9 climate outcomes in future as scienti�c models and growth in population, GDP per capita, and urbaniza- other information change. Because the range of both tion, which drive the demand for food, investment in climate and economic uncertainty tends to grow infrastructure, the bene�ts of protecting coastal zones, exponentially over time, the study examines expen- and so on. How would the costs of adaptation change ditures up to 2050 and limits the scope to adaptation with a different trajectory? Alternative assumptions to what may be broadly interpreted as the public sec- about population and economic growth have only a tor. The major impacts of climate change, such as the slight impact on estimates of the cost of adaptation in melting of ice sheets, are likely to occur after 2050, 2010–19, so the margins of error associated with the but the degree of uncertainty after this date requires development baseline are not very important in the a quite different approach to quantifying the costs of immediate future, although they will grow over time. adaptation.10 The United Nations publishes alternative population Hierarchy of uncertainty and timing projections that rely on different assumptions about fertility decline in developing countries. The varia- Rational resource allocation must take into account tion in population forecasts for developing countries the degrees of uncertainty about the nature and tim- in 2050 is approximately +/–14 percent for alterna- ing of climate outcomes. When will it be optimal to tive fertility assumptions. The United Nation’s central start building sea walls in coastal areas and how high projection has consistently been revised downward should they be? The uncertainties concern the rate over the last two decades as fertility rates have fallen of future sea level rise, the potential height of storm faster than anticipated. Thus, the plausible range of surges, the damage that may be caused to agricultural uncertainty might be +/–10 percent. The range of or urban assets in the coastal zone, and the future uncertainty for growth in GDP per capita is larger, cost of upgrading sea walls. These must be balanced ranging from –26 percent to +40 percent in 2050. The against the need to allocate resources that could be variation for developing countries is even larger— used to meet other social or economic goals instead. from –40 percent to +50 percent – so the range of Further, there may be alternative measures which variation in total GDP might be –45 percent to +75 are less expensive than sea walls but which provide percent, a huge margin of uncertainty. These errors are temporary or less effective protection. Developing a compounded by the con�dence intervals of projec- strategy involves choices about the selection, scale tions of demand as functions of population and GDP and timing of adaptation actions when both climate per capita. On this basis, it is very difficult to calculate change and economic development are uncertain. A potential margins of error in the estimates of the costs model is needed that allows governments to prioritize of adaptation. Yet, there is nothing unique in the pro- and sequence adaptation strategies in a �nancially cedures adopted here. They are assumptions widely constrained environment and which takes account of adopted in similar exercises. The very same uncertain- social, institutional, and cultural factors (Kellerer et ties apply in the analyses of all economic sectors that al. 2004).11 A pilot exercise was carried out at a sub- have such extended time horizons. basin water level in Bolivia. Future technologies Economic forecasts With the exception of agriculture this study does A key contribution of this study is to separate the not allow for the unknowable effects of innovation costs of adaptation from those of development by and technical change on adaptation costs. Hence, the de�ning an explicit development baseline. The study reported costs are based on what is known today rather assumes just one future development path, based on than what might be possible in 20–40 years. Sustained 10 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT growth in per capita GDP for the world economy over 40 years that can be used as a basis of compari- rests on technical change, which is likely to reduce son with the climate change scenario. the real costs of adaptation over time. The exclusion of technical change is one factor that imparts an Using a time frame of 2050, development baselines upward bias to the reported estimates of the costs of are �rst developed for each sector using a common adaptation. In the case of agriculture, growth in agri- set of GDP and population forecasts for 2010–50.12 cultural productivity, based on historical trends and The population trajectory is aligned with the United expert opinion, is built into the IMPACT model, and Nations Population Division’s middle-fertility pro- explicit account is taken of investment in agricultural jections for 2006. To ensure consistency with emis- research as an element of the cost of adaptation. sions projections, the GDP trajectory is based on the average of the GDP growth projections of the three Methodology major integrated assessment models of global emis- sions growth—Climate Framework for Uncertainty, Critical methodological issues include establishing Negotiation, and Distribution (FUND; Anthoff and a baseline, choosing climate projections, predicting Tol 2008); PAGE2002 (Hope 2006); and Regional impacts, simplifying assumptions, �ne-tuning the Dynamic Integrated Model of Climate and the methodological approaches of the global and country Economy (RICE99; Nordhaus 2002)—and growth tracks, and introducing a social component into the projections used by the International Energy Agency country track. and the U.S. Energy Information Administration to forecast energy demand. All these sources provide The baseline growth estimates at a regionally disaggregated level. To estimate the impacts of climate change and then The global average annual real GDP per capita the costs of adaptation, it is necessary to compare, for growth rate constructed in this way is 2.1 percent, each time period, the difference between the world similar to global growth rates assumed in the United with climate change and the world without climate Nations Framework Convention on Climate Change change. To do this, we �rst have to project what the (UNFCCC) A2 emissions scenario. From the base- world will look like between now and 2050, our plan- lines, sector-level performance indicators (such as ning horizon. This projected world without climate stock of infrastructure assets, level of nutrition, and change is the baseline. It is a reasonable trajectory water supply availability) are determined. for the growth and structural change of the economy ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 11 Choosing climate projections necessarily the wettest and driest in every location) based on the climate moisture index. Climate scenarios were chosen to capture the larg- est possible range of model predictions. Although Predicting impacts model predictions do not diverge much in projected temperature increases by 2050, precipitation changes The changes in climate are used to predict what the vary substantially across models. For this reason, world would look like under the new climate con- model extremes were captured by using the model ditions without and with adaptation. This meant scenarios that yielded extremes of dry and wet cli- translating the impacts of changes in climate on the mate projections, although catastrophic events were various economic activities (agriculture, �sheries), on not captured. Among the models reporting minimum people’s behavior (consumption, health), on environ- and maximum temperature changes, NCAR was the mental conditions (water availability, oceans, forests), wettest and CSIRO the driest scenario (globally, not and on physical capital (infrastructure). FIGURE 2 STUDY METHODOLOGY: GLOBAL AND COUNTRY TRACKS Global Track Country Track Projections Projections Climate Climate Sub-national Global Data Sets Data Sets Water Run-off Water Run-off Baseline GDP/Population Baseline GDP/Population Economic, Social and Economic, Social and Environmental Impacts Environmental Impacts Participatory Scenarios w/vunerable Identi�cation of Adaptation Identi�cation of Adaptation groups Measures Measures Decision Rule Decision Rule Cost of Adaptation Cost of Adaptation Sectors Cross-Sectors Agriculture Coastal Zones National Macroeconomic Forestry Extreme Weather Events Analysis Fisheries Infrastructure Water Resources Sectors Cross-Sectors Health Agriculture Coastal Zones Ecosystem Services Extreme Weather Events Forestry Fisheries Infrastructure Water Resources 12 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT Simplifying assumptions important information for national level deci- sion making. Second, the country track featured a Analyses of complex decisions under uncertainty social component in six of the seven country case easily become intractable. The various dimensions studies. As a companion piece to the EACC Syn- of climate change combined with many countries, thesis Report, an EACC-Social Synthesis Report sectors and agents reinforce this danger. Hence, has been produced that presents the findings of it is necessary to adopt a variety of assumptions the social component, which was conducted in all whose purpose is to ensure that useful results can case study countries except Samoa. be generated from the models. First, it is assumed that policy makers know what the future climate Social analysis will be and act to prevent its damages. Second, in costing the adaptation options, the study focuses The social component in the country track on hard options (building dams, dykes) and not focuses on preferred adaptation strategies from a soft options (early warning systems, community bottom-up, local-level perspective. The method- preparedness programs, watershed management, ology involved a combination of analytical meth- urban and rural zoning). This approach was delib- ods, including participatory scenario development erately chosen because the former options can be (PSD) workshops to reveal local stakeholders’ assessed and their costs estimated, not because assessments of adaptation pathways in the context soft options are less important. In reality, soft of uncertainty. In the workshops, participants rep- options should be adopted whenever they are less resenting the interests of vulnerable groups iden- expensive than hard options. Third, the adaptation tified preferred adaptation options and sequences costs are based on current knowledge. This implic- of interventions based on local and national cli- itly assumes that there will be no future innova- mate and economic projections. The findings on tion and technical change beyond current trends, what forms of adaptation various groups consider except in the agricultural sector. But we know that to be most effective—including soft adapta- economic growth and thus development depend tion options such as land use planning, greater on technical change, which is likely to reduce the public access to information, and institutional real costs of adaptation over time. All of these capacity building—have implications for the assumptions give an upward bias to the estimates costs of adaptation. of the adaptation costs. We return to these points in the limitations discussion. In addition, the social component generated new evidence on how vulnerability is socially differen- Differences between the global and tiated; identified the risks and benefits of adapta- country tracks tion options for a range of actors in an integrated and cross-sectoral manner; and highlighted the The steps identified apply to both the global and importance of social accountability and good gov- country tracks. But given their different objectives, ernance for achieving pro-poor, climate-resilient two steps of the methodologies differ. First, for development. The focus of the EACC-social most country studies a macroeconomic modeling analysis went beyond planned adaptation and framework was used allowing for the analysis of considered the potential of autonomous forms macroeconomic and cross-sectoral effects of the of adaptation undertaken by households, NGOs, impacts and adaptation to climate change. This and the private sector to inform future adapta- integrated approach has been less successful in tion planning. This approach was not viable in the generating lessons at the sector level, but provides global track. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 13 Limitations plus political factors: for example, how to influence the location of people away from high-risk or increas- The EACC study makes use of mathematical tools, ingly unproductive areas, how to improve the alloca- which impose intellectual discipline. Examples of this tion of water and land, or how to improve the quality discipline are the use of a well-de�ned baseline and of education. The goal of this study was to focus on the requirement under CGE models that the national the economics of adaptation, but this approach pres- income accounting identities balance at the end of ents only part of a much larger story. each year. This approach is required to provide a quan- titative evaluation of costs and bene�ts. The models Similarly, previous work in each country influenced can be used to assess the relative importance of dif- both the direction of EACC research and what it has ferent factors and the marginal impacts of changes been possible to accomplish.13 Where researchers, in speci�c policy variables on outcomes. Such analy- data, and models already existed, the EACC project sis provides an essential foundation for formulating naturally built upon prior work. The consequence is policies and making decisions. Nonetheless, the usual that the level and detail of the study’s modeling and limitations of relying upon econometric and other analysis varies across sectors and countries. In most mathematical models apply. cases, this reflects the relative importance that coun- tries and analysts have attached to different kinds of Path dependency. Formal models can encourage a adaptation. focus on questions that are amenable to analysis by the model at the expense of less tractable but perhaps Important limitations of this study fall under more important issues. Adaptation to climate change categories: (1) institutions, and (2) modeling involves responses that depend upon institutional or limitations. The crucial issue of uncertainty and the cultural factors or, more likely, a combination of these limitations it poses has already been discussed above. 14 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT Institutions attempted to identify opportunities for soft adaptation without trying to cost them. There is an additional Difficulty in addressing. From the outset the EACC consideration: the country studies suggest that draw- study did not attempt to incorporate institutional, ing a distinction between (a) what are good develop- political, and cultural factors in the analysis of adapta- ment policies, and (b) additional measures to adapt to tion costs. Without question, these factors are crucial climate changes is difficult under the best of circum- in understanding the process of adaptation and deter- stances. Hard adaptation measures can be identi�ed mining what is feasible as opposed to what might be and their costs estimated, whereas soft adaptation is desirable from an economic perspective. But there generally a matter of doing things that would be desir- was a clear tradeoff between extending the scope of able even in the absence of climate change. Sometimes, the study and ensuring that the economics of adapta- the focus has to be shifted or policies redesigned to tion could be examined in sufficient detail.14 take account of climate change, but it is rarely feasible to separate adaptation from development. Some types of adaptation are best implemented through effective collective action at the community Migration. One concern that is often expressed is level. “Soft� adaptation measures—early warning sys- that climate change will lead to substantial amounts tems, community preparedness programs, promoting of intra- or inter-country migration, which will education, and capacity building—require strong gov- imply substantial public expenditures to meet the ernance to be effective. If this can be achieved, they needs of migrants in their new places of residence. may go a long way in reducing vulnerability to climate Recent work suggests that social processes linked to change. However, estimating the costs of implement- poverty and marginality as well as the treatment of ing such options is difficult for individual countries and migrants may be more important determinants of the impractical at a global level. The global study focused amount and consequences of migration than envi- “hard� adaptation measures, while the country studies ronmental change (Barnett and Webber 2010). Good ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 15 development policies to reduce poverty and enhance sufficiently large number of measures/strategies and social inclusion are essential without any consideration compare their results. This has been the strategy of of climate changes, so the additional element of adap- this study. However, collaboration with govern- tation is a small part of a larger picture. If such policies ment has meant that for each country case study the are not implemented, environmental change may be �rst priority has been to simulate the government’s an important proximate factor in migration decisions preferred adaptation strategy. Because of time and leading to substantial costs of adaptation as a conse- resources constraints, and at times the reluctance on quence of wider policy failures. the part of government authorities to explore strate- gies outside the approved plan, few alternative strate- Modeling gies have been explored to date. The study has estimated the additional public sector These quali�cations do not mean that the study ignored (budgetary) costs that will be required for countries efficient adaptation. For each of the sectoral and coun- to adapt to climate change.15 Governments achieve try studies, a serious attempt was made to apply rules of climate adaptation at lowest cost when (a) they use thumb or other criteria that identify low cost, though cost-bene�t criteria to choose the most efficient proj- probably not least cost, strategies for adaptation. In any ects to meet the overall goal, and (b) they sequence case, an optimal investment program for adaptation in projects to maximize the net present value of their a country or for a sector is difficult to de�ne, let alone expected future investment streams. The models used calculate, when there is so much uncertainty about for this study cannot meet these efficiency conditions future climate and economic development. and therefore do not ensure adaptation at least cost. There is an important choice that has to be made None of the sector models used in the global study is when thinking about future work on adaptation. capable of choosing the best pro�le of government’s One approach would be to focus on efficient adapta- investment through time (inter-temporal optimiza- tion either by the use of optimization models across tion). Some of the models in the country studies sectors and over time, or by comparing the results have such capability, but most do not. In any case, of a wide range of alternative investment programs, inter-temporal optimization is difficult assuming including those that implement projects at differing certainty and is nearly unmanageable in a stochastic points in the future. An alternative approach would framework. In addition, sectoral adaptation plans be to look for robust rules of thumb that yield reason- were identi�ed independently in most cases for able or good adaptation strategies across a wide range both the global and the country studies. Identifying of climate outcomes and economic conditions. In whether the resources invested in one sector would view of the uncertainties about climate and economic have yielded higher adaptation bene�ts in another development as well as the limited information avail- sector, or whether cash transfers would maintain able to models, the second approach seems likely to welfare at lower cost, was beyond the capacity of be the better way forward in the immediate future. this exercise.16 Several of the country studies CGE models calculated the economywide effect of spe- There are two issues on which the economic frame- ci�c sectoral adaptation measures, but the adapta- work used to examine adaptation requires additional tion strategies themselves were not optimized, either work. The �rst issue concerns the treatment of ecosys- cross-sectorally or inter-temporally. tem services. Some of the services of ecosystems that are used as indirect inputs to the production of market One method to overcome the temporal and cross- goods and services were included implicitly or explic- sectoral limitations of models is to construct a itly in the sector. However, the role of ecosystems 16 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT such as coastal and inland wetlands in providing Nonetheless, the social component was invaluable as both nonmarket services—including protection from a complement to the quantitative analysis in assessing droughts, floods, or storms—and cultural or recre- the consistency of adaptation measures viewed from ational bene�ts was not addressed. Additional work national and local perspectives. is needed on flood protection services of wetlands other than mangroves and on the potential for using Future Work mangroves as an adaptation measure. With respect to biodiversity, it is difficult to separate the effects of As this section has emphasized, the EACC study has climate change from those of more general economic been a preliminary attempt to understand the eco- development. Even if that can be done, little is known nomic issues that arise in identifying and implement- about what adaptation measures are effective for pre- ing measures to adapt to climate change. The study serving biodiversity. has highlighted the wide range of uncertainty that hamper any attempt to draw immediate and speci�c The second issue is how to combine social analyses conclusions about the best policies and investments with the economic models. The original intent of the for adaptation. EACC was to translate the very rich, mostly qualita- tive information from �eld work into economic terms, It is sometimes self-serving to emphasize the need so that the adaptation measures indicated by the local for more detailed studies and further research. This populations could be included in the economic analy- is not the case when dealing with climate change. sis as explicit adaptation alternatives. This approach The highest priority in the immediate future must proved to be unworkable. Among the difficulties were be to reduce the range of uncertainty about future (a) the level of effort required to obtain the necessary climate impacts and to identify forms of adaptation economic information, (b) problems in scaling up that are robust across the range of uncertainty that very speci�c local and soft measures for incorporation will remain. Table 1 identi�es areas in which future in national models, and (c) the high degree of overlap work can contribute to this process of reducing between what local communities saw as immediate uncertainty. It is fundamentally based on the study’s development priorities and adaptation measures. limitations. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 17 TABLE 1 RECOMMENDATIONS FOR FUTURE WORK STUDY LIMITATION RECOMMENDATION Use of mathematical models Include institutional, social, cultural, and political perspectives to identify good policies Climate uncertainty Consider more scenarios, Monte Carlo simulations, and other probabilistic approaches Growth uncertainty Hard to improve other than through sensitivity analyses Technological uncertainty Incorporate better information from sector specialists and simulate the impact of potential advances Non-consideration to institu- Context-speci�c institutional capacity has to be assessed and considered to tional issues make recommendations realistic and feasible Limited focus on migration and Work with outside projections; limited current knowledge on cities and urbanization climate change Models not worked on Improve models to include inter-temporal, cross-sectoral, and cross-regional ef�ciency ef�ciency Limited range of adaptation Include a broader range of strategies No environmental services Pull better information and introduce more consistent estimates Integration with local, bot- Better understand economics of local actions tom-up perspectives 18 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT Three ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 19 Results From the Global Analyses But it is still a very low percentage (0.17 percent) Putting a price tag on adaptation of the wealth of countries (measured by their GDP, which was roughly $60 trillion in 2009). Overall, the study estimates that the cost between 2010 and 2050 of adapting to an approximately 2oC Total adaptation costs calculated by the gross sum warmer world by 2050 is in the range of $70 billion17 method average roughly $10–$15 billion a year more to $100 billion a year (Table 2). This sum is the same than the other two methods (the insigni�cant dif- order of magnitude as the foreign aid that developed ference between the X-sum and net sum �gures is countries now give developing countries each year. largely a coincidence). The difference is driven by TABLE 2 TOTAL ANNUAL COSTS OF ADAPTATION FOR ALL SECTORS, BY REGION, 2010–50 (2005 $ billions, no discounting) Europe & Latin Middle East Cost aggregation East Asia & Central America & & North South Sub- type Paci�c Asia Caribbean Africa Asia Saharan Africa Total National Centre for Atmospheric Research (NCAR), wettest scenario Gross sum 25.7 12.6 21.3 3.6 17.1 17.1 97.5 X-sum 21.7 11.2 18.7 2.4 12.4 15.1 81.5 Net sum 21.7 11.1 18.7 2.3 12.3 14.9 81.1 Commonwealth Scienti�c and Industrial Research Organization (CSIRO), driest scenario Gross sum 20.1 8.1 17.9 3.5 18.7 16.4 84.8 X-sum 17.9 6.9 14.8 2.5 15.0 14.1 71.2 Net sum 17.7 6.5 14.5 2.4 14.6 13.8 69.6 Notes: (a) The gross aggregation method sets negative costs in any sector in a country to zero before costs are aggregated for the country and for all developing countries. The X-sums method nets positive and negative items within countries but not across countries and includes costs for a country in the aggregate, as long as the net cost across sectors is positive for the country. The net aggregate measure nets negative costs within and across countries. (b) NCAR is The National Center for Atmospheric Research, and CSIRO is the Commonwealth Scienti�c and Industrial Research Organisation. Source: World Bank 2010a. 20 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT countries that appear to bene�t from climate change in infrastructure and coastal zones; for Sub-Saharan in the water supply and flood protection sector, espe- Africa, water supply and flood protection and agri- cially in East Asia and Paci�c and in South Asia. culture; for Latin America and the Caribbean, water supply and flood protection and coastal zones; and The drier scenario (CSIRO) requires lower total adap- for South Asia, infrastructure and agriculture. tation costs than does the wetter scenario (NCAR), largely because of the sharply lower costs for infra- Not surprisingly, both climate scenarios show costs structure, which outweigh the higher costs for water increasing over time, although falling as a percent- and flood management. In both scenarios, infrastruc- age of GDP—suggesting that countries become less ture, coastal zones, and water supply and flood protec- vulnerable to climate change as their economies grow tion account for the bulk of the costs. Infrastructure (Figures 4 and 5). There are considerable regional adaptation costs are highest for the wetter scenario. variations, however. Adaptation costs as a percent- age of GDP are considerably higher in Sub-Saharan On a regional basis, for both climate scenarios, the Africa than in any other region, in large part because East Asia and Paci�c Region bears the highest adap- of the region’s lower GDP, but also due to higher tation cost, and the Middle East and North Africa costs of adaptation for water resources (not shown) the lowest. Latin America and the Caribbean and driven by changes in patterns of precipitation. Sub-Saharan Africa follow East Asia and Paci�c in both scenarios (Figure 3). On a sector breakdown, Turning to the EACC analyses of sectors and extreme the highest costs for East Asia and the Paci�c are events, the �ndings offer some insights for policy FIGURE 3 TOTAL ANNUAL COST OF ADAPTATION AND SHARE OF COSTS, NCAR AND CSIRO SCENARIOS, BY REGION ($ billions at 2005 prices, no discounting) NCAR CSIRO 15.1 14.1 17.9 18% 21.7 20% 27% 25% 12.4 15% 21% 10% 14% 15 6.9 3% 11.2 2.4 23% 3% 21% 2.5 14.8 18.7 EAP ECA LAC MENA SAS SSA Note: EAP is East Asia and Paci�c, ECA is Europe and Central Asia, LAC is Latin America and Caribbean, MNA is Middle East and North Africa, SAS is South Asia, and SSA is Sub-Saharan Africa. Source: World Bank 2010a. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 21 FIGURE 4 TOTAL ANNUAL COST OF ADAPTATION FOR THE NATIONAL CENTRE FOR ATMOSPHERIC RESEARCH (NCAR) SCENARIO, BY REGION AND DECADE ($ billions at 2005 prices, no discounting) 30 25 20 EAP ECA US Billions 15 LAC MNA 10 SAS 5 SSA 0 2010-19 2020-29 2030-39 2040-49 Years Note: EAP is East Asia and Paci�c, ECA is Europe and Central Asia, LAC is Latin America and Caribbean, MNA is Middle East and North Africa, SAS is South Asia, and SSA is Sub-Saharan Africa. Source: World Bank 2010a. FIGURE 5 TOTAL ANNUAL COST OF ADAPTATION FOR THE NATIONAL CENTRE FOR ATMOSPHERIC RESEARCH (NCAR) SCENARIO, BY REGION AND DECADE ($ billions at 2005 prices, no discounting) 0.70% 0.60% Costs as percent of GDP 0.50% 2010-19 0.40% 2020-29 0.30% 2030-39 0.20% 2040-49 0.10% 0.00% EAP ECA LAC MNA SAS SSA World Bank Region Note: EAP is East Asia and Paci�c, ECA is Europe and Central Asia, LAC is Latin America and Caribbean, MNA is Middle East and North Africa, SAS is South Asia, and SSA is Sub-Saharan Africa. Source: World Bank 2010a. 22 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT makers who must make tough choices in the face of activity, yet they are also subject to a number of great uncertainty. climate risks, including sea level rise and possible increased intensity of tropical storms and cyclones. Infrastructure.18 This sector has accounted for the larg- These factors make adaptation to climate change est share of adaptation costs in past studies and takes critical. The EACC study shows that coastal adapta- up a major share in the EACC study—in fact, the big- tion costs are signi�cant and vary with the magnitude gest share for the NCAR (wettest) scenario because the of sea-level rise, making it essential for policymakers adaptation costs for infrastructure are especially sensitive to plan while accounting for the uncertainty. One of to levels of annual and maximum monthly precipita- the most striking results is that Latin America and the tion. Urban infrastructure—drainage, public buildings, Caribbean and East Asia and the Paci�c account for and similar assets—accounts for about 54 percent of the about two-thirds of the total adaptation costs. infrastructure adaptation costs, followed by railways at 18 percent, and roads (mainly paved) at 16 percent. East Water supply.20 Climate change has already affected Asia and the Paci�c and South Asia face the highest the hydrological cycle, a process that is expected to costs, reflecting their relative populations. Sub-Saharan intensify over the 21st century. In some parts of the Africa experiences the greatest increase over time, with world, water availability has increased and will con- its adaptation costs rising from $0.9 billion a year for tinue to increase, but in other parts, it has decreased 2010–19 to $5 billion a year for 2040–49. and will continue to do so. Moreover, the frequency and magnitude of floods are expected to rise, because Coastal zones.19 Coastal zones are home to an ever- of projected increases in the intensity of rainfall. growing concentration of people and economic Accounting for the climate impacts, the study shows ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 23 that water supply and flood management ranks as one Extreme weather events.23 Without reliable data of the top three adaptation costs in both the wetter on emergency management costs, the EACC study and drier scenarios, with Sub-Saharan Africa foot- tries to shed light on the role of socioeconomic ing by far the highest costs. Latin America and the development in increasing climate resilience. It Caribbean also sustain high costs under both models, asks: As climate change increases potential vulner- and South Asia sustains high costs under CSIRO. ability to extreme weather events, how many addi- tional young women would have to be educated to Agriculture.21 Climate change affects agriculture by neutralize this increased vulnerability? And how altering yields and changing areas where crops can much would it cost? The �ndings show that by be grown. The EACC study shows that changes in 2050, neutralizing the impact of extreme weather temperature and precipitation from both climate sce- events requires educating an additional 18 million narios will signi�cantly hurt crop yields and produc- to 23 million young women at a cost of $12 billion tion—with irrigated and rainfed wheat and irrigated to $15 billion a year. For 2000–50, the tab reaches rice the hardest hit. South Asia shoulders the biggest about $300 billion in new outlays. This means that declines in production, but developing countries fare in the developing world, neutralizing the impact worse for almost all crops compared to developed of worsening weather over the coming decades countries. Moreover, the changes in trade flow patterns will require educating a large new cohort of young are dramatic. Under the NCAR scenario, developed women at a cost that will steadily escalate to sev- country exports increase by 28 percent, while under eral billion dollars a year. But it will be enormously the CSIRO scenario they increase by 75 percent rela- worthwhile on other margins to invest in education tive to 2000 levels. South Asia becomes a much larger for millions of young women who might otherwise importer of food under both scenarios, and East Asia be denied its many bene�ts. and the Paci�c becomes a net food exporter under the NCAR. In addition, the decline in calorie availability Putting the �ndings in context brought about by climate change raises the number of malnourished children. How does this study compare with earlier studies? The EACC estimates are near the upper end of Human health.22 The key human health impacts of estimates by the UNFCCC (2007), the study clos- climate change include increases in the incidence of est in approach to this study, though not as high vectorborne diseases (malaria), waterborne diseases as suggested by a recent critique of the UNFCCC (diarrhea), heat- and cold-related deaths, injuries study by Parry and others (2009). A comparison and deaths from flooding, and the prevalence of of the studies is limited by methodological differ- malnutrition. The EACC study, which focuses on ences—in particular, the use of a consistent set of malaria and diarrhea, �nds adaptation costs fall- climate models to link impacts to adaptation costs ing in absolute terms over time to less than half the and an explicit separation of costs of development 2010 estimates of adaptation costs by 2050. Why do from those of adaptation in the EACC study. 24 But costs decline in the face of higher risks? The answer the major difference between them is the nearly lies in the bene�ts expected from economic growth six-fold increase in the cost of coastal zone manage- and development. While the declines are consistent ment and defense under the EACC study. This dif- across regions, the rate of decline in South Asia and ference reflects several improvements to the earlier East Asia and Paci�c is faster than in Sub-Saharan UNFCCC estimates under the EACC study: better Africa. As a result, by 2050 more than 80 percent of unit cost estimates, including maintenance costs, health sector adaptation costs will be shouldered by and the inclusion of costs of port upgrading and Sub-Saharan Africa. risks from both sea level rise and storm surges. 24 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT TABLE 3 COMPARISON OF ADAPTATION COST ESTIMATES BY THE UNFCCC AND THE EACC, $ BILLIONS EACC Study Scenario Sector UNFCCC (2007) NCAR (wettest) CSIRO (driest) Infrastructure 2–41 27.5 13.0 Coastal zones 5 28.5 27.6 Water supply and flood protection 9 14.4 19.7 Agriculture, forestry, �sheries 7 2.5* 3.0* Human health 5 2.0 1.5 Extreme weather events — 6.7 6.4 Total 28–67 81.5 71.2 Note: *In the agriculture, forestry, and �sheries sector, the cost of adaptation has changed as compared to the estimates presented in the EACC global report (World Bank 2010a), in which these costs stood at $7.6 billion for the NCAR and $7.3 billion for the CSIRO scenarios. The current costs are estimated as the difference in public spending in the scenario with climate change and adaptation as compared to the no-climate-change scenario, and use the same methodology as has been applied to the other sectors. In World Bank (2010), the costs were incorrectly reported as reflecting the difference in public spending in the scenario with climate change and adaptation as compared to the scenario with climate change but no adaptation. The difference lowers the EACC lower bound estimate of the global cost of adaptation from $75 billion reported in WB (2010) to $71.2 billion per year, rounded to $70 billion per year. Source: UNFCCC (2007) and revised estimates World Bank 2010a. Another reason for the higher estimates is the higher of adaptation against a consistently projected devel- cost of adaptation for water supply and flood pro- opment baseline and use of a smaller multiplier on tection under the EACC study, particularly for the baseline investments than in the previous literature, CSIRO drier climate scenario. This difference is based on a detailed analysis of climate proo�ng, explained in part by the inclusion of riverine flood including adjustments to design standards and main- protection costs under the EACC study. Also push- tenance costs. ing up the EACC study estimate is the study’s com- prehensive sector coverage, especially inclusion of the The one sector where the EACC estimates are actu- cost of adaptation to extreme weather events. ally lower than the UNFCCC’s is human health. The reason for this divergence is in part because of the The infrastructure costs of adaptation in the EACC inclusion of the development baseline, which reduces study fall in the middle of the UNFCCC range the number of additional cases of malaria, and thereby because of two contrary forces. Pushing up the adaptation costs, by some 50 percent by 2030 in the EACC estimate is the more detailed coverage of EACC study. infrastructure. Previous studies estimated adaptation costs as the cost of climate-proo�ng new invest- The bottom line: calculating the global cost of adap- ment flows and did not differentiate risks or costs tation remains a complex problem, requiring projec- by type of infrastructure. The EACC study extended tions of economic growth, structural change, climate this work to estimate costs by types of infrastructure change, human behavior, and government invest- services—energy, transport, communications, water ments 40 years in the future. The EACC study tried to and sanitation, and urban and social infrastructure. establish a new benchmark for research of this nature, Pushing down the EACC estimate are measurements as it adopted a consistent approach across countries ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 25 and sectors and over time. But in the process, it had under the CSIRO (driest) scenario. Allowing for to make important assumptions and simpli�cations, to development between 2000 and 2050 greatly reduces some degree biasing the estimates. the numbers of people killed under both scenarios. The �ndings are similar for the number of people Lessons affected by floods and droughts. The sector estimates of adaptation costs presented In the health sector analysis, allowing for develop- in the global track report point to a few important ment reduces the number of additional cases of lessons. malaria, and thereby adaptation costs, by more than half by 2030 and more than three-quarters by 2050. A. Development is imperative ... Development dra- matically reduces the number of people killed by The greater the baseline level of development in each floods and affected by floods and droughts, quite period, the smaller is the impact of climate change apart from the impact of climate change (Figure 6). and the smaller are the costs of adaptation. Develop- If development is held constant at 2000 levels, the ment must be inclusive, however, to have these effects. number of people killed by floods increases over time And development can also increase vulnerabilities: under the NCAR (wettest) scenario and decreases the more developed the country, the greater the value FIGURE 6 DEVELOPMENT LOWERS THE NUMBER OF PEOPLE KILLED BY FLOODS AND AFFECTED BY FLOODS AND DROUGHTS, 2000–50 Killed by Floods (Per Million) Affected by Floods 0.7 0.0045 0.6 0.0035 0.5 0.4 0.0025 0.3 0.2 0.0015 2000 2010 2020 2030 2040 2050 2000 2010 2020 2030 2040 2050 Affected by Droughts 0.066 CSIRO Static Historical 0.045 NCAR Static CSIRO Development NCAR Development 0.036 2000 2010 2020 2030 2040 2050 Source: World Bank 2010a. 26 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT of infrastructure and personal property at risk from of raw water. This advice is almost invariably ignored climate change and therefore the greater the cost of due to deeply embedded political and social interests. climate-proo�ng such assets. However, these costs But the costs of misallocation of water resources will decrease with development as a percentage of GDP. escalate even without climate change and could be overwhelming under conditions of climate change. A B. …but not simply development as usual. Adaptation large share of the costs of adaptation in water supply will also require a different kind of development— and flood protection could be avoided by adopting breeding crops that are drought- and flood-tolerant, better management policies. climate-proo�ng infrastructure to make it resilient to climate risks, reducing overcapacity in the �sheries For good practical reasons, this study focuses on the industry, accounting for the inherent uncertainty in costs of adaptation that are likely to fall on the public future climate projections in development planning, sector, and it assumes limited or no change in tech- and others. nology, except in the agriculture sector analysis. But the boundary between public and private (autono- Consider water supply. Adapting to changing con- mous) adaptation is almost in�nitely flexible. So long ditions in water availability and demand has always as governments and the public sector ensure that been at the core of water management. Traditionally, incentives for innovation, investment, and private though, water managers and users have relied on his- decisions reflect the scarcity of resources once the torical experience when planning, such as consistency impact of climate change is taken into account, expe- in flood recurrence periods. These assumptions no rience demonstrates that the costs of adaptation may longer hold under climate change. Water manage- be dramatically reduced by a combination of techni- ment practices and procedures for designing water- cal change and private initiative. related infrastructure need to be revised to account for future climate conditions. Similarly, dikes and D. Uncertainty remains a challenge. The inherent other coastal protection measures will need to be uncertainty in climate projections—and the uncertain- built in anticipation of rising sea levels. ties about economic impacts and adaptation respons- es—makes climate-resilient development planning a C. Though adaptation is costly, costs can be reduced. The challenge. While the science is clear on general global clearest opportunities to reduce the costs of adaptation trends of climate change, current climate science can are in water supply and flood protection. Almost every provide only limited guidance to public investment in developed country has experienced what can happen speci�c countries or sectors, with the exception of sea when countries fail to shift patterns of development level rise. This study has estimated the cost of adapta- or to manage resources in ways that take account of tion under 2 (of 26) global climate models associated the potential impacts of climate change. Often, the with the A2 scenario of the IPCC Special Report reluctance to change reflects the political and eco- on Emissions Scenarios. The costs were estimated as nomic costs of changing policies and (quasi-) property though the countries knew with certainty what the rights that have underpinned decades or centuries of climate outcome would be. This is clearly not the case. development. Countries that are experiencing rapid Also, the study estimates the costs relative to a devel- economic growth have an opportunity to reduce the opment baseline, which in turn assumes a certain rate costs associated with the legacy of past development of growth of per capita GDP between 2010 and 2050. by ensuring that future development takes account of This is also not the case. changes in climate conditions. Economists and others regularly urge the adoption of mechanisms for man- This implies that climate adaptation must be limited aging water resources that recognize the scarcity value to robust measures such as education and climate- ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 27 related research. For durable climate-sensitive invest- for maintenance or earlier replacement of assets if cli- ments, a strategy is needed that maximizes the mate conditions are worse than anticipated. Preparing flexibility to incorporate new climate knowledge as for the worst might not be that expensive if the cost it becomes available. Hedging against varying climate of adjusting design standards to accommodate future outcomes, for example by preparing for both drier climate conditions is relatively small, as is the case for and wetter conditions for agriculture, would raise the many infrastructure assets. Insurance is more com- cost of adaptation well above the estimates here. plicated, because uncertainty about climate change also involves regional shifts in temperature and rain- There are three ways to deal with this uncertainty: fall. What might be large uncertainties for individual wait for better information, prepare for the worst, and countries might become much smaller when the costs insure. Countries will select among these options, of adaptation are pooled, particularly across regions. A depending on speci�c investment decisions and their funding mechanism that permits the reallocation of level of risk aversion. Since climate change is gradual, funds across regions as better information is collected designing for limited or no change in climate condi- about the actual outcome of climate change would tions while waiting for better information might save provide a basis for pooling risks across countries. money today but will likely result in high future costs 28 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT Four ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 29 Results From the Country Analyses The purpose of the country case studies is to help Mozambique is subject to flooding and extreme the governments understand the potential economic events, including tropical cyclones. Both Mozam- impacts of climate change and to support their efforts bique and Ghana are on the receiving end of water to develop sound policies and investments in response flowing out of major international river basins. With to these potential impacts. This chapter summarizes most of their economic activity and population con- the results of the country studies and their main les- centrated along the coast and in low-lying deltas, sons. For each country, the analyses consist of (a) a Vietnam and Bangladesh are Asian countries widely brief description of the nature and degree of vul- recognized as among the world’s most vulnerable to nerability of the country to climate change; (b) the climate change, particularly from extreme weather EACC study approach and main results from the events and flooding, with particular impacts on modeling exercises, as well as local-level perspectives; poorer populations. Bolivia is a poor Latin Ameri- and (c) a summary of country-speci�c lessons and can country traditionally dependent on the Andean recommendations. glaciers to supply good portions of water demand, and consisting of a wide range of agroecosystems— Choice of countries. The seven country case studies were from small-scale family agriculture on the Altiplano selected based on overall vulnerability to major climate (largely composed of native indigenous populations) change impacts; differing environmental, social, and to large-scale commercial agriculture in the lowlands economic conditions; and adequate data at the national of Santa Cruz. Finally, Samoa represents a low-lying level. Government interest was also fundamental to Paci�c island at increased risk of sea level rise and select the countries. Although it was difficult to identify storm surge. See Table 4 for information on which the best set of candidates in advance, as is always the sectors were covered by country. case in similar exercises, it was considered important to have representativeness in terms of continents, size, Methodology in African countries. The overall meth- population, and income level of the country, as well as odology adopted in the three African countries richness of data and local capacity to work with the closely follows the one used in the global track. Using EACC core team to apply the proposed methodology a time frame of 2050, development baselines are �rst in the country. developed for each sector. The baseline represents the growth path the economy would follow in the Mozambique, Ghana, and Ethiopia represent nearly absence of climate change. It is a reasonable trajec- the full range of agricultural systems in Africa. tory for growth and structural change of the economy 30 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT TABLE 4 SECTOR ANALYSES CARRIED OUT IN EACH COUNTRY CASE STUDY Mozambique Ethiopia Ghana Bangladesh Vietnam Bolivia Samoa Agriculture Yes Yes Yes Yes Yes Yes Water Yes Yes Yes Yes Yes Roads Yes Yes Yes Yes Hydropower Yes Yes Yes Coastal Yes Yes Yes Yes Yes Extreme events Yes Yes Yes Yes Yes Social Yes Yes Yes Yes Yes Yes CGE/MIP CGE CGE CGE CGE CGE MIP Macro Note: The forestry and �sheries sectors were only carried out for Vietnam. over a period of 40 years that can be used as a basis studies use a common core dynamic CGE model for comparison with the climate change scenario. The (Box 1), incorporating comparable approaches to cli- baselines for each sector utilize a common set of GDP mate change impacts and adaptation strategies. There and population forecasts for 2010–50 and a common are signi�cant differences across the three countries, set of climate scenarios to project temperature and given their very different economic structures—for precipitation changes to 2050. The changes in climate example, Ethiopia has no coastline, while Ghana and are provided by a few different climate models that Mozambique are subject to coastal impacts of climate attempt to represent the most extreme variations in change—but the common modeling framework sup- the main two variables—temperature and precipita- ports comparative analysis of sensitivity to shocks and tion. The different scenarios typically consist of the adaptation strategies. two considered in the global analyses (CSIRO and NCAR), plus two other country-speci�c extreme cli- Mozambique25 mate scenarios. They are used to predict impacts on economic sectors (agricultural output, consumption, Vulnerability to climate change water availability, and infrastructure). The �nal steps involve identifying and costing adaptation options for Mozambique is subject to frequent droughts, floods, the key sectors. The costs of adaptation comprise the and tropical cyclones. These events threaten the coun- costs of public policy adaptation measures and exclude try’s economic performance, which is already highly the costs of private (autonomous) adaptation. affected by high rainfall variability. Drought is the most frequent disaster, with an average incidence of every The modeling of the impacts of climate change in the 3–4 years. Floods in Mozambique are characterized by selected sectors is carried out using a suite of models. a number of geographical factors. More than 60 per- Output parameters from these models are then fed cent of Mozambique’s population lives in coastal areas, into a common dynamic computable general equilib- which are very susceptible to flooding because they are rium (CGE) model where the economic implications in low-lying regions of river basins, and in areas with of the modeled data are assessed. The African country poor drainage systems. In the period 1958–2008, 20 ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 31 BOX 1 CGE MODELS—PARTIAL AND GENERAL EQUILIBRIUM ANALYSIS The impact of climate change is simulated using a dynamic computable general equilibrium (CGE) model. These models have features making them suitable for such analyses. First, they simulate the functioning of a market economy—including markets for labor, capital and commodities—and provide a useful perspective on how changes in economic conditions are mediated through prices and markets. Secondly, the structural nature of these models permits consideration of new phenomena, such as climate change. Thirdly, these models assure that all economywide constraints are respected. This is critical discipline that should be imposed on long-run projections, such as those necessary for climate change. For instance, suppose climate change worsens growing conditions, forcing Ethiopia to import food. These imports require foreign exchange earnings. CGE models track the balance of payments and require that a suf�cient quantity of foreign exchange is available to �nance imports. Finally, CGE models contain detailed sector breakdowns and provide a “simulation laboratory� for quantitatively examining how various impact channels influence the performance and structure of the economy. In CGE models, economic decision making is the outcome of decentralized optimization by producers and consumers within a coherent economywide framework. A variety of substitution mechanisms occur in response to variations in relative prices, including substitution between labor types, capital and labor, imports and domestic goods, and between exports and domestic sales. The relatively long time frame considered (40 years into the future) means that dynamic processes are important and need to be captured in the dynamic CGE model. To the extent that climate change reduces agricultural or hydropower output in a given year, it also reduces income and hence savings. This reduction in savings translates into reduced investment, which translates into future reduced production potential. In the same vein, increased infrastructure maintenance costs imply less infrastructure investment, which further implies less infrastructure both now and in the future. Extreme events, such as flooding, can wipe out economic infrastructure; that infrastructure is gone, both in the period in which the event occurs and all future periods. Generally, even small differences in rates of accumulation can lead to large differences in economic outcomes over long time periods. The CGE model employed is well-positioned to capture these effects. The baseline development path adopted reflects development trends, policies, and priorities in the absence of climate change. It provides a reasonable trajectory for growth and structural change of the economy over about 50 years (the period 2003–50 is modeled) that can be used as a basis for comparison. We can, for example, run the CGE model forward imposing the implications of future climate on dry-land agricultural productivity. Within the model, the decisions of consumers, producers, and investors change in response to changes in economic conditions driven by a different set of climate outcomes. 32 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT major flood events were recorded, affecting more than surges directly and indirectly related to sea level rise 9 million people (RMSI 2009). These extreme events (Figure 7). These sectors are vulnerable to current cli- have been followed by outbreaks of disease, causing mate variability and are most likely to remain vulner- even more death and economic loss. Sea level rise able to future climate change. Other sectors or issues is predicted to increase the negative effects of storm of importance—such as health and urban infrastruc- surge and flood events along the coast. Over the next ture—were not included in the EACC analyses due 40 years, all such consequences of climate change are to lack of data at the required scale. likely to complicate the already considerable develop- ment challenge in Mozambique. Findings from the social component suggested that livelihood activities most sensitive to climate change The most vulnerable sectors to the impacts from impacts continued to take place in areas most exposed climate change in Mozambique are agriculture, to these impacts. For example, subsistence farmers which employs over 70 percent of the Mozambican continued to farm in areas prone to drought, render- population; energy, particularly hydropower genera- ing them even more vulnerable. In the case of �shing, tion which is dependent on water runoff; transport artisanal �shers reported venturing further out to sea infrastructure, notably roads; and coastal areas, which in search of better �sh stocks, even though this was do not conform to a “sector� but characterize speci�c increasingly dangerous due to the occurrence of more geographical areas vulnerable to floods and storm frequent and intense storms. FIGURE 7 POPULATION DENSITY AND COASTAL AREA (land less than 30 meters above mean sea level, in red) ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 33 EACC approach and results years, climate change would lead to a 2–4 percent decrease in yields of the major crops. This, combined Impacts. Changes in precipitation and temperature with the effects of more frequent flooding on rural from four GCMs (the two global scenarios plus two roads, would result in an agricultural GDP loss of 4.5 extreme scenarios for Mozambique, MOZDRY and percent (conservative) to 9.8 percent (most pessimis- MOZWET) were used to estimate (a) the changes tic), translating into a total GDP loss between 0.8 to in yield each year for both irrigated and rainfed 1.6%. In addition, the potential energy de�cit due to crops, as well as irrigation demand for six cash crops climate change relative to the base generation poten- and eight food crops; (b) flow into the hydropower tial (2005–50) is approximately -110,000 GWh. generation facilities and the consequent changes in generation capacity; and (c) the impact on transport The results from the integrated models of coastal sys- infrastructure and the increased demand and costs of tems (DIVA) show that in the 2040s, if there is no road maintenance. Simulations of sea level rise were adaptation, Mozambique could lose up to 4,850 km2 constructed independently of the climate scenarios.26 of land existing today (or up to 0.6 percent of national Two approaches were undertaken. First, an integrated land area) and a cumulative total of 916,000 people model of coastal systems was used to assess the risk could be forced to migrate away from the coast (or and costs of sea level rise in Mozambique. Second, 2.3 percent of the 2040s population). In the worst analyses of the interactions between cyclone risk and case, the total annual damage costs are estimated to sea level rise were undertaken for Beira and Maputo, reach $103 million per year in the 2040s, with the the two largest cities in Mozambique. forced migration being a large contributor to that cost. These damages and costs are mainly concen- Analysis at the sector level suggests, for example, trated in Zambezia, Nampula, Sofala, and Maputo net negative changes in crop productivity over all of provinces, reflecting their low-lying topography and Mozambique in all scenarios (Figure 8), with central relatively high population. Mozambique being hit hardest. Over the next 40 34 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT The analysis of the interactions between cyclone risk has potential implications for rates of economic and sea level rise performed for Beira and Maputo growth. These growth effects accumulate into sig- illustrate that relatively small levels of sea level rise can ni�cant declines in national welfare by 2050. In the dramatically increase the probability of severe storm worst case scenario, the net present value of damages surge events under the assumption of no change in (discounted at 5 percent) reaches about $7.6 billion, the intensity and frequency of cyclone events. Results equivalent to an annual payment of a bit more than are more dramatic for Beira than for Maputo City. $400 million. The probability of a cyclone strike in Maputo is lower due to greater latitude and the positioning of Maputo Figure 9 breaks down the climate change shocks into City relative to Madagascar. three groups: (1) crop yields and sea level rise (the latter is very small), (2) transportation system, and (3) The estimated impacts on agriculture, transport, hydropower. The graph illustrates the dominant role hydropower, and coastal infrastructure are fed into played by transport system disruption, principally as a macroeconomic CGE model that complements a result of flooding. The global dry scenario is in fact the sector models by providing a complete picture of a very wet scenario for the Zambezi water basin as a economic impacts across all sectors within a coherent whole, and thus causes signi�cant and enduring dam- analytical framework. The CGE model looks at the age to roads. By contrast, the local dry scenario is a impact of climate change on aggregate economic per- very dry scenario for Mozambique and causes greater formance. As indicated in Figure 8, climate change damage to agriculture. FIGURE 8 FIGURE 9 AGRICULTURE: EFFECTS ON YIELD IN 2050 DECOMPOSITION OF IMPACT CHANNELS COMPARED TO BASELINE FROM A MACROECONOMIC PERSPECTIVE Global Dry Global Wet Moz. Dry Moz. Wet (CSIRO) (NCAR) (UKMO) (IPSL) % Change in Per Capita Absorption Moz. -0.0 Growth Rate from Baseline -0.1 Moz. -0.2 Dry Global -0.3 Wet -0.4 Global Dry Falling crop yields and rising sea level Deteriorating transport system -6.0% -4.0% -4.0% 0.0% 2.0% 4.0% Declining hydropower generation Central South North Note: Bars represent the average change in crop productivity. Source: World Bank 2010g. Regional averages are weighted by historic crop yield rates per crop in the region. Source: World Bank 2010g. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 35 Adaptation. After calculating the impacts, the CGE FIGURE 10 then considers potential adaptation measures in three REDUCTION IN CC DAMAGES, 2003–50 sectors (hydropower, agriculture, and transportation).27 (5% discount rate, constant 2003 $) Four adaptation strategies are then introduced in the 8 Discounted US $ billion (const.2003) model to minimize the damages: (1) transport policy 7 change,28 and then the transport policy change plus (2) increased agricultural research and extension, or 6 transport plus (3) enhanced irrigation, or transport 5 6.1 6.1 plus (4) enhanced investment in human capital accu- 4 mulation (education). 3 2 0.6 The results are shown in Figure 10. Sealing unpaved 1 1.5 1.5 1.5 1.5 roads reduces the worst-case climate damages substan- 0 tially, restoring approximately one-�fth of lost welfare,29 Transport Expanding Agriculture Primary and with little additional cost; it is thus a no-regret Infrastructure (3) Irrigation (4) R&D (5) Education (6) action advisable even under the baseline. Remaining Agriculture R&D Irrigation Sealing Unpaved welfare losses could be regained with improved agri- or Education Roads cultural productivity or human capital accumulation Source: World Bank 2010g. (education). Irrigation investments appear to be a poor alternative: 1 million hectares of new irrigation land forced to migrate could be reduced by a factor of 140 would only slightly reduce CC damages. In terms of to 7,000 people. Hence, the total annual damage cost softer adaptation measures, investing in agricultural is reduced by a factor of 4 to $24 million per year. research in order to raise agricultural productivity However, the total investment required to achieve (through increases in crop yield and reductions in these adaptation options is estimated at $890 million post-harvest loses) by 1 percentage point each year over per year in the 2040s for the high sea level rise sce- baseline productivity trends offsets remaining damages nario, which appears much higher than the bene�ts to agriculture (for example, a further 50 percent maize of the adaptation in terms of damages avoided. At yield increase by 2050). Increasing rates of human the same time, more targeted investments in high capital accumulation by slightly more than 1 percent- value and more vulnerable locations can provide posi- age point also offsets damages. Increased investments tive returns. The range of costs of more economically in education helps increase aggregate labor force pro- viable adaptation options in the 2040s varies from ductivity and this translates into higher GDP. Lastly, $190 million to $470 million per year, depending on investment costs required to restore welfare losses are the sea level rise scenario. Note that these costs are subject to debate, but are reasonably less than $390 higher than the bene�ts from adaptation that accrue million per year over 40 years. through 2050. But these adaptation measures such as dikes are long-lived and would also yield long-term With respect to speci�c coastal adaptation measures, bene�ts well beyond 2050, the scope of this analysis, the integrated coastal system analysis examined two and in fact through 2100 as sea-level rise and storm protection measures: beach/shore nourishment and surge risks accelerate. sea and river dike building and upgrading. When these are applied, the physical and socioeconomic The analysis of the interactions of cyclone risk and sea impacts are signi�cantly reduced. For instance, the level rise for Beira and Maputo provides more impetus total land area lost could be reduced by a factor of for investment in the near term, particularly for Beira. more than 80 to 61km2, and the number of people While the full cost of the necessary infrastructure for 36 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT protecting Beira city and port has not been estimated year changes in consumption to which households to date, the dramatic fall in return periods for sea must adjust. A value of 0.56 in the baseline indicates inundation due to sea level rise strongly suggests that that poor households must manage annual swings protection schemes should be reassessed. in the change in consumption of 56 percent. These swings in income in the base model reflect current Equity issues. The incidence of impacts from climate climate variabilities—droughts, floods, and tropical change between households categorized as poor and cyclones. In all scenarios, the CVs for poor house- non-poor in the base year are approximately similar. holds are slightly higher than those for non-poor This result is somewhat surprising but follows from households—poor households must deal with more the fact that the biggest impacts of climate change are income variability than the non-poor. The impact of on infrastructure which permeate across the economy the climate change scenarios on the CVs is signi�- and impacts the poor and the non-poor. The same cant—rising to about 0.70 in the two global scenarios. holds true for adaptation measures; poor and non- However, it either remains constant or falls in the two poor households both bene�t from the adaptation Mozambique scenarios. measures, and the incidence of these bene�ts is not substantially different. Poor and non-poor do appear Local-level perspectives on adaptation. Results from to differ in terms of their vulnerability to shocks. the social component in Mozambique were remark- ably consistent with the economic analyses and with Figure 11 shows the impact of the extreme wet adaptation priorities previously identi�ed in the and dry scenarios, with and without road network Mozambique NAPA. The most common adaptation adaptation investments, on the coefficient of varia- preferences emerging from participatory scenario tion (CV)30 of the year-to-year growth rates of total development workshops and �eldwork results are household consumption. They represent the year-to- presented in Table 5. FIGURE 11 HOUSEHOLD CONSUMPTION: COEFFICIENT OF VARIATION OF YEAR-TO-YEAR GROWTH RATES 0.8 Co-Efficiant of Variation (SD/Mean) 0.7 0.70 0.71 0.6 0.63 0.63 0.5 0.56 0.56 0.51 0.51 0.51 0.49 0.4 0.3 0.2 0.1 0.0 Poor Poor Poor Poor Poor Non-Poor Non-Poor Non-Poor Non-Poor Non-Poor Baseline Extremely Wet Extremely Dry Extremely Wet Extremely Dry plus Adapation plus Adapation Source: World Bank 2010g. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 37 TABLE 5 KEY ADAPTATION OPTIONS IN MOZAMBIQUE Planned Adaptation Autonomous Adaptation Hard ■ Flood control dikes and levies ■ More robust buildings ■ Coastal flood control gates ■ Farm-scale water storage facilities ■ Dams and irrigation channels ■ Deep wells to provide drinking water for ■ Improved roadways people and animals ■ Improved communication infrastructure ■ Grain storage facilities ■ Improved hospitals and schools ■ Improved food processing equipment Soft ■ Improved early warning of climatic hazards, ■ Better utilization of short-season, and of dam releases drought-resistant crops to prepare for drought, ■ Better planning and management of forest, floods, and cyclones �sh, and other natural resources ■ Diversi�cation of flood and drought risk by ■ Resettlement of populations to lower risk maintaining �elds in both highland and lowland zones areas ■ More credit and �nancial services for small busi- ■ Better household and community management nesses and rural development and use of natural resources, including wild fruits ■ Better education and information for rural areas ■ Practice of soil conservation agriculture ■ Improved health care, social services, and social ■ Migration to lower risk areas support for all people ■ Diversi�cation of livelihoods away from agriculture ■ Better planning of how much grain to save for personal consumption, and how much to sell for income generation Note: The options in plain text respond directly to climate hazards, while those in italics represent measures to increase the adaptive capacity of the population, or to make them more resilient to shocks to their livelihoods in general. Source: World Bank 2010g. All of the planned options in the left column rep- Education and overall knowledge about climate resent potential government interventions. The events are also needed so that vulnerable groups can right column represents autonomous measures that expect disasters to be a constant feature in the future. people can undertake on their own. These results Speci�cally, more technical assistance for improving indicate that more vulnerable groups will not have land management practices and access to real-time the resources or skills to undertake all measures they weather forecasts—effective early warning— will be deem a priority. This is particularly true for the hard crucial to enhancing local adaptive capacity. Box 2 options that require resources. For example, during describes the social component in Mozambique. participatory scenario development workshops, the most frequently mentioned approach for reduc- Lessons and recommendations ing impacts was the construction of irrigation sys- tems, while the most frequently listed barrier to this Several important lessons emerge from the Mozam- was lack of �nance. However, in the absence of an bique work. enabling economic and political environment, many of the soft options are also challenging to undertake. Adaptation entails increasing the climate resil- For example, many participants noted the fact that ience of current development plans, with particu- people would like to diversify income, but there are lar attention to transport systems and agriculture few opportunities for diversi�cation. and coastal development. 38 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT BOX 2 Changes in design standards, such as seal- ing unpaved roads, can substantially reduce the METHODOLOGY AND FIELD WORK OF SOCIAL COMPONENT IN MOZAMBIQUE impacts of climate change even without addi- tional resources. In Mozambique, the vulnerability assessment included a literature review; the identi�cation The imperative of increasing agricultural produc- of six sociogeographic “hotspots;�and �eldwork tivity and the substantial uncertainties of climate in 17 districts across 8 provinces, including 45 focus group discussions, 18 institutional change argue strongly for enhanced investments stakeholder interviews, and a survey of 137 in agricultural research. households. Fieldwork included the use of participatory rural appraisal (PRA) exercises It is unlikely to be cost effective to protect the vast (village history; focus group discussions of majority of coastal regions of Mozambique from men, women and different age groups; wealth sea level rise; however, high value and vulnerable ranking; and community risk mapping), as well as key informant interviews with local locations, such as cities and ports, merit speci�c government of�cials, NGOs, and traditional consideration, especially those at risk for severe leaders. Household interviews also were storm surge events. undertaken, with 10 per site from different income tiers, with questionnaire modules “Soft� adaptation measures are potentially power- covering household composition, income ful. Because the majority of the capital stock in sources, agricultural practices, household shocks and coping strategies, past climate 2050 remains to be installed, land use planning adaptation practices, and perceptions about that channels investment into lower risk locations climate change. can substantially reduce risk at low cost. In parallel, three participatory scenario Viewed more broadly, flexible and more resil- development (PSD) workshops were held ient societies will be better prepared to confront in Mozambique: one in Xai-Xai, one in Beira, and one national-level workshop in the challenges posed by climate change. Hence, Maputo. PSD workshops began with technical investments in human capital contribute both to presentations to characterize current climate the adaptation agenda and to the development and socioeconomic projections for the agenda. coming decades. Thereafter, participants characterized visions of a “preferred future� Ethiopia31 for 2050. They considered how climate change could impact this future vision, and then identi�ed autonomous and planned Vulnerability to climate change adaptation options necessary to achieve the desired vision. Finally, participants identi�ed With around 75 million inhabitants, Ethiopia is prerequisites, synergies, and tradeoffs among the second most populated country in Sub-Saharan their adaptation and development visions and Africa (SSA). The country is divided into �ve agro- prioritized action for the short, medium, and long term. The PSD component of the social ecological zones (Figure 12). Around 45 percent of component had a capacity-building emphasis, the territory consists of a high plateau—comprising including participation of national teams in zones 2 to 4—with mountain ranges divided by the regional training on workshop design and East African Rift Valley. Almost 90 percent of the implementation. population resides in these highland regions (1,500 meters above sea level). Within the highlands, zones 2 and 3 generally have sufficient moisture for the ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 39 cultivation of key crops, whereas zone 4 is prone to FIGURE 12 droughts. The arid lowlands in the east of the coun- AGROECOLOGICAL ZONES IN ETHIOPIA try—zone 5—are mostly populated by pastoralists. Ethiopia is heavily dependent on rainfed agriculture. Its geographical location and topography, in combination with low adaptive capacity, entail a high vulnerability to the impacts of climate change. Historically, the country has been prone to extreme weather variability—rainfall is highly erratic, most rain falls with high intensity, and there is a high degree of variability in both time and space. Since the early 1980s, the country has suffered seven major droughts, �ve of which have led to fam- ines, in addition to dozens of local droughts. Major floods also occurred in different parts of the country in 1988, 1993, 1994, 1995, 1996, and 2006. EACC approach and results To capture uncertainties on future climate, this study utilizes the two “extreme� GCMs used in the global track of the EACC (labeled Wet1 and Dry1), as well central highlands, and (b) more than 25 percent in as two additional models that are better suited to rep- most of the rest of the country. If the Wet2 scenario resent climate model uncertainty in the speci�c case is accompanied by an increase in the variability of of Ethiopia (labeled here Wet2 and Dry2). For the short-duration rainfall intensity, there would be an baseline (no-climate change scenarios), the analysis increased chance of severe episodic flooding caused uses historical monthly climate data and projects by storm runoff in highland areas. the historical pattern into the future. For the cli- mate change scenarios, stochastic representations of The economic analyses focus on three main channels weather variability in each global circulation model of climatic vulnerability that already affect the Ethio- (captured through a single draw) are superimposed pian economy, and are likely to be of major signi�- on the baseline to capture the variability of the future. cance under the climate of the future as well. These The scenarios include projections of extreme weather channels include (1) agriculture, which accounted for events such as droughts and floods. 47 percent of Ethiopian GDP in 2006 and is highly sensitive to seasonal variations in temperature and Climate projections obtained from these four GCMs moisture; (2) roads, the backbone of the country’s suggest an increase in rainfall variability with a rising transport system, which are often hit by large floods, frequency of both severe flooding and droughts due causing serious infrastructure damage and disrup- to global warming. The Dry2 scenario shows reduc- tions to supply chains; and (3) dams, which provide tions in average annual rainfall over 2045–55 of (a) hydropower and irrigation and are affected by large 10–25 percent in the central highlands, (b) 0–10 per- precipitation swings. cent in the south, and (c) more than 25 percent in the north. The Wet2 scenario shows increases in average Speci�cally, changes in precipitation and tempera- annual rainfall of (a) 10–25 percent in the south and ture from the four GCMs were used to estimate 40 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT (a) changes in yields for major crops and impacts on FIGURE 13 livestock, (b) flow into hydropower generation facili- ties and the consequent changes in power generation, DEVIATIONS OF GDP FROM BASE SCENARIO (c) the impact of flooding on roads; (d) the effects 2015 2025 2035 2045 of more frequent droughts on government expendi- 0.00 Percent Deviation from Base ture on vulnerability and food security (VFS); and -2.00 (e) the loss of irrigation and hydropower due to con- -4.00 flicts among competing demands. The analyses assess -6.00 deviations in GDP and other variables from the no- -8.00 climate-change baseline growth path for the four cli- -10.00 mate change scenarios mentioned above. -12.00 Wet 2 Wet 1 Dry 1 Dry 2 Impacts. Economy wide impacts of climate change were assessed using a CGE model. The results of the model- Source: World Bank 2010e. ing (Figure 13) suggest a large loss of GDP under the FIGURE 14 Dry2 scenario (6 percent to 10 percent) that is fairly evenly distributed across the entire time horizon. In con- AGRICULTURAL YEAR-TO-YEAR GROWTH RATES: STANDARD DEVIATIONS trast, in the case of the Wet2 scenario, the loss of GDP is 7.00 quite substantial in the 2040–49 decade, because of the costs of coping with damage caused by extreme weather 6.00 events, especially floods, from the 2030 decade onward. Percentage Points 5.00 The 10-year average GDP for the �nal decade is nearly 8 percent lower than the baseline. While these are not 4.00 forecasts of future climate impacts, they highlight the 3.00 high degree of vulnerability of Ethiopian agriculture and infrastructure to future climate shocks. 2.00 Climate change brings about increased weather vari- 1.00 ability, which translates into large swings in the growth 0.00 rate of agriculture GDP, illustrated by the increase in Base Dry 2 Wet 2 standard deviation compared to the baseline in Figure 14. While the simple means of annual growth rates Source: World Bank 2010e. are similar across the scenarios, high variability leads based highlands zone 2 (R2) and in urban areas. The to signi�cant welfare losses. latter reflects the downstream consequences of flood- ing and weather variability. The dry scenarios have Variability in agricultural income tends to affect the reverse impacts, with the arid lowlands and livestock poor more, with standard deviation values on average suffering greatly. some 10 percent higher than for the non-poor under both wet and dry scenarios. As shown by Figure 154 In addition to analysis of the three priority sectors for the Wet2 scenario, climate change impacts also (agriculture, roads, and hydropower), the study also are likely to vary signi�cantly across regions. The arid analyzed potential conflicts under climate change in lowland zone 5 (R5) derives substantial bene�ts from the use of water across sectors. A water planning model the increase in total rainfall, which supports livestock, was used to evaluate the potential interactions among while relative losses are concentrated in the cereals- growing municipal and industrial (M&I), irrigation, ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 41 FIGURE 15 REGIONAL GDP, DEVIATION FROM BASE, WET2 STANDARD DEVIATIONS 40.00 30.00 Percent Deviation from Base 20.00 10.00 0.00 2015 2025 2035 2045 -10.00 -20.00 rrR1 rrR2 rrR3 rrR4 rrR5 rrURB Source: World Bank 2010e. and hydropower demands under climate change. The efforts are required to attenuate climate change model evaluates these inter-sectoral effects between impacts. Adaptation strategies were therefore iden- 2001 and 2050, and generates time series of impacts ti�ed as additions to—or modi�cations of—current on irrigated agricultural yields and hydropower gen- government programs. eration under each of the climate scenarios. More specifically, adaptation in agriculture The results suggest that hydropower production included increasing the cropland to be equipped is impacted by irrigation and M&I withdrawals. for irrigation and investing in agricultural research Under the Dry2 scenario, giving priority to agricul- and development. In the transport sector, adapta- tural demands results in a loss of hydropower capac- tion options included increasing the share of paved ity equivalent to 100 percent of the 2000 installed and hardened roads, as well as “soft� measures such capacity and 10 percent of the hydropower capacity as changes in transportation operation and main- planned by the government for the period 2011–15. tenance, development of new design standards that If, on the other hand, priority is given to hydropower, consider projected climate changes, transfer of up to a billion m3 of water might be taken away from relevant transportation technology to stakehold- irrigated agriculture. That would cause a 30–40 per- ers, and the enhancement of transportation safety cent yield drop in an area of some 250,000 hectares measures. Adaptation policy considered in the that would be forced to revert to rainfed conditions. hydropower sector included altering the scale and timing of planned projects, as well as constraining Adaptation. The investment program included in the total downstream flow and irrigation flow. no-climate-change baseline established in consulta- tion with the government is likely to enhance Ethio- These strategies were �rst assessed on a sector-by sector pia’s resilience to climate change. However, additional basis. When the full set of economy wide linkages 42 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT is taken into account, direct plus indirect adaptation TABLE 6 costs increase signi�cantly, as indicated in Table 6. ADAPTATION COSTS (annual average, 2010-50, $ billions) To evaluate its welfare implications, the adapta- Direct, sector Indirect Total level costs costs costs tion strategy was analyzed in a CGE framework by comparing a no-climate-change baseline—reflecting Wet2 0.19 0.60 0.79 existing development plans—with climate change Wet1 0.17 0.77 0.94 scenarios reflecting adaptation investments. The main �ndings are that adaptation (a) reduces, but does not Dry1 0.16 2.30 2.46 eliminate, welfare losses; (b) that such welfare gains can be achieved at relatively low cost; and (c) that Dry2 0.26 2.55 2.81 adaptation lowers income variability. Source: World Bank 2010e. As shown in Figure 16, adaptation greatly reduces the FIGURE 16 welfare loss due to climate change (measured here by the difference from the baseline of total absorption— NET PRESENT VALUE (NPV) OF WELFARE GDP plus imports minus exports, discounted over DIFFERENCES the 40-year time horizon). Wet 2 Dry 2 Wet 1 Dry 1 0.0 Ratio (%) to NPV of Base GDP Finally, adaptation restores the variability of agri- -2.0 culture GDP growth close to the baseline scenario (Figure 17). -4.0 -6.0 While the bene�ts of adaptation investments are sig- ni�cant, they do not fully offset the negative impact -8.0 of the climate change scenarios. Two options were explored to close the “welfare gap� caused by climate -10.0 change. The �rst is to estimate the “residual damage No Adaptation Adaptation costs� as the transfers (in $) that would be required to completely offset the loss of absorption from CC Source: World Bank 2010e. shock, after implementing adaptation investments. Closing the “welfare gap� through residual compen- labor-upgrading program appears to be able to more sation would entail mobilizing signi�cant resources than offset the negative impacts of climate change. compared to direct project-level adaptation costs (Table 7). While no information was available within the time frame of the analysis to properly cost the skill upgrad- The second approach is to include an additional ing program, this �nding points to the signi�cant labor-upgrading program in the adaptation strategy. potential bene�ts of accelerating the diversi�cation In this scenario, 0.1 percent of rural unskilled labor of the economy away from highly climate sensitive is assumed to be transferred to the urban region, with sectors such as agriculture. In the Wet2 scenario, for additional upgrading so that all the urban labor cat- any value of the program cost below $0.4 billion/ year, egories, skilled and unskilled, grow uniformly faster a development strategy including a skill upgrading than in the base run. When tested under the Wet2 program like the one considered here would appear to scenario, an adaptation strategy including such a be preferable to the residual compensation approach. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 43 FIGURE 17 communities living on already degraded lands, and pastoral communities who face severe conflicts over STANDARD DEVIATION OF YEAR-TO-YEAR AGRICULTURE GDP GROWTH RATES, WITH natural resources (especially access to land for herd AND WITHOUT ADAPTATION mobility) with agriculturalists and the state. 6.00 Local participatory scenario development (PSD) 5.00 workshops identi�ed soil and forest rehabilitation, 4.00 irrigation and water harvesting, improved agricultural 3.00 techniques and drought-resistant varieties, education, 2.00 and land use rights for pastoralists as adaptation 1.00 preferences. Regional development and the need for 0.00 structural shifts toward service and industry sectors Baseline No With No With Adaptation Adaptation Adaptation Adaptation to improve employment outcomes were also raised Wet 2 Dry 2 as issues. At the national level, similar options were identi�ed, along with a focus on early warning sys- Source: World Bank 2010e. tems and flood control measures, agricultural tech- nology, �nance and market development, renewable TABLE 7 energy, and urban planning. The adaptation options identi�ed at local and national levels generally aligned ADAPTATION COSTS AND RESIDUAL with the natural resource and agriculture focus in the DAMAGE (annual average, 2010–50, $ billions) NAPA, which also identi�es needed investments in Adaptation Residual crop insurance, wetlands protection, carbon liveli- Scenario costs damage Total hoods, agroforestry, and anti-malaria initiatives. Wet 2 0.79 0.43 1.22 Wet1 0.94 0.81 1.75 Lessons and recommendations Dry1 2.46 1.52 3.97 The �ndings of this analysis suggest that impacts of Dry2 2.81 3.03 5.84 climate change will be quite signi�cant, particularly Source: World Bank 2010e. as Ethiopia approaches the middle of the century. The magnitude of the impacts remains considerable, irre- Local-level perspectives on adaptation. Land and spective of whether the climate of the future will be water management are central concerns in Ethiopia, wetter or drier. Given the large uncertainty on future which is subject to extremes of drought and floods. climate outcomes, the approach to enhance Ethio- Vulnerable groups identi�ed through community pia’s climate resilience should be couched in terms discussions included asset-poor households with very of a gradual, adaptive, and learning paradigm. Such limited means of coping with climate hazards, the an approach could be articulated for both the shorter expanding group of rural landless who lack income term—including the implementation of the Growth opportunities, the urban poor living in flood-prone and Transformation Plan (GTP) recently issued by areas of cities, and the elderly and the sick due to the government—and for the long term. their limited adaptive capacity. Women and children left behind as male adults migrate for employment Shorter term (up to 2015). By and large, the Growth during drought-related production failures were also and Transformation Plan supports a number of identi�ed as vulnerable during and after extreme actions that—by boosting growth—will contribute to events. Other vulnerable groups identi�ed included the enhancement of Ethiopia’s resilience to climatic 44 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT shocks. Robust growth based on infrastructure invest- railways, energy, telecommunication, irrigation, ment is likely to be the �rst line of defense against drinking water and sanitation and basic infrastruc- climate change impacts. Relatively small deviations ture developments; (…) With regard to roads, rural from the ambitious investment targets set forth by roads will be constructed on all regions and all rural the government for roads, dams, hydropower, water kebeles will be connected (through) standardized all management, and irrigation would signi�cantly weather roads with main highways.� increase long-term vulnerability to climate change and thus make adaptation costlier. Modeling results show that existing infrastructure design standards—that is, the level of prevention However, there are a number of additional issues that against extreme events such as local and regional the government could consider to further enhance flooding—are inadequate to address current climate the contribution of GTP to Ethiopia’s climate variability and will impact economic growth rates resilience—and thus, ultimately, to the ability of the in the near to medium term. Results from climate country to support sustained, long-term growth. change analyses show this issue is likely to become worse in the medium to long term. The government Agriculture. The GTP purports to “continue the should consider enhancing infrastructure design ongoing effort of improving agriculture productivity standards as soon as possible. in a sustainable manner so as to ensure its place as the engine of growth.� The analysis of this report indicates Even under current climate, the direct bene�ts—in that agricultural production as an engine of growth is terms of increased lifetime—of roads designed fol- vulnerable to climate change and climate variability. lowing higher standards outweigh the corresponding While the more pronounced effects on crops and live- costs in a discounted bene�t/cost analysis. The case stock are likely to materialize in later decades, efforts for improved design standards is even stronger under to enhance the resilience to climate shocks of crop climate change, irrespective of climate outcomes: the yields and livestock production should be stepped up bene�t/cost ratio of adopting higher design standards as soon as possible, particularly on account of the lead is 17 to 75 percent higher than in the baseline under time needed to strengthen research systems, and to the Wet2 scenario, and 16 to 55 percent in the Dry2 transfer and adapt �ndings from the lab to the �eld. scenario (Figure 18). In addition, there are important indirect economy wide bene�ts: a more climate- Investments in improved agricultural productivity— resilient road network can avoid costly disruptions such as watershed management, on-farm technology, of communications links and supply chains that access to extension services, transport, fertilizers and increased flood frequency might bring about. improved seed varieties, and climate and weather forecasting—will enhance the resilience of agricul- Energy. Current water resources and Ethiopian topog- ture, both to droughts and to waterlogging caused raphy indicate an overall potential of more than 30,000 by floods. National and local actions will need to megawatts in economically viable hydropower genera- be supported by international efforts—for example, tion capacity. The GTP approach is to focus on “the through the CGIAR system—to develop climate development of water and wind energy options to ful- resilient agricultural technologies, given the global �ll the energy demand of the country,� with targets for public good nature of these innovations. hydropower of 6,000 to 8,000 MW in additional gen- eration capacity. The hydropower analysis of this report Road infrastructure. The GTP aims to expand the (conducted at the monthly scale, which is adequate for coverage and enhance the quality of infrastructure: sector-wide planning purposes, although not for plant- “Focus will be given to the development of roads, level design and operation) provides support, from a ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 45 FIGURE 18 BENEFIT/ COST RATIO OF UPGRADING ROAD STANDARDS Wet 2 Scenario Dry 2 Scenario 5.00 4.00 3.50 4.00 3.00 3.00 2.50 2.00 2.00 1.50 1.00 1.00 0.50 0.00 0.00 2020 2030 2040 2050 2020 2030 2040 2050 No Climate Change With Climate Change Source: World Bank 2010e. climate change perspective, to the GTP targets. The and to facilitate the economy’s transition from being projects likely to go online in the next 5 years have very highly agriculture-dependent to having a broader low risk of being impacted by climate change. productive base in industry and services. Given the vulnerability of the agricultural sector to current cli- While in the longer term (see below) hydropower mate shocks (let alone those to be expected in the development will become increasingly more climate future), strengthening of the electricity sector, and sensitive, projects in the current pipeline are likely to in particular the promotion of regional and Africa- be less vulnerable to shocks as the overlap between wide power grids to receive Ethiopia’s excess power, their life span, and the time when stronger climate should be a priority in the investment strategy. change effects will materialize, is relatively limited. Strengthened hydropower development can both Some climate change scenarios actually project an increase near-term economic growth and make the increase in Ethiopian runoff, resulting in larger vol- energy system more climate-resilient. Additional umes of hydropower generation, and thus making the reservoir storage distributed over the country also case for investment in hydropower stronger. will provide more reliability and protection from regional droughts. In the nearer term, the economics of hydropower investments will be influenced less by climate, and Medium to long term. As Ethiopia looks into the next more, on the demand side, by the evolution of domes- stages of development—starting with preparation of tic and external markets. A sustained expansion of the next growth plan following the GTP 2011–15—it national and foreign demand for power will be key to might want to evaluate more closely the implications support the expansion of Ethiopia’s hydropower sec- of climate change for its overall policies and infra- tor, which in turn will be vital to support the country’s structure development programs. Early planning for accelerated economic growth. the more severe climate impacts of mid-century is desirable, so as to avoid locking the country into a cli- In the short run, expansion of hydropower genera- mate-vulnerable development trajectory, particularly tion should be accelerated as a way to support growth when it comes to economic processes with a high 46 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT degree of inertia, or investment decisions concerning infrastructures with a long life span. Due to the uncertainty of future climate, a risk-based investment planning approach should be adopted. Robust decision-making principles are needed to minimize the “regrets� of climate-sensitive decisions. As climate shocks become more frequent and severe, the “opportunity costs of capital� invested in projects and programs that are viable only under a limited set of climate outcomes becomes too large. Some key areas for consideration to develop a climate risk management approach to support long-term devel- opment include the following. Macroeconomic management. Historically, the Ethio- pian economy has been vulnerable to climate fluc- tuations. The analysis of this report shows that climate variability will increase under all scenarios. Since agriculture (the economy’s most climate sen- sitive sector) is likely to remain for some time one of Ethiopia’s main engines of growth, climate-induced shocks will continue to be a threat to macroeco- nomic stability because of the impacts on income, employment, �scal revenues, capital formation, the drain on government expenditures, and aid flows to support disaster relief. Under climate change, renewed efforts will be neces- sary to buffer the economy from more frequent and/ or severe climate shocks. These include strengthen- ing social safety nets, access to relief funds, drought early warning systems, crop insurance programs, grain banks, and strengthening infrastructure design. Promote diversi�cation. In the longer term, however, accelerated diversi�cation of income and employ- ment sources away from climate-sensitive sectors such as agriculture is likely to become increasingly important under a more erratic climate. It be should explored in closer detail, particularly because it holds promise to be a cost-effective way to eliminate residual welfare damage caused by climate change. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 47 The government may want to look into ways to Ghana32 accelerate the absorption of the rural labor force into non-agriculture activities, including through skills- Vulnerability to climate change upgrading programs and encouragement of growth poles around medium-size municipalities. Ghana’s economy is particularly vulnerable to climate change and variability because it is heavily depen- Evaluate the climate resilience of large infrastructure dent on climate-sensitive sectors such as agriculture, projects. As we move toward mid-century, the range forestry, and hydropower. The agricultural sector, in of possible climate futures broadens to encompass particular, is highly vulnerable because it is largely markedly different “wet� and “dry� scenarios. This rainfed with a low-level of irrigation development. has implications for the optimal timing of dams and The country’s 565km coastline is inhabited by about other investments in water infrastructure, which is a quarter of the population and is the location of sig- likely to be quite sensitive to climate outcomes. Large ni�cant physical infrastructure. projects of this type should be subject—on account of the large capital outlays involved—to careful climate- The inheritance system, local governance and custom- robustness tests. ary law, and multiple forms of land tenure systems disproportionately harm both women and migrants’ To adequately inform the design of subsequent adaptive capacity. Rural-rural migrants, for example, generations of water infrastructure projects, invest- forgo income by not planting long-gestation cash ments in enhancing national hydro-meteorological crops for lack of a secure title in receiving areas. Sea- services, data collection, and analysis are crucial to sonal floods, which are indicated by projected climate assist identifying which climate change path Ethio- change scenarios, could cause signi�cant impacts pia is actually on, and to provide inputs to the adap- in highly populated urban and peri-urban areas in tive management process for resource management. Greater Accra, particularly given poor housing and Better data on hydro-meteorological processes, and the possibility of disease outbreaks in the “zongo� stronger capacity to analyze and model the data, is slums dominated by in-migrants. key to making more informed decisions on issues such as the number of hydropower plants, the design In line with the approach taken in the global track of individual plants, and the operation of the grid. study, climate projections from the NCAR and CSIRO models were used to generate the “Global Proactively address conflicts in water uses. Under “dry� Wet� and “Global Dry� scenarios for the Ghana future climate scenarios, competition among water case study. In addition, the climate projections from users —municipal and industrial consumption, hydro- the two GCM/SRES combinations with the lowest power generation, and irrigation—might become and highest CMI for Ghana were used to generate more acute, particularly in certain river basins. The a “Ghana Dry� and a “Ghana Wet� scenario (Table availability of water to downstream riparian countries 8). Note that in the case of Ghana, the globally “wet- might also be affected. test� GCM actually projects a drier future climate for Ghana than the globally “driest� GCM under emis- Given the signi�cant pay-off of addressing internal sion scenario A2. and transboundary conflicts on water use before they arise, the government might want to consider invest- The projections indicate fairly wide fluctuations in ments in river basin planning systems and institu- annual temperatures in all four Ghana agroecological tional arrangements that can facilitate information regions (northern savannah, southern savannah, for- sharing, dialogue, and dispute resolution. est, and coastal) for all the four scenarios. However, 48 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT the trend over the period 2010–50 indicates warming TABLE 8 in all regions, with temperatures increasing the most GCM SCENARIOS FOR THE GHANA CASE in the northern savannah region—with increases of STUDY up to 2.2–2.4°C, leading to average temperatures as CMI high as 41°C—while also presenting the widest range Scenario GCM SRES Deviation of temperature variability (5.7°C range). All agroeco- Global Wet ncar_ccsm3_0 A2 -17% logical regions show signi�cant precipitation variabil- Global Dry csiro_mk3_0 A2 9% ity compared to the baseline scenario. The coefficient Ghana Wet ncar_pcm1 A1b 49% of variation of annual precipitation in Ghana varies between -9 percent (global wet scenario) to -14 per- Ghana Dry ipsl_cm4 B1 -66% cent (Ghana dry scenario). Source: World Bank 2010f. As shown in Figure 19, there may be wide variations EACC approach and results in stream flows and runoff changes. The southwestern part of Ghana is expected to experience increases in Impacts. All four GCM scenarios suggest signi�cant runoff under both Ghana speci�c scenarios, with the adverse economywide effects, which become stronger opposite occurring with the Black Volta basin. The toward 2050. Although there is considerable variation fluctuations in stream flows and runoffs, particularly in real GDP growth over the simulation period, the in the Volta River, increase the risk of floods and/or overall trend relative to the baseline is clearly down- droughts in urban and rural areas. Given that Ghana ward. Toward 2050, annual real GDP is projected has very little control over the upper streams of rivers to be 1.9 to 7.2 percent lower than in the baseline across its borders in Burkina Faso and Togo, there is scenario without anthropogenic climate change. Real a need for regional cooperation in the management household consumption also declines relative to the of water resources among the neighboring countries base scenario in all the four climate change scenarios sharing the Volta basin. analyzed in this study. FIGURE 19 SURFACE FLOW AVERAGE DIFFERENCE FROM THE NO-CLIMATE-CHANGE SCENARIO, 2010–50 (million m3/month) Ghana Dry Scenario (ipsl_cm4-B1) Ghana Wet Scenario (ncar_pcm1 A1b) Source: World Bank 2010f. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 49 In comparison to the baseline growth path without FIGURE 20 climate change, the output of the agricultural sector is DEVIATION OF REAL TOTAL AND AGRICUL- estimated to decline by between 10.1 percent (Global TURAL GDP FROM BASE, TERMINAL YEAR (%) Wet) and 3.0 percent (Ghana Wet) by 2050 (Figure 20). The projections for cocoa show considerable variation across the climate scenarios, regions, and Ghana Wet decades. Under the Ghana Wet climate, cocoa pro- duction is projected to experience signi�cant adverse effects, while under the Global Dry and Global Wet Ghana Dry climates the impacts turn out to be predominantly positive from a nationwide perspective. The variance of annual cocoa yields rises across all climate change Global Wet scenarios. Ghana’s coastal zone is of immense signi�cance to Global Dry the economy. There are �ve large cities located in the coastal zone and about a quarter of the population lives in this area. It is estimated that over 240,000 -12.00 -10.00 -8.00 -6.00 -4.00 -2.00 0.00 people living in the coastal zone are at risk of seal Agriculture level rise (Ghana Statistical Services, 2008). Like most coastal cities around the world, Ghana’s coastal Total areas are vulnerable to extreme events above the cur- rent defense standards of structural protection, and Source: World Bank 2010f. are especially vulnerable to coastal flooding. Addi- tional threats include coastal erosion and reduction in freshwater resources in deltas and estuaries. billion from a 2050 baseline output of $2.33 billion. This represents a decline of between 3 and 6%. The total cost of damage from flooding, land loss, and forced migration is estimated to reach $3.7 million/ Adaptation. Adaptation actions were considered in yr by the 2020s, rising to $6.5 million/yr by the 2040s four key sectors in Ghana—roads, agriculture, hydro- using the high sea level rise scenarios. Damage costs power, and coastal. In each case it is important to look are estimated at $4.7 million/yr for the 2040s using at the resource envelope available to fund adaptation, the low sea level rise scenario. which can begin with the existing government budget (or projected budget), and increase to a higher level if Ghana’s water and energy sectors have already shown one assumes that funding will be available from dif- signs of vulnerability to climate change, particularly ferent sources. the effect of highly variable precipitation patterns on hydropower production. The 1980–83 drought not Road Transport. Adaptation of road infrastructure is only affected export earnings through crop losses, but considered in order to make the road network more also caused large-scale human suffering and called climate-resilient at no additional cost compared to into question the nation’s continued dependence on the baseline. That is, the baseline road infrastruc- hydroelectric power. By 2050, the annual average ture budget is just reallocated through changes in output of the water and energy sector is expected to road design standards. This is more costly initially decline to within a range of $2.19 billion to $2.26 and reduces the amount available for the expansion 50 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT of the road network, but at the same time there is less climate change damage to the road network later on. The economywide simulation analysis assumes that the same road infrastructure adaptation strat- egy is adopted in all adaptation scenarios under consideration. Agriculture. The whole adaptation resource envelope considered is spent on gradual expansion of irrigated land area from 2012 onwards. The assumed upfront investment cost of irrigation is $18,000 per ha, taking account of Ghana-speci�c cost estimates for recent and planned irrigation projects, plus the need for complementary investment in water harvesting etc, as this strategy requires the irrigable land area to expand. For example, under the Ghana Dry scenario, the share of irrigated land rises gradually from less than 0.4 to 23 percent of the current total cultivated area. The resulting average annual factor productivity increase for crop agriculture as a whole is an additional 0.54 percent above baseline productivity growth. This scenario can also be interpreted as representing other productivity-rising agricultural adaptation measures with a comparable yield impact per dollar spent. Energy. In this sector, part of the available resource envelope is spent on additional investments in hydro- power relative to the baseline, minimizing negative climate change impacts on power generation. The remaining part of the resource envelope is spent on agricultural productivity improvements. The pres- ent value of the additional power investment up to 2050 is estimated to be $859 million, which reduces the amount available for agricultural investment. For example under the Ghana Dry scenario, 32 percent of the resource envelope goes to power and the rest to agriculture. Coastal Zone. Coastal adaptation analysis and options have been used as an example for now, while a more detailed analysis is being completed using the DIVA Model and SRTM 90-m resolution data. The larg- est cost component is the construction of sea dikes, estimated to be about $87 million per year under a ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 51 high sea level rise scenario, and $30–34 million a year available for adaptation measures is spent on addi- under a low sea level rise scenario. Annual mainte- tional broad-based education and training that raises nance costs of the sea dikes under the high sea level labor productivity across all skill groups. rise scenario will be about $9 million in the 2010s, rising to $35 million by the 2040s. Total annual adap- Results from simulations. Table 9 shows that in the tation costs for the coastal zone are estimated to be absence of adaptation, aggregate real welfare losses up $19–$156 million. These results are based on a partial to 2050 will range (in present value terms) from $2.7 equilibrium model (assuming no interaction between billion (Ghana Dry) to $13.1 billion (Global Dry). In the coastal sector and the rest of the economy) and annualized values, these estimates range from $157 are not part of the CGE adaptation analysis. They million (Ghana Dry) to $765 million. On a per capita also do not include other adaptations measures, such basis, they amount to $6.50 and $31.46 for Ghana Dry as protection for the �shery industry. Compared to and Global Dry, respectively. The equivalent annual 1990, sea levels were assumed to gradually increase value of the real welfare impacts with no adaptation from 4 cm in 2010 to 15.6 cm by 2050 using the low range from $157.9 million (Ghana Dry) to $764.5 sea level rise scenario (and from 7.1 cm to 37.8 cm million (Global Dry). These results translate to an using the high sea level rise scenario). annual equivalent per capita impact that range from $6.5 (Ghana Dray) to $31.46 (Global Dry). Macroeconomic/integrated analysis. In the dynamic CGE analysis, it is assumed that the maximum Table 10 reports deviations of the present value of resource envelope available for adaptation measures welfare from the baseline for the three alternative over the simulation period (2010–50) is equal to the adaptation strategies (these strategies are built in present value of the aggregate welfare loss due to cli- combination with the road adaptation strategy). In mate change in the absence of adaptation measures order to generate a meaningful comparison across (Table 9). From an economywide perspective, these alternative adaptation investment paths, the total �gures represent the lump-sum income transfers resource envelope for adaptation investments is the Ghana would have to mobilize from external sources same across the different strategies (but different in order to be fully compensated for the economic across the four climate scenarios). impacts of climate change.33 Similar to the other case studies for Africa, the CGE analysis also includes an Changes in road design standards alone provide sig- adaptation scenario in which the resource envelope ni�cant reductions in welfare losses with the notable TABLE 9 REAL WELFARE IMPACTS WITH NO ADAPTATION INVESTMENTS Present value of lost Equivalent annual value Annual equivalent Climate Scenario welfare ($ Billion) ($ Million) per capita Global Dry 13.118 764.5 31.46 Global Wet 10.095 588.3 24.21 Ghana Dry 2.709 157.9 6.50 Ghana Wet 4.050 236.0 9.71 Notes: (i) Discount rate = 5%. Welfare is measured by real absorption, the constant-price value of domestic and imported �nal goods and services available for household consumption, government consumption, and capital stock investment. Second column: constant annual flow with same present value. Third column: second column / 2010 population (UN medium projection: 23.4 Million). Source: World Bank 2010f. 52 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT TABLE 10 DEVIATIONS OF REAL WELFARE FROM BASELINE UNDER ALTERNATIVE ADAPTATION STRATEGIES (present value, in $ billions) Adaptation Investment in No Adaptation scenario Road Design Agriculture Hydro /Agric. Education Global Dry -13.118 -10.308 -0.121 -0.941 -2.090 Global Wet -10.095 -5.854 -2.973 2.116 0.584 Ghana Dry -2.709 -3.009 -1.193 -1.782 -1.308 Ghana Wet -4.050 -0.766 1.936 1.358 1.795 Note: Key assumption: Resource envelope is externally �nanced and does not reduce Ghana’s baseline investment path. Source: World Bank 2010f. exception of Ghana Dry. In this scenario, the reallo- to hydropower investments. This is the climate sce- cation of funds from road network expansion to road nario with the strongest adverse impacts of climate hardening slows down road network growth without change on hydropower generation. generating net bene�ts, because climate shocks to the road system turn out to be very mild. Thus, this result Finally, the results for investment in education serve suggests, that in the case of Ghana (and contrary to to represent an adaptation strategy that does not the case of Mozambique), road design change is not directly address climate change impacts in particular an unequivocal no-regret adaptation measure. sectors, but is aimed at spurring growth performance in general in order to reduce vulnerability to nega- The simulated adaptation investments in agriculture tive climate change shocks. The illustrative results in combination with road design slightly overcom- reported here suggest that a broad-based education pensate for the climate change damages in a macro- strategy is cost-effective under the wet scenarios but economic sense under Ghana Wet. This means that in not under the dry scenarios. As with the strategy this case the total cost of returning aggregate welfare based on agricultural investments this argues for a to the baseline is actually lower than the assumed more targeted education-based adaptation strategy, adaptation investment expenditure of $4.05 billion. targeting sectors where labor productivity gains are In the other three scenarios, the agriculture-focused likely to be large. strategy restores aggregate real absorption close to the baseline level, but the negative signs in Table Local-level perspectives on adaptation. While the 9 indicate some residual damage. In these cases it economic analyses prioritized improved road infra- would appear advisable to channel the investments structure, energy, and regional integration (including selectively to crops and regions with high expected transboundary water management), these issues were returns and use the remaining part of the resource not raised in the social investigation. However, iden- envelope for lump-sum compensation payments. ti�ed adaptation options in the areas of agriculture and coastal zone development did largely echo those The comparison of the combined hydropower/agri- raised by participatory scenario development (PSD) culture adaptation strategy with the pure agriculture stakeholders and respondents in �eldwork. In addi- adaptation strategy suggests that only under the tion, adaptation preferences expressed in the PSD Global Wet climate scenario is it preferable to divert workshops largely coincided with priority action areas a fraction of the adaptation envelope from agriculture in the National Adaptation Plan of Action (NAPA) ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 53 as well as related climate strategy priorities in-country. adaptation to climate change go hand-in-hand with Discussions also focused on the need for improved “good� development policies: governance, social protection, land tenure, and training and education in order to accelerate development and Agriculture: Invest in R&D related to impacts of cli- build resilience to climate change. mate change on crops and livestock products and pest control, as well as early-maturing varieties; improve Interestingly, local participants in the zonal workshops water storage capacity to utilize excess water in wet were more concerned with declining living standards years; and improve agricultural extension services due to degraded natural resources and with the lack and marketing networks. Other required measures of public services as drivers of vulnerability than with include construction of small to mid-size irrigation exposure to climate-related events. Speci�c priorities facilities, improvement of the land tenure system, included a focus on improving agricultural production and improved entrepreneurial skills to generate off- techniques and land management practices; manag- farm income. ing migration; closing the gap in gender equity; and strengthening governance and institutional structures. Roads: Proper timing of road construction (for National workshop participants also focused on adapta- example, during dry season); routine and timely road tion measures that would offer cobene�ts with sustain- maintenance; upgrade road design speci�cations, able development, yet preferred adaptation measures including choice of materials; and consider drain- that were often more expensive and left little room for age and water retention, road sizes, and protection integration of inputs of local communities. of road shoulders. Lessons and recommendations Energy: Diversify current thermal and large hydro sources to include renewable sources such as the In view of the expected change in temperature and planned mid-size hydro Bui Dam and mini hydro. precipitation, strategic planning in Ghana should take regional climate change variability into consid- Coastal Zone: Improve shoreline protection in areas eration. At the national level, the National Develop- with economically important urban and port infra- ment Planning Commission’s draft Medium-Term structure; upgrade peri-urban slums and control National Development Policy Framework for 2010 development of new ones; protect, manage, and sus- to 2013 lays out the priorities of the government tainably use coastal wetlands; and review Ghana’s installed since February 2009. This framework was coastal development plans to take into consideration used to establish the baseline scenario of develop- climate change adaptations, including coastline and ment upon which this study is based. As the gov- port protection, flood protection, and coastal commu- ernment moves to implementation of this new plan, nities and �shery industry protection. recommendations from adaptation options presented in this study should be considered. Social: Improve social safety nets, community- based resource management systems, and disaster For each of the ten regions in Ghana, the possible preparedness. It is also necessary to accelerate the sets of climate change impacts described need to decentralization process to devolve decision making be addressed through the Regional Coordinating to the local level to promote local-level adaptation Councils, and at District-level through District and preparedness. Development Plans. Speci�c needs in each sec- tor are discussed below. As in the other countries West Africa Regional Integration: Ghana needs of the EACC study, policy recommendations for to enhance dialogue with neighboring countries 54 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT regarding the management of shared water resources, annual costs from natural disasters to the national and explore possible regional water resource manage- economy—in terms of damages to infrastructure and ment coordination in order to effectively deal with livelihoods and losses from forgone production— the challenges of climate change such as droughts, have been estimated at 0.5 percent to 1 percent of floods, and possible regional migration. GDP. These statistics do not include the signi�cant loss of life that has also occurred during these events. Bangladesh34 These damages and losses are geographically concen- trated in areas that also have higher concentrations of Vulnerability to climate change the poor, who are the most vulnerable and have the lowest capacity to address the impacts, hence are also Bangladesh is one of the most vulnerable countries in affected disproportionately. They live in thatch or tin the world to climate risks. Two-thirds of the nation houses that are more susceptible to direct damages is less than 5 meters above sea level and is susceptible from cyclones, storm surges, and floods. Additionally, to river and rainwater flooding, particularly during most rural households depend on weather-sensitive the monsoon. Due to its location at the tail end of sectors—agriculture, �sheries, and other natural the delta formed by the Ganges, Brahmaputra, and resources—for their livelihood. Destruction of their Meghna (GBM) rivers, the timing, location, and assets and livelihoods leaves the poor with a limited extent of flooding depends on the precipitation in the capacity to recover. entire GBM basin, not just on the 7 percent of the basin that lies within the country. Nearly 80 percent The importance of adapting to these climate risks of the country’s annual precipitation occurs during to maintain economic growth and reduce poverty is the summer monsoon season, when these rivers have thus very clear (Figure 21). Since the sixties, Bangla- a combined peak flow of 180,000 m3/sec, the second desh has invested $10 billion on disaster reduction highest in the world. Once every three to �ve years, measures, both structural and non-structural, and up to two-thirds of Bangladesh is inundated by floods enhanced its disaster preparedness system (Box 3). that cause substantial damage to infrastructure, hous- These measures have signi�cantly reduced damages ing, agriculture, and livelihoods. Low-lying coastal and losses from extreme events over time, especially in areas are also at risk from tidal floods and severe terms of deaths and injuries. In addition, public sec- cyclones. On average once every three years, a severe tor agricultural support services have enabled rural cyclone makes landfall on the Bangladesh coastline, households to reduce exposure to these risks and to either before or after the monsoon. The largest dam- maintain their livelihoods. For instance, they have ages from a cyclone result from the induced-storm adapted to “normal floods� by switching from low- surges, sometimes in excess of 10 meters. Bangladesh yielding deepwater rice to high-yielding rice crops, is on the receiving end of about 40 percent of the resulting in increased agricultural production. How- impact of total storm surges in the world. Crops and ever, it is the low-frequency, high-magnitude floods that the livelihoods of the rural poor in low-lying coastal have adverse impacts on livelihoods and production, areas are also devastated by saline water intrusion into particularly of the poorest and most vulnerable. The aquifers and groundwater and land submergence. In cost of strengthening and expanding these measures to addition, seasonal droughts occasionally hit the further reduce the risks from existing climate-related northwestern region. hazards is less than the avoided damages. Climate-related disasters continue to result in large A warmer and wetter future climate that goes beyond economic losses, reducing economic growth and historical variations will exacerbate the existing cli- slowing progress in reducing poverty. The direct mate risks and increase vulnerability by increasing the ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 55 extend and depth of inundation from flooding and EACC approach and results storm surges and by reducing the arable land due to sea level rise and salinity intrusion. The median pre- The Bangladesh case study builds on a parallel study dictions from these models are for warming of 1.55°C on the impacts of climate change and food security and an increase in precipitation of 4 percent by 2050. (Yu et al. 2010) and focuses on two speci�c climate The median temperature predictions exceed the 90th hazards: storm surges induced by tropical cyclones and percentile of historical variability across GCMs dur- inland flooding. The study (a) estimates the additional ing the summer months by the 2030s. Unlike temper- damages that would result in key economic sectors and ature changes, the predicted changes in precipitation in the overall economy if no additional adaption mea- (and discharges) through 2050 are not distinct from sures are put in place to address current and expected the historically observed variability for all months hazards, and (b) estimates the costs of additional and seasons, reflecting the large historical variability investments that would be needed to protect against in precipitation levels and the even larger uncertainty these hazards. The study also analyzes the differential in future precipitation predictions. Current trends for impact of climate change on vulnerable populations water levels in coastal areas suggest a rise in sea levels and how they cope with such impacts. of over 27 cm by 2050. Further, the increased sever- ity of cyclones in the Bay of Bengal is expected to Tropical cyclone-induced storm surges. The poten- increase risks of inundation in coastal areas by 2050. tial damages and the adaptation cost necessary to FIGURE 21 MAPS OF POVERTY, FLOODING AND TIDAL SURGES Poverty Map Map of Flooding and Tidal Surges Source: BBS, World Bank, WFP 2009. 56 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT BOX 3 PAST EXPERIENCE ADAPTING TO EXTREME CLIMATE EVENTS IN BANGLADESH Given its vulnerability to extreme climate events, a number of adaptation measures are already in place in Bangladesh, including both hard infrastructure as well as soft policy measures combined with communal practice. Hard infrastructure has included coastal embankments, foreshore afforestation, cyclone shelters, early warning systems, and relief opera- tions; soft measures have included design standards for roads and agricultural research and extension, such as the introduction of high-yielding varieties of aman and boro rice crops. Both types of adaptation measures have made the country more resilient in facing hazards, as evidenced by the decline in the number of fatalities and the share of GDP lost as a result of these events. Coastal embankments. In the 1960s and 1970s, 123 polders (including 49 facing the sea) were constructed to protect low-lying coastal areas. Polders have been an effective measure for protection against storm surges and cyclones, but breaching of embankments has been a recurring phenomenon due to overtopping, erosion, inadequate O&M, and other problems. Foreshore afforestation to protect sea-facing dikes. Foreshore afforestation is a cost-effective way to reduce the impacts of cyclonic storm surges on embankments by dissipating wave energy and reducing hydraulic load on the embankments during storm surges. The limited damages from Cyclone Sidr (2007) and Aila (2009) have been partially attributed to foreshore afforestation. Government of�cials have recommended that the existing forest belt include at least a 500-meter wide mangrove forest. Currently 60 km of forest belts exist on the 49 sea-facing pol- ders, span a total combined length of 957 km, leaving over 90 percent of the polder length unprotected. Cyclone shelters. Cyclone shelters are currently essential to protect human lives and livestock during cyclones hitting the coast. During the Cyclone Sidr in 2007, 15 percent of the affected population took refuge in cyclone shelters, sav- ing thousands of lives. Focus group interviews with area residents revealed that shelters have been limited in their use and effectiveness, mainly due to existing design, distance from the homestead, dif�cult access, the unwilling- ness to leave livestock behind, lack of user-friendly facilities for women and people with disabilities, overcrowding, and lack of sanitation facilities. Although the need for cyclone shelters is expected to decline with more effective protection through embankments combined with autonomous adaption with rising incomes, cyclone shelters will nevertheless be needed in areas where dikes are not cost effective (such as in small less inhabited islands). Early warning systems. Early warning and evacuation systems have played an important role in saving lives during cyclones. The Bangladesh Meteorological Department tracks cyclones and issues a forewarning that indicates areas that are likely to be affected by the cyclone storm. These warnings are broadcast through newspapers, television, and radio stations throughout the affected area. The existing evacuation operations managed by the local govern- ments can be improved by increasing the spatial resolution of the warning and indicating the severity of expected inundation. Repeated warnings in areas that are not ultimately affected reduce the con�dence of the inhabitants in the early warning system. Decentralization of relief operations. Relief operations were historically centralized in Dhaka, away from the actual impacts and affected population, resulting in a long chain of command and delayed effective relief. Recent efforts to decentralize operations have proven quite successful. They include the establishment of a forward operation center with a government appointed commander-in-chief to oversee operations, the use of high frequency and ultra high frequency transceiver radios, and cell phones as an emergency communication system. Pre-positioning of emer- gency relief materials and life-saving drugs and medical supplies is playing an increasingly important role in quickly initiating relief and rehabilitation activities. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 57 avoid these damages are estimated separately for current population density is lower in areas with a two scenarios—a baseline scenario without climate higher risk of inundation (Figure 21). However, not change and another with climate change. The base- all households are able or can afford to migrate away line scenario is developed from all 19 major historical from higher risk areas. Poverty rates are also highest cyclones making landfall in Bangladesh between 1960 in the higher risk areas (Figure 23). and 2009, and represents the risk in the absence of climate change currently, as well as in 2050. Climate The population and assets at risk and the damages change is expected to increase the severity of cyclones from cyclone-induced storm surges in 2050 under the by 2050. The storm surge conditions in 2050 under two scenarios are computed assuming an economy the climate change scenario are simulated based on with “normal� development patterns that is the same three anticipated effects: ( a) sea level rise of 27cm, (b) for both the baseline and climate change scenarios. an increase in the observed wind speed by 10 percent, The population is projected to increase 1.15 percent and (c) landfall during high tide. annually (reaching replacement fertility by 2021) with increases concentrated in urban areas. GDP is Cyclone-induced storm surges due to climate change projected to continue to grow at the current annual are expected to inundate an additional 15 percent rates of growth of 6 to 8 percent, but is expected to of the coastal area and also increase the inundation be less dependent on agriculture by 2050. Damages depth in these areas (Figures 22 and 23). Households from storm surges are based on spatially disaggre- have adapted to the existing risks by moving further gated projections of the population and assets poten- inland into areas with lower current risks; as a result, tially at risk. If additional public adaptation measures FIGURE 22 INUNDATION RISK FROM STORM SURGES IN COASTAL AREAS WITH AND WITHOUT CLIMATE CHANGE IN RELATION TO CURRENT POPULATION DENSITY Source: World Bank 2010c. 58 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT FIGURE 23 INUNDATION RISK FROM STORM SURGES IN COASTAL AREAS WITH AND WITHOUT CLIMATE CHANGE IN RELATION TO CURRENT POVERTY RATES Source: World Bank 2010c. are not put in place, the damages from a single typi- Increased monsoon precipitation, higher transbound- cal severe cyclone with a return period of 10 years is ary water flows, and rising sea levels resulting from expected to rise nearly �vefold to over $9 billion by climate change are expected to increase the depth and 2050, accounting for 0.6 percent of GDP, with the extent of inundation. The impacts of climate change burden likely falling disproportionately on the poor- are measured by comparing the inundation levels pre- est households. dicted by simulations using the MIROC 3.2 GCM predictions under the A2 emission scenario for 2050 Inland flooding. Rural households have adapted their with the inundation levels in the 1998 floods. Cli- farming systems to the “normal floods� that typi- mate change places an additional 4 percent of land cally inundate about a quarter of the country every area at risk of inundation. Further, inundation depth year by switching to high-yielding rice crops instead increases in most areas currently at risk, with increases of low-yielding deepwater rice. As a result, agricul- greater than 15 cm in about 544 km2, or 0.4 percent of tural production has actually risen over the past few the country (Figure 24).35 These are underestimates of decades. Severe flood events, however, continue to the actual increased risk from climate change as they cause signi�cant losses, both to agriculture and to the do not account for the frequent river course changes. transportation and communication networks and to The total inundation risks in 2050 are actually sub- the livelihoods of the poor once every three to �ve stantial, considering the increased risks are measured years. The 1998 flood inundated over two-thirds of relative to the 1998 flood. Despite the higher risks, Bangladesh and resulted in damages and losses of the rural population exposed to flooding is expected over $2 billion (4.8 percent of GDP), approximately to decline from current levels due to the rural-to- equally split between infrastructure, agriculture, and urban migration projected to occur by 2050. These industry/commerce. risks are in addition to the substantial baseline risks ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 59 FIGURE 24 CHANGE IN INUNDATION DEPTH (CM) FROM MONSOON FLOODING (WITH AND WITHOUT CC) Source: World Bank 2010c. that currently exist from inland flooding. Damage crop damages from severe flooding, and losses of cul- estimates from the agriculture component indicates tivable land due to rising sea levels. Considering all that climate change increases the existing damages by climate impacts (CO2 fertilization, temperature and about one-third, suggesting that actions to manage precipitation changes, flooding, and sea level rise), current severe floods are a good no-regrets strategy cumulative rice production is expected to decline by for adapting to future climate change. 80 million tons (about 3.9 percent each year) over 2005–50, driven primarily by reduced boro crop pro- Agriculture. The climate change and food security duction (Figure 25). Agricultural GDP is projected study examines the impacts of predicted changes in to be 3.1 percent lower each year ($36 billion in lost climate on crop yields, agricultural production, GDP, value-added) and total GDP $129 billion lower due and household welfare. Crop yields are separately to climate change over the 45-year period 2005–50. modeled for 16 different agroecological regions, with rice split by seasonal varieties using climate predic- Adaptation tions from 16 GCMs for three emission scenarios. In addition, the impacts of severe flooding on agri- The costs of adaptation under the two climate sce- cultural production are assessed using �ve GCMs narios are estimated through a gap analysis taking and two emission scenarios. The models predict that into account the adaptation investments already in higher yields of the main rice crops aman and aus— place. The costs under the baseline scenario corre- resulting from higher concentrations of CO2, rising spond to the adaptation de�cit, while the cost differ- temperature, and heavier precipitation—will be more ence between the two scenarios represents the cost of than offset by declines in the yield of the boro crop, adaptation due to climate change. 60 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT FIGURE 25 PROJECTED NATIONAL RICE PRODUCTION WITH AND WITHOUT CLIMATE CHANGE 90,000 Average Annual Growth Rates, 2005-50 2.0 Optimal Scenario: 3.03% 80,000 1.8 Existing Variability Scenario: 2.71% Share of Discounted GDP (%) Average CC Scenario: 2.55% 1.6 70,000 1.4 60,000 1.2 1.0 50,000 0.8 40,000 0.6 Baseline Scenario A2 & B1 0.4 30,000 Average CC Scenario A2 only 0.2 Optimal Scenario B1 only 20,000 0.0 2005 10 15 20 25 30 35 40 45 50 2005 10 15 20 25 30 35 40 45 50 Source: Yu et. al 2010. Tropical cyclone-induced storm surges. Since the inundation on existing infrastructure—road network 1960s, Bangladesh has made signi�cant investments and railways, river embankments, and embankments in embankments, cyclone shelters, coastal afforesta- to protect highly productive agricultural lands, drain- tion and in disaster preparedness to address the risks age systems, and erosion control measures for high- from cyclones and storm surges. However, these value assets such as towns. The total cost of adaptation investments are not sufficient to address the exist- due to climate change to address inland flooding risk ing risks and much less the future risk from climate is $2.7 billion in initial investment and $54 million in change. Adaptation measures evaluated were (a) annual recurrent costs (Table 12). Full protection in embankments, (b) afforestation, (c) cyclone shelters, 2050 will also require addressing the existing baseline and (d) early warning systems. Protecting Bangladesh risks of flooding, which are likely to be at least of the against existing storm surge risks requires $3,090 mil- same of order of magnitude or larger. lion in initial investments and $62 million in annual maintenance costs. Addressing the additional risks Agriculture. While the public sector cost of adapt- due to climate change will require additional invest- ing in the agriculture was not estimated, the rela- ments of $2,426 million and annual maintenance tive merits of a number of short-term adaptation costs of $50 million by 2050 (Table 11). Despite dif- measures—namely the extension of currently avail- ferences in methodology, climate scenarios, economic able options into new areas—are examined from the assumptions, and scope of coverage, these costs are of farmer’s perspective. Part of the longer term adapta- the same order of magnitude as the adaptation costs tion strategy will be to control the damages from estimated for Bangladesh from the global track of the inland floods. This has been partially costed in the study of around $13 billion over the 40-year period. inland floods component of the study. In addition, longer term adaptation has to also include develop- Inland flooding. The analysis focuses on adaptation ment of alternatives, particularly to the boro crop in measures to avoid further damage from additional the southern region. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 61 TABLE 11 COST OF ADAPTING TO TROPICAL CYCLONES AND STORM SURGES BY 2050 ($ millions) Baseline Scenario (additional risk CC Scenario (existing risks) (1) due to CC) (2) (total risk = (1) + (2)) Adaptation Option IC AMC IC AMC IC AMC Polders 2,462 49 893 18 3,355 67 Afforestation 75 75 Cyclone shelters 628 13 1,219 24 1,847 37 Resistant housing 200 200 Early warning system 39 8 39 8 Total 3,090 62 2,426 50 5,516 112 Note: CC = climate change; IC = investment cost; AMC = annual maintenance cost Source: World Bank 2010c. TABLE 12 TOTAL ADAPTION COST FOR INLAND FLOODING BY 2050 ($ Million) Adaptation Option Investment Cost Annual Recurrent Cost Transport – Road height enhancement 2,122 42 Transport – Road cross-drainage 5 - Transport – Railway height enhancement 27 1 Embankment – height enhancement 96 2 Coastal Polders – cross drainage 421 8 Erosion Control Program 1 Total Costs 2,671 54 Source: World Bank 2010c. Local-level perspectives on adaptation options for livelihood diversi�cation and also have low social capital. Past adaptation practices by households vary accord- ing to hazard type and asset base holdings. The most Participants in local and national participatory sce- common form of adaptation is temporary migration nario development workshops identi�ed preferred for day labor work by adult men (undertaken by 37 adaptation options in environmental management percent of surveyed households). Storage of food and (mangrove preservation, afforestation, coastal green- drinking water before extreme events is also common, belts, and waste management); water resource man- and 25 percent of surveyed households also reported agement (drainage, rainwater harvesting, drinking building livestock platforms to guard animals during water provisions, and flood control); infrastructure such events. Adaptive capacity among all �eld sites (roads and cyclone shelters); livelihood diversi�ca- was low; in particular, poor urban dwellers face few tion and social protection for �shers during cyclone 62 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT season; education; agriculture (development of salt tolerant and high-yield varieties, crop insurance); �sheries (storm resistant boats, conflict resolu- tion between shrimp and rice farmers); governance (especially access to social services for the urban poor); and gender-responsive disaster management (separate rooms for women in cyclone shelters, mini-shelters closer to villages, use of female voices in early warning announcements, and mobile medi- cal teams in Char areas). Lessons and recommendations Given the pervasive impacts of climate-related risks over time, Bangladesh is also one of the most climate resilient countries and can provide many lessons on developing climate resilient strategies for other devel- oping countries. Yet, damages from recent cyclones and floods indicate that substantial risks remain. De�ciencies in costal protective measures weaken resilience to existing cyclone-induced storm surges, and climate change is expected to nearly double these risks. Further, the aggregate additional costs of the proposed adaptation measures needed to mitigate climate change risk from extreme events are generally smaller than the expected damages. As a result, a no- regrets strategy would be to begin by addressing the adaptation de�cit and strengthening early warning systems. Additional embankments and shelters can be constructed in the medium term as the geographic incidence of risk becomes more certain. The impacts of existing climate variability are concen- trated in areas that also have higher concentrations of poor and socially vulnerable populations. Climate change does not shift these distributions, but just exac- erbates them. The rural poor in the southern region in particular are expected to face the largest declines in per capita consumption, as well as declining productivity in the aus and aman rice crops, severe yield losses in the boro crop, and land losses due to increased salinity brought forth by sea level rise. Though the government has made substantial investments to increase the resil- ience of the poor (e.g. new high-yielding crop varieties, ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 63 protective infrastructure, disaster management), the and droughts, all of which have serious implications scale of the current efforts remains limited and will for food security and water supply, are common cli- need to be scaled up commensurate with the probable mate-related events. impacts from climate change. Figure 26 shows the influence of accumulating By 2050, the number of people living in cities will extreme events on agricultural GDP. The negative triple while the rural population will fall by 30 per- impact of strong El Niño events (red) is clear in the cent. The long-term challenge is to move people and years 1982–83, 1991–92, and 2005–06. Also visible is economic activity into less climate-sensitive areas. A the slight improvement in agriculture management in strategic balance between protecting existing popula- 2003–04. The less severe effects from La Niña (blue) tions and encouraging the mobility of future popula- also are evident in the years 1985–86, 1988–89 and tions must be sought. Current policies will determine 1994–95. where this urban population settles and how prepared it is to adapt to a changed climate. Good policy will Most of the climate models for future projections do encourage future populations to move away from not agree with regard to rainfall estimates in terms of areas of high natural risk. This requires avoiding the sign of the change, intensity, and geographical dis- perverse incentives to remain in high-risk areas and tribution within Bolivia (Figure 27), showing a range adopting positive incentives to promote settlement of plausible wet and dry scenarios.37 The economic and urban growth in low-risk areas. and population welfare impacts under a changing climate are thus somewhat uncertain. Higher tem- Lastly, although Bangladesh accounts for only 7 per- peratures and fewer frosts might stimulate agricul- cent of the Ganges-Brahmaputra-Meghna (GBM) tural production in the Altiplano and the valleys. The basin, due to its geographical location at the tail end key uncertainties concern the total amount, timing, of the basin, flooding in Bangladesh depends on and intensity of precipitation. If the dry scenarios the rainfall in the entire GBM basin. Institutional are correct, then the bene�ts of higher temperatures arrangements on the sharing and management of will be more than offset by more frequent and severe water resources with its neighbours will be just as periods of low rainfall, especially in the southwest, important in managing floods. together with an uncertain effect in the north, mak- ing the case for improved water storage and irrigation Bolivia36 infrastructure. Vulnerability to climate change On the other hand, if the wet scenarios are correct, then agricultural yields should increase throughout The Bolivian population has always been exposed much of the country, but this would also require to hydro-meteorological extremes and climate vari- upgrades in infrastructure (water storage and flood ability, particularly because of the influence of the El control) together with improved agricultural prac- Niño oscillation (ENSO), which—regardless of cli- tices and land management. mate change—occurs periodically in different areas across the country. The impact of El Niño 2006–07 Bolivia’s economic mainstays—mining and hydro- in Bolivia cost approximately $443.3 million in dam- carbon extraction—suggest the nation is relatively ages. Direct damage to property accounted for about insensitive to climate change. However, it is the half of the total; the remaining 45 percent included large majority of its rural and urban population who losses in cash flow, declines in production, reduced are quite vulnerable to changes in climate. Bolivia’s income, and disruption of services. Floods, landslides, rural population relies mostly on rainfed agriculture, 64 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT FIGURE 26 ANNUAL PERCENT CHANGE IN AGRICULTURAL GDP FROM EL NIÑO AND LA NIÑA EFFECTS 15,0 10,0 5,0 Percent Change -0,0 1980 1981 1982 1983 1984 1985 1986 1988 1989 1990 1991 1992 1993 1994 1995 1996 1998 1999 2000 2001 2002 2003 2004 2005 2006 1987 1997 2007 -5,0 -10,0 -15,0 El Niño La Niña Historical Variablity Source: World Bank 2010d. FIGURE 27 PROJECTED PRECIPITATION CHANGES TO 2050 UNDER DIFFERENT CLIMATE SCENARIOS Source: World Bank 2010d. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 65 small-scale livestock farming, and seasonal labor in percent, whereas the dry scenario shows a tempera- agriculture. Approximately 30 percent of Bolivia’s ture increase of 2.4˚C and a decrease in precipitation rural residents reside in the valleys and high plateau of -19 percent averaged across the Bolivian terri- areas, where water availability is already problematic. tory. Models indicate that the frequency of extreme In addition, a high proportion live in extreme con- weather events might increase, including the onset ditions without the necessary resources to adapt to of El Niño and La Niña events. The accumulation of climate change. such extreme events within shorter time frames can easily threaten the development-as-usual patterns in Climate change will not only affect rural areas within Bolivia, given the serious public sector �nancial limi- Bolivia. Several major cities located in the upper tations to reconstruct and recover to previous levels watersheds in the Altiplano and Valley regions—such of welfare. as La Paz-El Alto, Sucre, Potosí, and Cochabamba— are signi�cantly vulnerable to climate variability and As one of its main objectives, the EACC study evalu- water scarcity. These cities are highly vulnerable to ated the robustness of planned adaptation measures decreasing rainfall trends, to unexpected changes in under different climate scenarios in terms of �nan- seasonality, and to prolonged droughts. Water short- cial values and socioeconomic terms. Results dem- ages have already incited social conflicts in Cocha- onstrated that most current planned investment in bamba, Sucre, and Tarija. agriculture and water resources continue to be robust to climate change, at least under extreme conditions. Investment in better water management will enhance Thus, current adaptation measures in Bolivia are rep- the resilience of Bolivia’s population, both to system- resentative of primarily no-regret development strat- atic changes associated with annual levels of rainfall, egies under climate variability. as well as greater year-to-year volatility in rainfall patterns. Improved water management practices are The EACC in Bolivia consisted of three different conducive to smart development even in the absence economic assessments regarding the costs, bene�ts, of climate change, making this a robust no-regrets and sequencing of alternative adaptation measures at investment given the prevailing uncertainty about different levels. The �rst exercise assessed the robust- future climate change in the area. ness of planned investments in the water sector by evaluating costs and bene�ts of government-selected EACC approach and results adaptation measures reflecting the types of needed adaptation actions previously identi�ed by the Bolivia Based on continuous dialogue with the government, National Mechanism of Adaptation. The second exer- the EACC study in Bolivia focused on two vulnerable cise evaluated the cost of stylized adaptation options sectors: agriculture and water resources. In addition, for the water sector—mainly infrastructure needs for a social component complemented the sector-based increased irrigation. economic analysis and shed light on the implications of different adaptation options on poor and vulner- The third assessment comprised the development of a able groups. The study considered two extreme cli- planning investment tool to evaluate the sequencing mate scenarios in terms of water availability in order and prioritization of adaptation options under cli- to simulate the worst case scenarios, assuming that mate and development constraints. The analysis was changes in the Bolivian climate are likely to occur able to identify the most vulnerable population, and somewhere between these two. The wet scenario how to restore watershed-level bene�ts to their base- forecasts an average temperature increase of 1.5˚C line levels through accelerated investment. This type and an annual mean precipitation increase of +22 of planning model permitted a detailed comparison 66 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT of investment alternatives and the potential effect 2,000 mm in the warm lowlands, and maximum values of climate change on each alternative. The approach of 5,000 mm in certain transition areas from the valley facilitates investigation of the sequencing and priori- to the lowlands. tization of actions in a certain time frame, as well as the robustness of alternative investment and policy Rural Areas. According to most future climate strategies to possible climate outcomes. The water- projections, access to water resources in rural shed planning model developed for this investigation areas will be impacted by two major water-related is a practical, useful planning tool that can be used climate risks: gradual changes in the magnitude by Bolivian authorities, re�ned, and updated as addi- and distribution of precipitation and temperature, tional climate and watershed data become available. and changes in the frequency and magnitude of extreme events. In addition, local evidences of Impacts rapidly melting glaciers may exacerbate water shortages in the arid and semi-arid valleys and in Water. Water resources are abundant in Bolivia. Aver- the highlands, which already lack water storage age rainfall is about 1,200 mm38, and despite high capacity. Glaciers act as a buffer for water avail- evaporation rates, average water allocation is high at ability during dry periods, and in Bolivia they are approximately 45 m3 per capita per year.39 However, shrinking at an alarming rate. natural water supply presents both a marked geo- graphical and seasonal variability: 45 percent of the Water Supply and Sanitation in Urban Areas. In rainfall falls within 3 months (December–February), many cases—such as in Cochabamba, Sucre, or with values from 100 to 600 mm in the cold Altiplano Tarija—the competition for water resources is and less cold central and southern valley regions, up to high, and social conflicts are frequent between ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 67 the urban utility and different user communi- While water resources are abundant for the whole ties. The case of La Paz-El Alto, is particularly country, improving the storage efficiency of wet peri- worrying due to disappearance of the glacial con- ods to meet irrigation demands in de�cit areas such as tribution to the super�cial runoff, which, though the south of the Altiplano and El Chaco is essential. not properly quanti�ed, will provoke a reduction Improvements in irrigation need to be accompanied in the amount of natural water supply and pose by better water management, particularly integrated an extra threat on this metropolitan area, where watershed management where the resource compe- demand has already matched supply.40 tition between rural and urban populations is likely to increase. In addition, there is a need to reinforce, Agriculture. The crops analyzed were quinoa, potato, improve, protect, and diversify water sources in order maize, and soy, which are cultivated from the Altiplano to strengthen the production capacity of the urban to regions of lower elevation. Generally, Bolivia’s agri- utilities, especially in cities of the arid regions like La culture would bene�t signi�cantly from a warmer and Paz, Cochabamba, or Sucre. wetter climate, so long as the varieties and crops that are grown can be adjusted to changes in rainfall pat- Under the wet scenarios, there will be an increase terns during critical phenological time periods and/or in flooding, especially in the valleys and the eastern any shortening in the growing season. Yields for maize lowlands. Reforestation, as well as the development and soybeans would increase by 40–45 percent, while of flood warning and disaster prevention systems, can that for potatoes and quinoa by 60–90 percent. On the help reduce the economic and social costs of flood- other hand, the dry scenarios would lead to a substan- ing in lowland areas. More expensive forms of flood tial reduction in agricultural yields in the Altiplano, the prevention such as dikes are rarely justi�ed. valleys, and the El Chaco regions. The effects of less rainfall and higher evaporation could only be offset Agriculture. According to the estimated impact of by (a) a substantial investment in water storage and climate change, similar adaptation options for the irrigation infrastructure, and (b) the adoption of more four studied crops were identi�ed as crucial, irriga- drought-resistant varieties and crops in the lower lands. tion being clearly the most important. In addition, Potential losses from a drier climate are on the order of for quinoa, the application of de�cit irrigation and 25 percent for maize and 10–15 percent for soybeans, changes in the sowing dates and crop varieties are potatoes, and quinoa. This suggests that rapid and viable options; for potatoes, better management of timely implementation of irrigation (at least at the ini- the different varieties, changes in sowing dates, and tial phases of crop development), would be even more application of irrigation in critical phenological attractive under a scenario of warmer climate. periods; for soybeans, investments in flood control measures as well as the introduction of input saving Adaptation varieties; and for maize, speci�c additional adaptation measures include flood control in wet periods, as well Water. Investment in better water management will as improved soil management practices. Most adap- enhance the resilience of Bolivian agriculture both to tation strategies will require signi�cant institutional systematic changes in annual levels of rainfall and to support in order to avoid negative social and ecologi- greater year-to-year volatility in the rainfall patterns. cal impacts due to intensi�cation of crop production. Such investments are “no-regret� measures and would be desirable under most development strategies for Adaptation measures indicated by local popula- a stable climate, so that climate change is likely to tions include the need for better information reinforce the bene�ts of such investments. and capacity building initiatives geared toward working with new and adapted seed varieties, as 68 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT well as better infrastructure for conservation and usually excluded from national budgets and mostly storage of crops during warm periods. Extension �nanced by international cooperation. services, crop insurance, and improved access and availability to hydro-meteorological data will also The analysis was made in terms of �nancial (market) be vital to improve agriculture adaptation policies values and in socioeconomic terms (shadow prices), and meet the needs of livelihoods based on rain- and integrated climate change variables (tempera- fed agriculture. Some measures that remain to be ture and precipitation) under a dry (worst case) and explored include the potential role of investments a no-climate-change scenario. The objective was not in rural roads in providing the infrastructure to evaluate the projects themselves, but rather their required to facilitate shifts in the location of agri- economic feasibility and robustness as appropriate cultural production linked to changes in compara- adaptation measures to climate variability in Bolivia tive advantage. Robust adaptation options should (Table 13). be accompanied by large cobene�ts and interlink- ages. For example, investments in rural roads can The results suggest that the Altiplano will be favored be an important adaptation strategy because they by increased temperatures, while the oriental and increase access to markets for agricultural inputs Chaco zones will be negatively affected by increased and outputs. The implementation of policies that temperatures and reduced precipitation. These results increase access to markets—bridging two vulner- are in accordance with the spatial distribution of the able sectors, infrastructure and agriculture—could projects where, depending on the area, the Internal be implemented for small-scale farmers, and com- Rate of Return (IRR) is reduced due to these regional plemented by additional macroeconomic measures impacts. The agriculture projects show a slight increase for large-scale farmers. of the IRR under the climate change scenario in the highland zones (except the B.R.Paraisito project). This Finally, the agriculture study suggests that planned suggests that current planned investment in agriculture adaptation actions—including increased irrigation and water resources continue to be robust to climate resources and flood control measures, as well as the change, at least under dry extreme conditions.41 Thus, implementation of knowledge support for improved adaptation measures in Bolivia represent primarily analysis of climate science—are highly costly and good development strategies under climate variability. hard to implement in Bolivia, based on qualitative assessments of cost, bene�ts, and viability of these Two other economic exercises explored the possible adaptation actions. effects of climate change on a long-term irrigation program (National Watershed Program – PNC by its Economic analysis of adaptation investment Spanish acronym) at the watershed level. options The �rst exercise considered the cost of providing The cost-bene�t analysis illustrates the use of an eco- the required level of additional water storage infra- nomic tool for the evaluation of robustness of invest- structure to meet PNC’s planned irrigation expansion ment projects under a changing climate. The analysis to 2011 and estimated up to 2050. The analysis was evaluated projects selected primarily based on the based on balances of water de�cit and water surplus availability of data. Water projects included water months, and therefore the necessity and potential to supply and water management, and the agricultural reallocate additional water through storage, under a projects consisted primarily of irrigation projects. wet and a dry extreme climate scenario. The estimated The analysis excluded the assessments of larger infra- cost of the additional water storage required to match structure projects in urban areas, as these projects are monthly water de�cits due to climate change under ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 69 the wet climate scenario would be on the order of of families receiving irrigation versus maximizing the an additional $12 million over the projected baseline economic bene�ts from irrigation. (estimated at around $480 million for 2050 under no climate change), and an additional $60 million under Results showed that incorporating the effects of cli- the dry climate scenario up to 2050. mate change appears to slightly modify the original national watershed development plan, suggesting The second exercise42 explored the effect of climate that most of the potential irrigation investments in change on PNC’s planned investment program for the the Mizque watershed are robust to climate impacts. Mizque watershed. This is a watershed that has been This is the case because major vulnerability problems identi�ed as being particularly susceptible to climate are upstream and related to urban water supply, sani- effects. The study evaluated the effect of a changing tation, and threats of floods and droughts. This analy- climate on decisions to make durable investments. As sis suggests that farther downstream in the Mizque irrigation must compete with potable water and water River watershed, annual rainfall would remain suffi- for agriculture practices, two policies were explored cient for nearly all the irrigation projects identi�ed in under extreme climate conditions and budget con- the PMIC-Mizque study, assuming sufficient storage straints: (1) decentralization of budgets to the sub- was built as part of the program. Under a “dry sce- basin level versus centralized Mizque watershed-level nario,� the effect would be to reduce potential social watershed planning; and (2) maximizing the number bene�ts by 1–3 percent as water becomes scarce TABLE 13 COST-BENEFIT ANALYSIS OF ADAPTATION MEASURES IN THE AGRICULTURE AND WATER SECTORS Investment Project Costs (000) Bene�ciaries NPV1 (000) IRR (%) NPV* (000) IRR (%) Baseline Dry scenario WATER Distribution in Sapecho 3,440 2,199 persons 3,428 24 3,331 24 Potable water S.P. Cogotay 408 140 persons 8 13 3 13 Well drills Chapicollo 317 50 families 187 17 151 17 Flood Control Caranavi 4,052 528 houses 2,658 22 2,658 22 AGRICULTURE Irrigation dam S.P.Aiquile 11,476 147 ha 2,583 16 4,195 18 Dam restoration Tacagua 313,623 907 ha (184,275) 3 (171,580) 3 Wall elevation Tacagua dam 120,457 907 ha 9,705 14 21,563 16 Irrigation B.Retiro S Paraisito 3,686 178 ha 17,260 71 14,874 63 Catchment Atajados/Aiquile 1,951 32 ha 115 14 347 16 Notes: parenthesis values indicate a negative NPV, suggesting that dam resto- ration is not economically feasible in this location. * NPV = Net present value. Source: World Bank 2010d. 70 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT and there are diminishing economic returns from tended to favor more comprehensive strategies that sequenced investment projects. The effect of the wet consisted of a mix of hard and soft options. scenario is to increase bene�ts by 1–3 percent. These results vary with varying assumptions regarding the Lessons and recommendations budget available each year and the degree of budget- ary decentralization policies. There is little practical difference between Bolivia’s development agenda and the adaptation agenda. Thus, In general, the effect of budgetary decentralized there is a need to accelerate the development agenda, management43 at the subwatershed level overwhelms as in most cases, good development policies are the the effect of a dry or a wet climate change scenario, most robust adaptation policies. While the country regardless of whether the objective is to maximize has always experienced a high degree of climate vari- social bene�ts or to maximize the number of families ability, climate change is expected to intensify the directly bene�tting from the projects. According to phenomenon. Since the Bolivian economy is heav- the model exercise, decentralizing budgets at the sub- ily dependent on minerals and gas, it is not expected watershed level actually reduces social bene�ts and/or to be highly impacted by climate change. However, the number of families directly bene�tting from the the majority of the rural and indigenous populations projects by between 2 percent and over 30 percent. are dependent on agriculture, which in turn is highly impacted by changes in climate. Climate change Local-level perspectives on adaptation. Communities therefore will tend to intensify the already severe dis- in the valleys and highlands considered drought to be tributional problems of the country, thus calling for the principal threat to their livelihoods and prioritized an even stronger people-centered development. adaptation measures related to water management— including improving water storage capacity and irriga- The two possible climate trends—warmer and wet- tion infrastructure—followed by improved agricultural ter, and warmer and drier—will imply quite different and livestock practices. In contrast, communities from outcomes. Even in the more optimistic scenario of the Chaco and Plains regions asserted that improved wetter conditions, agricultural productivity can only agricultural practices were most important and consid- increase if the capacity to store and use the needed ered water management measures to be of secondary additional water is available for farmers and poor signi�cance. These local perceptions coincided with the peasants. Given the great uncertainties about future adaptation measures identi�ed in the sector analyses. precipitation patterns, strategies that will work well under both wet and dry conditions are called for. Yet results from participatory scenario development Such a strategy should include a combination of workshops and �eldwork demonstrate that communi- improved water resources management and building ties view adaptation strategies not as isolated “hard� water storage and irrigation infrastructure. These are measures nor as single projects, but rather as a set of no-regret strategies that should be pursued irrespec- complementary measures comprised of both hard tive of climate change. and soft adaptation actions. Thus, while infrastructure investments would be necessary, they emphasized that Vietnam44 these would be insufficient if complementary efforts are not made to promote capacity, institutional develop- Vulnerability to climate change ment, and in many cases, fundamental transformation to underlying logic and livelihood strategies. Notably Vietnam is a long narrow country consisting of too, local authorities tended to favor investment in an extensive coastline, two major river deltas, and discrete, hard measures, while community members mountainous areas on its eastern and northeastern ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 71 BOX 4 LOCAL FACTORS INFLUENCE LOCAL ADAPTATION PREFERENCES IN BOLIVIA The social component in Bolivia identi�ed the very wide variety of envisaged livelihood strategies in fourteen com- munities and highlighted the importance of past experience and support from local institutions in determining local adaptation preferences. More speci�cally, adaptation strategies tended to reflect the order of priority assigned to the same type of adaptation measure in the past. In effect, this shows how preferred adaptation strategies depend on the recent history of a particular community. For example, communities that have bene�ted from investments in water management schemes that have resulted in safer drinking water do not consider water management for improved drinking water as necessary for their future, as they do not view the current system as inadequate. The presence or lack of institutions is a second determinant for identifying, prioritizing, and sequencing adaptation strategies in Bolivia. Where local authorities and privatized institutions have a history of supporting development, community members will count on their continued support and prioritize measures that require external support. Where institutions do not have a strong presence, prioritized adaptation options will not be based on major external support. borders. As such, Vietnam is heavily exposed to the with the March–May rainfall reductions being higher risks of weather variability and climate change. Its in the southern part of the country; the wet seasons vulnerability to weather risks has given the country are projected to get wetter, with the June–August experience in designing and implementing measures rainfall increases being higher in the northern part of to mitigate the effects of droughts, flooding, storms the country. Hence, it is expected that rainfall will be and similar events on agriculture and other sectors concentrated even more than now in the rainy sea- of the economy. Assessing the potential impacts of son months, leading to an increase in the frequency, climate change and determining how best to adapt intensity, and duration of floods, and to an exacerba- represents a new challenge, for which past experience tion of drought problems in the dry season. Sea level may be a guide but which is accompanied by large is projected to rise approximately 30 cm by 2050 and uncertainties. up to 75 cm by 2100 under the medium scenario. In June 2009, the Ministry of Natural Resources and An analysis of vulnerability to climate change at the Environment (MoNRE) published Vietnam’s official regional level was carried out as part of the social scenario for climate change. The MoNRE scenario analysis. Exposure to climate change is assessed by falls in the middle of a range of alternative climate considering the numbers of households potentially scenarios for Vietnam when these are arranged by threatened by the effects of storms, flooding, salinity their climate moisture indexes. In addition to the intrusion, sea level rise and storm surges, landslides MoNRE scenario, the EACC study has made use of and flash floods, and drought. Each region is assigned two other climate scenarios—Dry (IPSL-CM4) and to categories ranked from 0 (low exposure) to 4 Wet (GISS-ER)—which represent the extremes of (severe exposure). Similarly, sensitivity to the impacts the distribution by climate moisture indexes. of climate is assessed on criteria that reflect vulner- ability to the consequences of climate change based Rainfall projections across seasons are of particular on speci�c socioeconomic characteristics—poverty, interest. The dry seasons are projected to get drier, economic diversi�cation, education, and health and 72 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT sanitation—and for speci�c social groups, such as of climate change is highest in the North West and ethnic minorities, women and children, migrant pop- Central Highland regions. The correlation between ulations, and urban populations. Again, each region is exposure and sensitivity is negative, so that regions assigned to categories ranked from 0 (low sensitivity) with high exposure tend to have low sensitivity and to 4 (extreme sensitivity). Unweighted averages of the vice versa. The only region with indexes that are classi�cations were computed to give indices of expo- above the average on both measures is the Mekong sure and sensitivity. These are shown in Table 14. River Delta. The analysis indicates that exposure to the effects EACC approach and results of climate change is highest in the Central Coastal regions (NCC & SCC) and in the Mekong River Methodology. The sector studies concentrated pri- Delta. On the other hand, sensitivity to the effects marily on Vietnam’s rural economy because that is TABLE 14 REGIONAL VULNERABILITY TO CLIMATE CHANGE Red North South Mekong North North River Central Central Central South River Region West East Delta Coastal Coastal Highland East Delta Exposure Storms 1 3 4 4 4 2 2 3 Flooding 1 1 4 4 4 2 2 4 Salinity 0 0 1 2 2 0 1 4 SLR 0 0 2 2 2 0 3 4 Landslides 3 3 1 3 3 2 1 1 Drought 2 2 1 4 4 4 2 2 Average 1.2 1.5 2.2 3.2 3.2 1.7 1.8 3.0 Sensitivity Poverty 4 3 2 4 2 4 1 2 Economic 4 4 2 4 3 4 2 2 diversi�cation Education 4 3 1 2 2 2 1 3 Health & sanitation 4 1 2 1 1 1 1 3 Ethnic minorities 4 3 0 1 1 4 1 2 Women & children 4 3 1 2 3 3 1 2 Migrants 0 0 2 2 1 4 4 1 Urban households 0 0 2 1 1 0 4 3 Average 3.0 2.1 1.5 2.1 1.8 2.8 1.9 2.3 Source: World Bank 2010i. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 73 where the effects of climate change are likely to be Step 2: Consider the relevant climate variables for most immediate and serious. Detailed studies were the sector and identify changes projected to 2050 or carried out for agriculture (crop production), aqua- beyond for each of the climate scenarios. This made culture, forestry, and coastal ports. Each of the sec- use of detailed information on precipitation by season tor studies follows a broadly similar approach that and/or region. involved the following steps: Step 3: Identify the impact of changes in climate on Step 1: Establish a baseline scenario consisting of resource productivity and land use. This included the projections of land use, production, value-added, effect of changes in seasonal temperatures on rice population growth, urbanization, and other variables yields or of seasonal precipitation on coffee yields, as without climate change. This provides a reference well as the effect of flooding or saline intrusion on the scenario against which the impacts of climate change amount of land that can be used for rice production without and with adaptation are measured. in the Mekong River Delta. 74 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT Step 4: Using GIS and other techniques, combine event have been economic under a wide range of cli- the information collected in steps 2 & 3 to estimate mate conditions. the overall impact of climate change on land use and production of crops, freshwater �sh, timber, and so Agriculture. The impact of the alternative climate sce- on by comparing estimates of yields and production narios on crop production has been examined using under no climate change and with climate change projections of runoff, which affects the availability but no adaptation. of irrigation water, plus agronomic models that take account of temperature and rainfall patterns, water Step 4A: For agriculture, incorporate the results from availability for rainfed and irrigated crops, and other Step 4 into a macroeconomic model to assess the factors to estimate the impact of climate change on consequences of changes in agricultural output on crop yields. agricultural prices, trade, GDP, and economic activity in other sectors and household consumption. Changes in yields without adaptation vary widely across crops, agroecological zones, and climate scenar- Step 5: Identify opportunities for (a) autonomous ios. As for other EACC studies, the results reported do adaptation undertaken by farmers and other producers not take account of CO2 fertilization, partly because in response to changes in climate and other conditions, of the uncertainties about the extent of the effect and and (b) planned adaptation, which is likely to be initi- partly to assess how far adaptation can mitigate the ated and at least partly funded by the government. worst-case outcome. For rice, the Dry scenario would lead to reductions in yields ranging from 12 percent Step 6: Estimate the production of crops, timber, and in the Mekong River Delta to 24 percent in the Red so on under the new climate conditions after the River Delta. The primary influences on rice yields are adaptation measures have been implemented. This the increase in average temperatures and reductions provides the basis for identifying (a) the effect of cli- in runoff in many months of the year (Figure 28). mate change with adaptation (the difference between the baseline scenario and the scenario of climate There would be more extensive inundation of crop change with adaptation), and (b) the impact of adap- land in the rainy season and increased saline intrusion tation itself (the difference between the scenarios of in the dry season as a consequence of the combina- climate change without and with adaptation). tion of sea level rise and higher river flooding. For the Mekong River Delta, it is estimated that about Step 6A: As for Step 4A, incorporate the results from 590,000 ha of rice area could be lost due to inunda- Step 6 into the macroeconomic model to assess the tion and saline intrusion, which accounts for about 13 bene�ts of adaptation in terms of aggregate and sec- percent of today’s rice production in the region. toral economic activity and household consumption. Table 15 shows the potential impact of climate Many of the adaptation options are “no regrets� change without adaptation under alternative climate options that increase yields or production even scenarios on production of six major crops or crop without climate change. This is not invariably the categories relative to the baseline for 2050 if there case, because there is no need to upgrade ports if sea were no climate change. Paddy rice production may level and storm surges do not change. However, for fall by 5.8 (MoNRE) to 9.1 (Dry) million metric tons agriculture and other sectors it is difficult to identify (mmt) per year. measures that are only justi�ed under a speci�c set of climate conditions. For these sectors, adaptation Note that these �gures are not forecasts of what is often a matter of doing things that would in any will actually happen. Farming involves a continuous ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 75 FIGURE 28 CHANGES IN RUNOFF FOR THE RED RIVER AND MEKONG DELTA BY CLIMATE SCENARIO Red River Basin - 2050 Mekong Delta Inflow Changes 2050 200 60 150 40 IPSL 20 IPSL 100 GISS 0 GISS 50 MONRE Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec HadCM3 -20 0 -40 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec -60 Source: World Bank 2010i. TABLE 15 CHANGE IN CROP PRODUCTION IN 2050 DUE TO CLIMATE CHANGE WITH NO ADAPTATION (million metric tons) Climate scenario Paddy rice Maize Cassava Sugar cane Coffee Vegetables Impact Yields Sea level Total Yields Yields Yields Yields Yields Dry -6.7 -2.4 -9.1 -1.1 -1.9 -3.7 -0.4 -1.7 Wet -5.8 -2.5 -8.4 -1.0 -2.6 -2.9 -0.4 -3.1 MoNRE -3.4 -2.4 -5.8 -0.3 -0.6 -1.4 -0.1 -0.9 Source: World Bank 2010i. process of adaptation to weather, technology, eco- Further, this assessment of the potential impact of nomic and other influences, so adaptation will cer- climate change on crop production needs to be inter- tainly occur to a greater or lesser extent in practice. preted in a larger context. Changes in diets and con- Rather, these projections provide a starting point— sumer preferences with falling demand for rice, market based on the best available information and subject liberalization, trade (which will expose Vietnam to to substantial uncertainty—for (a) understanding lower-cost competition), and conversion opportun- the potential importance of climate change for crop ities to aquaculture and more salt-tolerant varieties production, holding other factors constant; and (b) will all have important effects on the demand for and assessing the type and scale of adaptation that may be the supply of agricultural products over the coming required, which will require a combination of autono- decades. The impacts of climate change have to be mous adaptation (by farmers) and planned adaptation assessed against a background of wider economic and (as a consequence of government policy). social development. 76 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT Macroeconomic impacts. As in other country studies, The impact on household incomes is skewed, with a CGE model has been used to examine the mac- greater losses for those in the bottom rural quintile roeconomic impacts of climate change. In Vietnam (the poorest 20 percent of rural households arranged the GCE model was only used to examine the effects by expenditure per person) than for the top quintile. of climate change and adaptation for the agricultural Poor rural and urban households are most vulnerable sector, so it does not attempt to take account of all of because they rely more heavily on the agricultural the macroeconomic impacts of climate change. The sector for their incomes and they spend a higher CGE model establishes a baseline composition of proportion of their income on food, which becomes economic activity up to 2050 given data and assump- relatively more expensive. tions about inter-industry linkages for 158 sectors, including regional crop production for the six crops Adaptation in agriculture. The study examined a range examined above, consumption for 10 rural/urban of adaptation options that combine autonomous household groups, population, investment, and pro- adaptations undertaken by farmers with planned ductivity growth. This is used to simulate the effect of adaptation underpinned by government spending exogenous “shocks;� that is, deviations from the base- in areas that will enhance the capacity of farmers to line scenario such as a reduction in crop production adapt. The autonomous adaptations include changes due to climate change. The model is run assuming that in sowing dates, switching to drought-tolerant crops, the aggregate level of investment and savings remains adoption of salinity-tolerant varieties of rice, adop- constant in real terms, so that aggregate consump- tion of new varieties for other crops, and switching tion moves with national income (GDP). The model to rice-�sh rotations. The planned adaptations focus takes account of the effects of exogenous shocks on on (a) increased spending on research, development industry and services, international trade, commodity and extension, with the goal of raising average crop prices, and the distribution of consumption. A broad yields by 13.5 percent relative to the baseline; and (b) picture of its results may be obtained by examining extending the area of irrigated land by about 688,000 changes in total GDP, aggregate consumption, and ha, roughly half for rice and the remainder mainly other variables under the alternative climate scenarios for maize and coffee. The total cost of these measures in 2050 relative to a baseline with no climate change. is estimated at about $160 million per year at 2005 prices without discounting over the period 2010–50. As shown in columns 1–3 of Table 16, total GDP and aggregate consumption in 2050 with no adap- Deviations in GDP and other macroeconomic tation will be 2.4 and 2.3 percent lower than the variables from the baseline with adaptation for the baseline under the Dry/Wet scenarios, respectively, alternative climate scenarios are shown in columns but only 0.7 percent lower under the MoNRE sce- 4–6 of Table 16, while columns 7–9 give the net ben- nario. The reason for the reduction in GDP is the e�ts of adaptation after allowing for the costs that decline in agricultural value-added of 13.9/13.5 are incurred. The adaptation measures substantially percent under the Dry/Wet scenarios, which is reduce the impact of climate change under all sce- marginally off set by small increases in value-added narios. The expenditures on adaptation for agriculture in industry and services. There are signi�cant dif- are clearly justi�ed as the ratio of their bene�ts to the ferences between the impact of climate change on costs that are incurred is much greater than 1. The different regions, as illustrated by the estimates for combination of the MoNRE scenario with adapta- changes in regional GDP for the North-Central tion leads to an increase in aggregate consumption, Coast and South-East regions. The gain in the indicating that some, perhaps many, of the adaptation South-East is a consequence of the concentration measures are “no regrets� options that would be justi- of industry and services in the region. �ed even without climate change. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 77 TABLE 16 MACROECONOMIC EFFECTS OF CLIMATE CHANGE WITHOUT/WITH ADAPTATION IN 2050 (Percentage deviations from baseline with no climate change) No adaptation With adaptation Adaptation bene�ts Dry Wet MoNRE Dry Wet MoNRE Dry Wet MoNRE (1) (2) (3) (4) (5) (6) (7) (8) (9) GDP -2.4% -2.3% -0.7% -1.1% -0.7% 0.7% 1.3% 1.6% 1.3% Aggregate consumption -2.5% -2.5% -0.7% -1.4% -0.8% 0.6% 1.1% 1.7% 1.3% Agricultural value-added -13.9% -13.5% -5.8% -3.8% -3.4% 5.4% 10.0% 10.1% 11.2% Regional GDP North Central Coast -6.6% -6.1% -2.6% 0.5% -0.3% 4.8% 7.1% 5.8% 7.4% South East 1.1% 0.8% 1.0% 0.0% 1.1% 0.2% -1.1% 0.3% -0.9% Rural household consumption Bottom quintile -6.5% -6.3% -2.6% -1.9% -1.4% 2.4% 4.7% 4.9% 5.0% Top quintile -1.6% -1.7% -0.4% -1.5% -1.0% 0.0% 0.1% 0.7% 0.4% Source: World Bank 2010i. An important aspect of adaptation is that it offsets is expected to be about $2.8 billion in 2010. Higher most of the disproportionate impact of climate temperatures, increased frequency of storms, sea level change on poorer households. The bottom quintile rise, and other effects of climate change are likely to of rural households bene�t most from adaptation, affect �sh physiology and ecology as well as the opera- and the gap between the changes in household tion of aquaculture. Some �sh species, such as cat�sh, consumption for the bottom and top quintiles is may grow more rapidly with higher temperatures, but almost eliminated. Adaptation partly or wholly off- be more vulnerable to disease. The main impacts of sets both the reduction in agricultural incomes and climate change on aquaculture seem likely to be a con- the increase in food prices that accompany climate sequence of increased flooding and salinity. change without adaptation. Parts of the aquaculture sector, particularly cat�sh Investments in flood and coastal protection were farming, face uncertain economic prospects, particu- not incorporated in the macroeconomic analysis. larly as a result of rising prices for feedstuffs and the Separate studies have indicated that the costs of costs of maintaining water quality. Without adapta- building/upgrading sea dikes and flood defenses tion, it is likely that climate change will signi�cantly to protect urban infrastructure and the most valu- reduce the margins that can be earned from the busi- able agricultural land would be about 1 percent of ness, so that only the most efficient farmers who adopt total investment—about $540 million per year at best farming practices will survive. Successful adapta- 2005 prices. tion will require a combination of better feed conver- sion and improvements in marketing, together with Aquaculture. Aquaculture, especially in the Mekong investments in upgrading dikes to reduce flooding River Delta, is an important source of employment and salinity intrusion that will bene�t other sectors and rural income. It is estimated that some 2.8 million as well as aquaculture. Semi-intensive and intensive people are employed in the sector, while export revenue shrimp producers may incur additional costs of water 78 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT pumping to maintain water and salinity levels. Since level rise and storm surges. Impacts include acceler- the industry is both capital-intensive and growing ated depreciation of structures and flooding of port rapidly, adaptation is likely to be autonomous with facilities such as warehouses. the costs borne by operators. The total cost of adapta- tion is estimated at an average of $130 million per Adaptation options examined in the study include (a) year from 2010–50, which is equivalent to 2.4 percent raising quay walls, (b) improving surface drainage to of total costs. reduce flooding, and (c) increasing expenditure on the maintenance and replacement of port infrastructure. Forestry. The impact of climate change on forests The cost of adaptation for all ports would be less than is likely to be complex and long term. For natu- $500 million, or about $12 million per year without ral forests, the analysis suggests that there will be a discounting at 2005 prices. substantial reduction in the area of land that is suit- able for humid semi-deciduous forest, which would Social analysis. To this point, government policies replaced by other forest types. Mangrove forests will have focused on sector-wide assessments for the be affected by sea level rise unless they are able to whole country and on “hard� adaptation measures— migrate inland. The area of land under plantation for- sea dikes, reinforced infrastructure, and durable ests with short rotations has increased rapidly over buildings. Little attention has been paid to “soft� the past 20 years. A forestry growth model suggests adaptation measures, like increasing institutional that climate change will increase the variability of capacity or the role of collective action and social plantation yields across the country without having a capital in building resilience. Most adaptation major impact on the average yield. Thus, an important options identi�ed at the �eld sites and during par- aspect of adaptation will be to ensure the best match ticipatory scenario development workshops were between soil, climate, and management practices to aimed at improving response capacity and disaster obtain the highest yields from plantations. risk reduction—forecasting, weather monitoring— and managing climate risk. Notably, adaptation A range of adaptation options was considered. The options that reduce poverty and increase household key measures identi�ed were (a) changes in land use resilience or that integrate climate change into planning to facilitate the migration of mangroves, development planning were not emphasized. (b) adoption of plantation species and methods of silviculture that are more resilient to droughts, (c) Overall, many of the adaptation options observed improvements in pest management, including genetic at the field sites and/or proposed in workshops selection and integrated pest control strategies, and were highly cost-effective and do not require large (d) use of herbicides or biological controls to limit expenditures. Moreover, they were largely in line the effect of exotic weed species on tree growth. The with the adaptation options considered for the cli- money costs of adaptation are likely to be modest, but mate scenarios in the sector analyses. These adap- the institutional issues may be more difficult to deal tation measures included shifting planting dates, with. adopting drought-tolerant crops, and switching to salinity-tolerant varieties of rice. The diversity Coastal ports. Along its 3,200km coastline, Vietnam of preferred adaptation responses reflected the has a total of 116 ports. In addition, new terminals are impressive variety of Vietnam’s vulnerability zones being constructed and planned all along the coastline, and confirm the need for a mix of both autono- particularly in the south around Ho Chi Minh City mous and planned adaptation, a mix of hard and and in the north around Hai Phong. Given the nature soft options, and adaptation to be carried out at of its location, this infrastructure is at risk from sea- the national, subnational, and community levels. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 79 Lessons and recommendations infrastructure and improvements in operations and maintenance. Climate change will have a signi�cant impact on some regions and sectors of Vietnam’s rural economy. Increased spending on the maintenance and Still, in macroeconomic terms the impacts of climate extension of coastal and flood defenses to mini- change on agriculture and related sectors, even with mize the impacts of sea inundations, salinity no adaptation, appear to be relatively modest. In prac- intrusion, and river flooding, especially in the tice, there will be substantial autonomous adaptation Mekong River and Red River deltas. even without active government intervention, since farmers will change the crops and crop varieties they Much of these expenditures would be justi�ed even grow and their methods of farming. without climate change, so adaptation to climate change is primarily a matter of building upon no-regrets mea- The major concern is the extent to which climate sures. Under the intermediate MoNRE climate sce- change will hit poor households, partly because of the nario, the program of agricultural adaptation outlined in decline in agricultural incomes and partly because of this study will increase agricultural incomes relative to an increase in food prices relative to the general cost the baseline, especially in the Central Highlands region, of living. The lowest 20 percent of households—either illustrating the general bene�ts of the strategy. urban or rural—arranged by household expenditure per person will experience larger reductions in real If this program of adaptation were to be imple- standards of living due to climate change than the mented, the adverse impacts of climate change on top 20 percent of households. poorer households would largely be avoided. There would still be a net loss of agricultural value-added Thus, the primary focus of policies to adapt to climate and aggregate consumption in the Wet and Dry cli- change should be to protect the poor, the vulnerable, mate scenarios, but the magnitude of the losses would and those least able to respond to changing climatic be signi�cantly smaller and the skewed impact on the stresses. The goal should be to provide farmers and distribution of income would be corrected. others with the tools and resources that will enable them to respond to climate change itself and to the Year-to-year weather variability is much greater than new risks that will accompany climate change. The the long-term trends associated with climate change. key elements will be: Policies and systems that can cope effectively with weather variability will be more successful in adapt- Increased expenditures on research, development, ing to future climate change than those that cannot. and extension for crop production, aquaculture, Strengthening the capacity of the rural sector to cope and forestry to develop new crop varieties that are with current weather variability and build resilience more tolerant to drought, salinity, and higher tem- into such systems will yield bene�ts both now and in peratures early in the growing season. Both the the future. It is also important to collect, analyze, and public and the private sectors should be involved report data on how the climate is changing in dif- in efforts to increase yields and productivity. ferent regions of the country, so that those who have to take account of climate change in planning new Investment in expanding irrigation infrastruc- infrastructure or implementing investment programs ture, especially in the central regions where the should have access to the best possible information. opportunities for irrigation expansion are great- est. In the short term, this should build upon Climate change, including sea level rise, will affect the achieving fuller utilization of existing irrigation country’s infrastructure and require expenditures on 80 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT adaptation. The case study of coastal ports indicated storm activity (IWTC 2006). Still, many climate sci- the lesson that the costs of adaptation are likely to be entists believe that climate change will lead to some modest. The total cost of protecting existing ports that increase in the intensity of tropical cyclones—an are exposed to flooding—as a result of higher sea level increase of 3–5 percent in peak wind speed per 1°C combined with greater storm surges—is estimated rise in sea surface temperature—accompanied by at no more than $500 million in total over 40 years, greater variability of rainfall with more frequent epi- equivalent to $12.5 million per year (at 2005 prices sodes of very heavy rainfall and drought. without discounting), or about 1 percent of planned investment in ports over the period 2010–30. Approximately 70 percent of the population of Samoa lives in low-lying areas, which would be An equally important lesson from the case study is vulnerable to inundation as a consequence of the that it is essential to plan ahead for climate change. combined effects of sea level rise, more severe storm Ports that are built over the next 10–20 years should surges, and flooding caused by heavier rainfall. As an be designed to cope with sea levels and storms to illustration of the risks, two major cyclones (Ofa and which they may be exposed 50 or more years from Val) hit or passed near to one of the two main islands now. It may be cheaper to build margins of resilience in 1990–91, damaging a majority of buildings and and safety into new infrastructure than to upgrade causing a total economic loss including asset damage assets during the course of their life. The same lesson and the capitalized value of lost GDP of about $550 emerges from the countrywide analyses for infra- million at 2005 prices, equivalent to about 3.75 times structure and coastal protection undertaken as part of GDP in 1990. While these events were considered to the EACC Global study. The total cost of adaptation be unprecedented within the previous 100 years, an for these sectors amounts to about 2 percent of total increase in the probability of such large losses from investment for the Global Wet (NCAR) scenario and 1-in-100 years to 1-in-50 years or even 1-in-25 years about 1.3 percent of total investment for the Global would clearly be very signi�cant. Dry (CSIRO) scenario on the assumption that adap- tation measures are combined with new investments EACC approach and results anticipating climate change up to 2100. The EACC study focuses on the impact of, and 45 Samoa adaptation to, a shift in the probability distribution of tropical storms affecting the islands. The severity Vulnerability to climate change of such storms is measured by their peak wind speed over a period of 10 minutes. Wind speed is associated Samoa is a country at extreme risk from a variety of directly with the amount of wind damage caused by natural disasters, including tropical cyclones and tsu- a storm. Further, it serves as a proxy for the intensity namis caused by earthquakes. In addition, it is subject of precipitation and the height of storm surges, which to inter-annual climate fluctuations associated with are associated with flood damage in coastal and non- El Niño (ENSO), which affect precipitation as well coastal zones. Since storms that hit the islands and as air and sea temperatures. Periods of drought in the cause signi�cant damage are infrequent events, the islands have been linked to the ENSO. There is no analysis examines how climate change will affect the simple association between increases in mean surface expected annual value of storm damage expressed as temperature and the frequency of tropical cyclones, a percentage of GDP under the alternative climate partly because of the strong influence of ENSO cycles scenarios. The extent of such damage depends upon on tropical storms in the Paci�c and partly because a combination of (a) the resilience to storm dam- climate models have difficulty in simulating tropical age that is designed into buildings and other assets, ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 81 and (b) other measures to reduce the vulnerability of and April, leading to the likelihood that the length communities to flooding and wind damage. of the main cyclone season will increase. Because the impact of climate change on the intensity of the The study divided Samoa into four economic regions worst cyclones is uncertain, the analysis examined the (Figure 29). North Upolu has a population of about effect of an increase in the peak wind speed for a 1 in 110,000, while the populations of the other regions 100 year storm in a range from 4 percent (low) to 8 fall between 17,000 and 28,000. Table 17 shows base- percent (high) for 2050, and from 10 percent (low) to line values and changes in precipitation over differ- 25 percent (high) for 2100. In addition, the extent of ent periods—the whole year, the rainy season from potential damage caused by flooding at other times November to April, and the main cyclone season from was linked to changes in the amount of precipitation December to February—and mean temperatures by during the rainy season, using the NCAR climate region derived from the Global Wet (NCAR) and scenario as the low scenario and the CSIRO climate Global Dry (CSIRO) scenarios. The rise in mean scenario as the high scenario. temperature is consistent across regions and falls in the range of 0.8 to 1°C for the two climate scenarios. The data available for Samoa cannot sustain a con- However, total precipitation declines marginally in ventional CGE model, so a simple macro model of three out of four regions under the NCAR scenario, climate and economic growth has been used to exam- but increases signi�cantly in the CSIRO scenario. ine the effects of climate change on the economy. To For all regions and both scenarios, precipitation tends maintain the baseline level of economic growth, this to increase during the months of November, March, assumes that changes in the expected value of damage FIGURE 29 REGIONS OF SAMOA USED IN THE CLIMATE-ECONOMY MODEL Source: World Bank 2010h. 82 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT TABLE 17 DEVIATIONS IN PRECIPITATION AND TEMPERATURE IN 2050 BY CLIMATE SCENARIO Scenario Region Baseline values for NoCC Deviations in 2050 relative to NoCC Total Precipit. Precipit. Mean Total Precipit. Precipit. Mean precipit. Dec-Feb Nov-Apr temperat. precipit. Dec-Feb Nov-Apr temperat. mm mm mm °C mm Mm mm °C NCAR Savai’I N 2,958 1,062 1,921 26.87 -17 -39 5 0.99 NCAR Savai’i S. 3,002 1,107 1,971 26.86 -19 -41 3 0.99 NCAR Upolu N. 3,048 1,154 2,024 26.83 -21 -42 0 0.99 NCAR Upolu S. 2,929 1,090 1,942 26.67 106 -8 118 0.97 CSIRO Savai’i N 2,958 1,062 1,921 26.87 277 43 197 0.81 CSIRO Savai’i S. 3,002 1,107 1,971 26.86 343 65 215 0.83 CSIRO Upolu N. 3,048 1,154 2,024 26.83 344 68 218 0.83 CSIRO Upolu S. 2,929 1,090 1,942 26.67 335 66 213 0.83 Note: NoCC=no climate change. Source: World Bank 2010h. caused by storms fall on total consumption, so that the 1990s—are more important sources of risk for the economic impact of climate change is measured agricultural incomes than climate change. Managing by the changes in the present value (discounted at 5 non-climate risks better will also reduce the potential percent) of consumption over the period 2010–50 impact of future climate variability on agricultural relative to the no climate change (NoCC) baseline. incomes. Impacts. The gross economic losses under two sce- Adaptation. The key form of adaptation is the imple- narios of low and high impact when there is climate mentation of design standards to ensure that buildings change without adaptation are shown in Columns and other assets can cope with higher winds and more (1) and (2) of Table 18. The impact of climate change intense precipitation without damage. The effective- under the high-impact scenario amounts to $212 ness of this approach can be illustrated by analysis of million at 2005 prices in present value terms. On an the damage caused by Cyclone Heta in 2004. This was annualized basis, this is equivalent to $12.1 million at approximately a 1-in-11 year event with a peak wind 2005 prices per year, or 1.3 percent of total GDP in speed of 110 kph, but it caused very limited economic the baseline scenario. damage. Had design standards in force in 2004 been similar to those in 1990–91, when Cyclones Ofa and There is one sphere in which the impacts of climate Val hit the country, the economic loss would have been change may be smaller than is often assumed. Agri- much higher at 35–40 percent of GDP. The reduction culture provides the main source of employment for in potential damage was a consequence of changes in about one-third of the labor force, but it represents design standards and other measures that increased a small and declining share of GDP—about 6 per- the effective threshold for storm damage from 1-in-5 cent for agriculture in 2008. Variations in the ENSO year events (a peak wind speed of 90 kph) to 1-in-10 cycle have had a statistically signi�cant impact on year events (a peak wind speed of 108 kph). The esti- taro production and agricultural imports. Even so, mates of net economic losses with adaptation in Table non-climate factors—for example, the taro blight in 18 assume that the current standard of a 1-in-10 year ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 83 TABLE 18 LOSSES DUE TO CLIMATE CHANGE WITHOUT AND WITH ADAPTATION Gross losses without Net losses with Net bene�ts of adaptation adaptation adaptation NCAR CSIRO NCAR CSIRO NCAR CSIRO (1) (2) (3) (4) (5) (6) Present value @ 5%, $ million 103.9 212.4 4.5 5.4 99.4 207.0 Annualized equivalent, $ million per year 5.9 12.1 0.3 0.3 5.6 11.8 Loss/bene�t as % of baseline GDP 0.6 1.3 0.0 0.0 0.6 1.3 Source: World Bank 2010h. storm is applied in estimating the costs of adaptation and the residual damage after adaptation. However, analysis carried out for the EACC global study indicates that the bene�ts of strengthening buildings and infrastructure to withstand the wind and precipitation associated with a 1-in-50 year storm (peak wind speed ~ 148 kph) would exceed the costs involved by a large margin even under an assumption of no climate change. The present value of adopting the stricter design standard as a consequence of a lower expected value of storm damage would exceed $275 million after allowing for the additional costs of con- struction. This would be a “no regrets� form of adap- tation, because it would greatly reduce the economic impact of climate change under any climate scenario. With a 1-in-50 year standard, the present value of gross losses from climate change would fall from $212 million to $37 million for the high climate scenario. Further spending on adaptation to protect against 1-in-50 year storms up to 2100 under the high sce- nario (peak wind speed ~ 184 kph) would also pass a cost-bene�t test. The estimates of net economic losses with adaptation in Table 18 assume that a standard of a 1-in-50 year storm is applied in estimating the costs of adaptation and the residual damage after adaptation. For agriculture the key element of an adaptation strategy is to increase expenditures on research, development, and advisory services so as to mitigate the higher risks that are likely to be associated with 84 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT climate change. Again, this is building upon policies that would form part of a sound development strategy. Samoan agriculture was hit hard by the taro blight in the mid-1990s, which devastated taro production and eliminated export revenues from taro for nearly 15 years. Vulnerability to disease, pests, and storm dam- age means that diversi�cation of both crop varieties and crops is an important element in any policy to limit the impact of these risks on agricultural house- holds. Since climate change is likely to reinforce these risks, the appropriate level of expenditure will be higher to reflect the greater value placed on risk reduction. Other adaptation measures, largely drawn from NAPA, were considered in each of the regions. It is assumed that adaptation measures are only imple- mented in a particular region when or if the resulting reduction in the expected value of economic losses due to climate change exceeds the annualized cost of the adaptation measures. This is a simple cost-bene�t test designed to optimize the timing of expenditures on adaptation. Coastal zone infrastructure. These include the con- struction of sea dikes to protect infrastructure along vulnerable parts of the coast or the reloca- tion of key assets such as roads or schools out of potential flood zones. Measures to encourage vil- lages to relocate away from flood zones entirely, such as extending the national power grid and building new roads, were also included. This approach is being applied in parts of the south coast of Upolu in response to the 2009 tsunami. The largest costs are associated with the reloca- tion of utility infrastructure (power, roads, water reticulation, water treatment and telecommunica- tions) for a village—estimated at $32 million— but the initial investment is expected to be partly offset by greater income from tourism, planta- tions, and other activities. Water supply. Ensuring better access to good qual- ity water for communities was the main priority ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 85 in the NAPA. This is an example of the overlap the period 2030–35 in Upolu North, but not before between development priorities and adaptation 2050 in Upolu South. Again, several of the adaptation to climate change. There is little doubt that mea- measures —for example, moving coastal infrastructure sures such as decreasing leaks in the reticulated in Upolu South and improving water supply—would water supply, improvements in catchment man- be justi�ed for non-climate reasons. agement, and better water treatment at the source are justi�ed even without climate change. Table 19 shows the composition of adaptation costs by category and decade. Most of the cost of adapta- Tourism. Adaptation in tourism focuses on the tion for housing will fall upon households. This may provision of niche tourist facilities, including be regarded as a form of autonomous adaptation, so inland (rain-forest) resorts that are away from the table shows the total cost of adaptation including the coast. and excluding housing. The main costs are incurred for housing, municipal infrastructure—which covers Food security. This includes improvements in the public buildings and storm water drainage—and agri- operation of existing plantations, the promotion culture and �sheries. When expressed as a percentage of village based micro-enterprises, research into of the cost of providing the relevant services in the crop changes, and sustainable �shing. baseline scenario, the average increases vary from 1 percent of baseline costs for electricity and telecom to Urban development. This is particularly important 6 percent for housing. As might be expected, the bur- in Upolu North, where a better approach to plan- den of adaptation rises over time as the probability ning land use and urban development is needed. distribution of severe storms shifts. Over the whole In view of the vulnerability of urban infrastruc- period from 2010 to 2050, the increase is 3.5 percent ture to storm damage, adaptation must focus on for infrastructure excluding housing, and 4.1 percent enhancing the resilience of the key commercial including housing. and economic zones to extreme weather shocks. Local-level perspectives on adaptation. The EACC Columns 3 and 4 of Table 18 show the net economic study did not carry out a social assessment of cli- losses due to climate change with adaptation, allow- mate change in Samoa. Nonetheless, the country has ing for the cost of implementing the adaptation started to address the potential impacts of a greater measures. The net bene�ts of adaptation are shown frequency and intensity of cyclones through a com- in columns 5 and 6. They amount to $99 million at bination of stronger institutions, better governance, 2005 prices for the low scenario and $207 million for and robust planning. This will underpin a variety of the high scenario. soft adaptation actions, such as re-orienting coastal infrastructure management and developing commu- Adaptation measures involving the adoption of more nity disaster plans. stringent design standards are clearly justi�ed, even in the low scenario for which the gross losses due to cli- Samoa’s cultural context is an important factor when mate change are relatively small. However, under the selecting adaptation measures. The traditional model low scenario the other adaptation measures outlined of community decision making is by consensus under above do not reduce climate losses by a sufficient mar- the leadership of the matai (chief ). The authority of gin to cover their costs, though they may be warranted a village matai and customary land ownership rights for other reasons. Under the high scenario, these adap- are respected, so negotiations between the govern- tation measures should be implemented in the period ment and village matai can often take a long time. 2025–29 in both Savai’i North & Savai’i South and in There is a commitment to supporting village-based 86 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT TABLE 19 AVERAGE ADAPTATION COSTS BY DECADE, HIGH SCENARIO, 2005 $ MILLION/YEAR, NO DISCOUNTING 2010–19 2020–29 2030–39 2040–49 Agriculture & �sheries 0.99 1.47 2.03 2.66 Coastal protection 0.03 0.03 0.03 0.03 Education & health 0.13 0.18 0.30 0.44 Electricity & telecom 0.08 0.11 0.17 0.20 Housing 0.55 1.07 1.93 3.16 Municipal 0.86 1.39 2.22 2.98 Other transport 0.08 0.13 0.22 0.36 Roads 0.52 0.68 0.82 0.93 Water & sewer 0.03 0.04 0.07 0.10 Total 3.26 5.11 7.78 10.86 Total excl housing 2.71 4.04 5.85 7.70 Source: World Bank 2010h. consultations that include women and youth. Raising related development issues. Over the last decade it awareness of climate change and other development has focused on increasing the capacity of its institu- concerns through village-based consultation is an tions, which are necessary for the implementation effective and sustainable way of supporting the tradi- of soft approaches to adaptation, including land-use tional decision-making model. Nevertheless, women controls and coastal infrastructure management. and migrants in the poorer communities remain among the most vulnerable groups in the community. One key lesson is that extreme weather variability Stakeholders at workshops held during the prepara- in the coastal zone will involve signi�cant costs tion of the NAPA identi�ed the following areas as for either investments in coastal protection or the critical to a strategy for adapting to climate change: relocation of assets. In the longer term, the relo- the protection of community water supplies, early cation of assets—or, even whole villages—may be warning systems, support for agriculture and for- the best option as it can shift economic activity estry sectors, implementation of coastal infrastruc- such as tourism, crops, and other businesses away ture management plans, and integrated catchment from the coast. management. The uncertainty about climate outcomes and lack Lessons and Recommendations of baseline data has led to a focus on the col- lection of information in Samoa. More effort is Samoa is a small island nation with most of its popu- needed to support the collection and analysis of lation and infrastructure located along the coast, so it this information and use of the information to is highly vulnerable to extreme weather events. How- inform decision making. ever, Samoa is also among the more climate-resilient Paci�c Island countries, and there is much to learn Good development policies are a foundation from the way it is approaching climate change and for climate change adaptation. The participatory ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 87 consultations undertaken across the country in without climate change, but will yield even larger developing plans for managing coastal infrastruc- bene�ts if climate change leads to more intense ture are continuing with a focus on other develop- storms in future. ment and adaptation issues. Going beyond “no regrets� adaptation, the key The analysis suggests that the country should measure identi�ed in the study is the adoption of consider, as a good development policy even in design standards that will enable buildings and the absence of climate change, the adoption of other assets to cope with storms with higher peak design standards that would enable buildings and wind speeds and associated precipitation look- infrastructure assets to cope with 1-in-50 year ing forward to 2100 under alternative climate storms under historical climate conditions with- scenarios. Retaining the 1-in-50 year criterion out signi�cant damage. This would imply that would mean buildings and infrastructure should buildings and infrastructure should be designed be designed to withstand storms with a peak wind to withstand storms with a peak wind speed of speed up to 184 kph under the worst climate sce- up to 148 kph. Such a policy is a “no regrets� form nario. Under this strategy, the expected losses of adaptation whose bene�ts exceed its costs even from climate change would be greatly reduced. 88 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT Five ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 89 Lessons Extracting robust general conclusions across diverse across regions as well as decades. Our estimates of countries with respect to an uncertain and broad the overall cost of adaptation are 0.6–0.7 percent of phenomenon like climate change is a perilous task. GDP for the Sub-Saharan Africa region in 2010–19, Too much generality leads to banal and potentially falling to about 0.5 percent of GDP in 2040––49. uninformative conclusions. Excessive speci�city is In contrast, the equivalent �gures for the East Asia unhelpful as a basis for useful generalizations. We have and Paci�c region are 0.13–0.19 percent in 2010–19 attempted to strike an appropriate balance. The recom- and about 0.07 percent in 2040–49. Apart from Sub- mendations follow from the lessons of the country and Saharan Africa, the regions facing high relative costs global exercises. of adaptation are the Latin America and Caribbean region and (under the dry climate scenario) the South The costs of adapting to climate change Asia region. Past and prospective future emissions mean that some The absolute costs of adaptation increase over time amount of climate change is inevitable over the next cen- and will certainly continue to increase after 2050. Our tury, even though the extent and nature of the changes projections suggest that real GDP will increase more are uncertain. Adapting to a climate that is about 2°C rapidly than the costs of adaptation during the next warmer will be costly, but our country studies show that four decades, even on quite conservative assumptions the impacts of climate change without adaptation will about growth in GDP per person. However, it would be much more costly. The study puts the cost of adapting be unsafe to assume that this trend will continue into to climate change at an average of $70 billion to $100 the second half of the current century. billion a year at 2005 prices between 2010 and 2050. The cost amounts to 0.2 percent of projected GDP for all LESSON 1: developing countries in the current decade and falls to The cost of developing countries to adapt to climate about 0.12 percent of projected GDP for 2040–49. This change between 2010 and 2050 is estimated at $70 cost is large when compared to current levels of devel- billion to $100 billion a year at 2005 prices. This opment aid, as $100 billion is 80 percent of the total amounts to about “only� 0.2 percent of the projected disbursement of ODA in 2008. GDP of all developing countries in the current decade, and at the same time to as much as 80 percent of total disbursement of ODA in 2008. The averages across all developing countries hide a very uneven distribution of the burden of adaptation 90 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT In a few of the country studies—Bangladesh, Samoa vulnerability to climate, both in aggregate and for and Vietnam—it was possible to make limited com- the poor when their interests are built into devel- parisons between adaptation costs estimated in the opment strategies. global analysis and in the country study. Where like could be compared with like, the separate estimates Economic development generates both the were in reasonable agreement. However, the analysis resources and opportunities to adapt to climate suggested that the costs of strengthening infrastruc- change at a relatively low cost by ensuring that ture to provide resilience against the wind damage, the design and location of new infrastructure, precipitation and flooding caused by infrequent buildings, and other assets take account of the but severe tropical storms could increase the overall effects of climate change on their performance. cost of adaptation by 10-20% under the worst sce- nario relative to the global estimates for countries Our country studies show that a failure to adapt that are most exposed to such storms. This cannot to climate change may lead to very large weather- be directly linked to any particular climate scenario, related losses, both in terms of the destruction since the worst storm scenario depends upon (a) how of infrastructure and foregone opportunities for ocean temperatures respond to climate change, and future growth. In Ethiopia, for example, robust (b) changes in the tail of the distribution of extreme growth based on infrastructure investment is the weather events. This adjustment does not alter the �rst line of defense against climate change impacts. overall range of adaptation costs, since the countries At the same time the design of new infrastructure affected account while the range is based on estimates must take account of future weather stresses and with a strong upward bias. existing infrastructure must be upgraded and/or replaced where it cannot cope with such stresses. On the other hand, the analysis in the Ethiopia country study suggests that cross-sectoral effects of Bangladesh, Ghana, Mozambique and Vietnam rely adaptation measures (not considered under the global upon large rivers that are affected by the investment track) could increase the overall costs of adaptation. and operational decisions made by upstream riparian It is hard to generalize from the one country case to countries. Such decisions determine the availability estimate the extent of bias, but it is important that of water resources for irrigation, hydro-power and future research explore this bias which reflects the other uses as well as vulnerability to flooding. For opportunity cost or forgone bene�ts from allocating example, Vietnam’s capacity to respond to climate resources to adaptation measures instead of allocating change will, in part, depend upon upstream develop- them to other development initiatives. ments in the Mekong River basin, but this will also affect investments in and the management of irriga- Economic development and adaption to tion and aquaculture even without climate change. climate change More extensive and effective cooperation between the countries that share international river basins The link between economic development and adapta- will be crucial both for economic development and tion to climate change is fundamental. for adaptation to climate change. Economic development is the most basic and cost- The existing management of water resources and irri- effective method of adaptation, provided that it is gation systems was highlighted as a key issue for both properly managed. Richer countries are more resil- development and adaptation in the country studies ient to weather variability. Economic development for Bolivia, Samoa and Vietnam. In these studies, brings changes in economic activities that reduce access to and the availability of water for public water ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 91 supply and irrigation systems were identi�ed as key Climate change exacerbates this pattern. For example, concerns in the social vulnerability studies and for the rural poor in the southern region of Bangladesh are the economic impact of climate change. These studies expected to face the largest declines in per capita con- identi�ed a combination of more investment in and sumption, as well as declining productivity of subsistence better management of existing water infrastructure as crops and land losses due to increased salinity brought yielding immediate development bene�ts as well as forth by sea level rise. The Vietnam study suggested that enabling communities to adapt more easily to a range the impact of climate change falls disproportionately on of future climate scenarios. households in the lowest quintiles of the rural and urban income distributions. Adaptation through agricultural Another example of the overlap between development improvement and expansion of irrigation largely offset and adaptation concerns autonomous adaptation in this impact and may reduce inequality relative to the agriculture in countries such as Bolivia, Ethiopia, outcome with no climate change. Mozambique and Vietnam. The dissemination and adoption of improved agricultural practices including Rapid urban growth will be a key aspect of economic drought-resistant crop varieties, fertiliser use and bet- development in countries such as Bolivia, Ethiopia, ter water management will be critical for maintaining Ghana and Mozambique. Some of the country stud- agricultural production and incomes under different ies examined investment in human capital as part of climate scenarios. But, equally, enhancing the capacity strategies to adapt to climate change. Under most cli- of small farmers to take advantage of more produc- mate scenarios, expenditure on equipping rural pop- tive agricultural technologies is crucial for economic ulations and urban migrants with better education development under any scenario. This would include and market-driven skills is a cost-effective response crop insurance and other mechanisms for redistribut- to climate change. At the same time, these country ing the risks of weather and other agricultural risks studies highlighted the risk that the growth of cit- with or without climate change. ies will expose more people to the consequences of poor urban management, such as flooding and lack of sanitation, that are likely to be exacerbated by climate LESSON 2: change unless appropriate social and physical infra- Economic development is a central element of adapta- structure is put in place. tion to climate change, but it should not be business as usual. LESSON 3: Countries that reach the middle of the 21st century Invest in human capital, develop competent and flexible institutions, focus on weather resilience with large shares of their populations engaged in and adaptive capacity, and tackle the root causes of subsistence agriculture, with substantial illiteracy, and poverty. Eliminating poverty is central to both devel- with lethargic and/or inept institutions will be par- opment and adaptation, since poverty exacerbates ticularly vulnerable to the effects of climate change. vulnerability to weather variability as well as climate Rapid development leads to a more flexible and change. resilient society, so that building human and social capital—including education, social protection, and Climate uncertainty: the need for robust health and skills training—are crucial to adaptation. strategies In all case-study countries, the burden of existing climate The fundamental problem of making public policy in variability is especially heavy in areas that have high con- the face of climate change is one of uncertainty with centrations of poor and socially vulnerable populations. regard to both climate outcomes and longer term 92 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT projections of social and economic development. possible and focus on low-regret actions, while await- Even though the uncertainties regarding the socio- ing greater certainty about climate and socioeco- economic projections are more frequently discussed nomic scenarios. Low-regret actions are those actions in the context of broader development planning, they that are robust under most climate scenarios. They should not be entirely shadowed by climate uncer- are typically policies or investments that can be identi- tainties. In Bangladesh, for example, the number of �ed as priorities for development even without climate people living in cities by 2050 is expected to triple, change. Our country studies included a number of while the rural population falls by 30 percent. Current these strategies. For example, investments to expand policies will determine where this urban population the road system and increase the share of paved roads settles and how prepared it is to adapt to a changed in Africa yield high returns by lowering transport climate. Adaptation decisions to be made now can costs and expanding markets. At the same time they prove to be signi�cantly wrong, and thus costly, if lessen the impact of floods and enhance the ability of such socioeconomic projections end up being wrong. farmers to respond to changes in agricultural com- parative advantage. Similarly, better management of In terms of climate outcomes, such uncertainty is water resources; improved access to extension ser- particularly large for patterns of precipitation. Some vices, fertilizers, and improved seed varieties; and of the country studies have highlighted crucial dif- better climate and weather forecasting will enhance ferences between alternative wet and dry scenarios the resilience of agriculture, both to droughts, and to and their effects on agricultural production, water waterlogging caused by floods. resources, and transport infrastructure. This uncer- tainty about the underlying trends in climate variables Apart from promoting these low-regret measures, is exacerbated by the expectation that the variability which include many “soft� adaptation alternatives, it is of weather indicators around climate averages will also important to subject long-lived, expensive infra- increase, making it even more difficult to reach reli- structure such as dams and other water infrastructure able conclusions on what is weather variability and to careful climate-robustness tests. In Mozambique, what are climate trends. the recommendation coming out of our study is clearly toward delaying investments in large coastal protection The general economic equilibrium analyses in Ethio- schemes, and focusing more on the people affected pia, for example, suggested that the cost of adapta- than on the land lost. The expensive option of con- tion varies by a factor of 3, depending on the climate structing dikes would be justi�ed only for vital coastal scenario considered. Thus the cost of selecting the infrastructure, such as the port of Beira. In Bangladesh, “wrong� strategy may be considerable. Under these a no-regrets strategy would be to begin by addressing circumstances, the value of reducing uncertainty the adaptation de�cit and strengthening the early about future climate outcomes is extremely high, since warning systems. Additional embankments and shel- it would help better de�ne what kinds of adaptation ters can be constructed in the medium term as the (viewed as a form of insurance) are most appropri- geographic incidence of risk becomes more certain. ate. It also follows that making investment decisions based on any one climate scenario is no more justi- Bolivia and Ethiopia are two countries with large dif- �ed than basing it on another. Attempting to hedge ferences between the effects of the extreme Wet and against most or all climate outcomes obviously may Dry climate scenarios, especially in highland areas raise the cost of adaptation very substantially. such as the Altiplano and the Rift Valley plateau. These are areas that are already prone to both intermittent It also follows from this analysis that countries droughts and intense rainfall causing severe flooding. should want to delay adaptation decisions as much as Major investments to manage water resources might ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 93 be warranted under some climate scenarios but not by intermittent flooding is a major factor determin- under others. Hence, the studies identify a variety of ing the losses associated with alternative climate more limited measures which are designed to increase scenarios. This damage can be reduced, sometimes the resilience of agriculture and economic activity by a large amount, simply by reallocating the bud- to existing weather risks and which can be justi�ed gets for road construction and maintenance to build under a wide range of possible climate outcomes. In fewer but stronger roads. With a few exceptions this the longer term, larger investments for adaptation is a no regrets strategy that will pay off under a wide may be warranted if the uncertainty about the effects range of climate scenarios. of climate change has been greatly reduced. A similar conclusion emerges from the Samoa case study, in which strengthening buildings and infra- LESSON 4: structure so that they can withstand 1 in 50 or 100 Do not rush into making long-lived investments in year storms under today’s climate would reduce the adaptation unless they are robust to a wide range of climate outcomes or until the range of uncertainty potential losses due to climate change by more than about future weather variability and climate has nar- 80% without any further adaptation. While resilience rowed. Start with low-regret options. to extreme weather events was not examined in detail in all of the country studies, the analyses for Bangla- Current climate vulnerabilities desh, Mozambique and Vietnam all indicate that the adoption of construction standards and investment Climate change will always hide beneath weather vari- strategies which are more resilient to current weather ability. Systems that can effectively cope with existing risks can greatly reduce both storm damage in the weather risks will be more successful in adapting to future near future and the costs of adaptation to climate climate change than those that cannot. The short-term change in the longer term. priority is to better prepare for the weather risks that countries are already facing. LESSON 5: One clear example concerns the impact of storms, Adaptation to climate change should start with the adoption of measures that tackle the weather risks especially in coastal areas. Despite the uncertainty over that countries already face, such as more investment future rainfall, there is relative certainty that a warmer in water storage in drought-prone basins or protec- climate will lead to rising sea levels and an increased tion against storms and flooding in coastal zones intensity of storms. With the inevitable increase in and/or urban areas. Climate change will exacerbate urban populations, the costs of failing to protect these risks. coastal cities against major storms will increase rap- idly. At the same time, the de�ciencies of storm water The prospect of greater weather variability has an drainage in coastal or inland cities has already led to additional, rather more difficult, implication. Eco- avoidable—and sometimes large—losses caused by nomic development has been accompanied by a urban flooding that have disproportionate effects on tendency for more rapid urban growth in coastal the health and welfare of the poor. areas than in inland cities. This may reflect relative differences in transport costs as well as government Many of the country studies illustrate the contri- policies or individual preferences. There will be bution that better methods of managing existing many opportunities to reduce weather risks and the weather risks can make in lowering the costs of associated costs via intelligent urban and land-use adapting to climate change. In each of the Africa planning. Whether in rural or urban areas, the rule case studies damage to roads and transport caused of thumb is simple: wherever possible, ensure that 94 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT future growth and infrastructure takes place in loca- Similarly, the country studies for Bolivia and Ethio- tions that are less exposed to weather risks. pia highlight the importance of managing agricul- tural expansion in drought-prone areas, especially in In addition to the need to subject large investments places where droughts may become more frequent or in protecting coastal infrastructure to great scrutiny severe as a consequence of climate change. This may to ensure robustness to different climate scenarios, be economic with sufficient investment in developing appropriate incentives must be put in place that drought-resistant crop varieties and water infrastruc- discourage the accumulation of physical capital in ture, but too often such expansion has been a response the shadow of dikes considered to be “safe.� As the to pressure on land resources. Once such expansion tragedy of New Orleans dramatically illustrated, a has occurred it becomes more difficult and expensive sufficiently extreme event will breach a dike. The to adapt to adverse climate outcomes. combination of an increasing severity of extreme events, the high costs of providing physical protec- A �nal example of the same mistake is permitting the tion, and the accumulation of capital behind such development of housing and urban infrastructure in barriers can mean that the expected value of losses, flood plains, which is a signi�cant problem in Ban- including human suffering, may not be reduced, gladesh, Ghana and Mozambique. either at all or by as much as expected by invest- ments in protection. LESSON 6: Similar concerns apply to efforts to maintain the Beware of creating incentives that encourage devel- welfare of populations engaged in agriculture and opment in locations exposed to severe weather risks. Where possible, build future cities out of harm’s way, other resource-intensive activities that are sensitive particularly flood plains or coastal zones that are to climate variability and change. Short-term mea- exposed to sea level rise and storm surges. sures to prevent suffering must be complemented by long-term measures such as education, job train- ing, and resettlement designed to reduce reliance Hard vs. soft approaches to adaptation on resources and assets whose value may be eroded by climate change. Adaptation should not attempt The distinction between “hard� (capital-intensive) to resist the impact of climate change, but rather it and “soft� (institutions and policies) adaptation is should offer a path by which accommodation to its easily exaggerated. The reality is that both approaches effects can be made less disruptive and does not fall are necessary. There is no point in building the best disproportionately on the poor and the vulnerable. type of road in the wrong place, while the best insti- tutions will provide no protection against a storm The country study for Samoa highlighted the dif- that destroys buildings or power lines. Thus, the �culty of managing coastal development in a coun- challenge is to get the balance between hard and soft try with high exposure to natural hazards. Coastal adaptation right. In some �eld sites in Vietnam, re- locations are attractive and easy to develop but establishment or migration of mangroves was ranked buildings and infrastructure have been badly dam- above spending money on sea walls, given the lower aged by storm surges or tsunamis. On the other costs of mangrove planting and the potential for this hand, relocating existing infrastructure is expen- activity to be more pro-poor. sive and may be hard to justify. Either such invest- ments should be minimised or they should built/ Nonetheless, pouring concrete is often a very expen- upgraded so that they are capable of withstanding sive and relatively ineffective method of adaptation. more severe weather and other risks. The importance of keeping infrastructure and urban ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 95 development out of harm’s way is a key illustration of This is part of the larger theme that economic develop- the costs of creating perverse incentives that encour- ment is the best form of adaptation. Implementing good age behavior and investments that worsen rather than policies and developing effective institutions should reduce the prospective impacts of climate change. be pursued simply because they yield large economic Equally, however, experience shows that the diffi- and social bene�ts. Once this is done, the incremental culties in devising and implementing soft measures cost of planning for adaptation to climate change is are often underestimated because they may involve minimal, because it should form a regular element of changes in expectations or (quasi-) established prop- the responsibilities of institutions and the design of erty rights that are strongly resisted. policies. All EACC estimates rely upon the assump- tion that investments in adaptation take place within The analysis of the global costs of adaptation relies a framework of appropriate development policies and heavily on the costs of hard measures. It is much efficient management of the economic sectors. simpler to estimate the costs of new or replace- ment investment to provide protection against the The social analyses in all countries pointed to the effects of climate change than it is to estimate the need and importance of improvements in safety nets, costs of creating new institutions and implement- community-based resource management systems, and ing better policies. In many cases the monetary costs disaster preparedness. It is also necessary to acceler- of soft adaptation are zero or negative in the longer ate the decentralization process and devolve decision run, because the changes bring greater bene�ts than making to the local level to promote local-level adapta- merely adaptation to climate change. Unfortunately, tion and preparedness. Some of the economic models the political and social costs may be perceived as incorporated soft adaptation measures in agriculture, being very high, while the wider bene�ts are hard to such as improvements in extension services and mar- assess and, sometimes, to achieve. keting networks. Others should be exploited, such as 96 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT research and development of early maturing plant Conclusion varieties and other innovations related to impacts of climate change on crops, livestock products, and pest The related messages of uncertainty, flexibility, and control, as well as improvements in land tenure sys- time are central to this report. Some speci�c conclu- tems and improved entrepreneurial skills to generate sions may be drawn about the implications of climate off-farm income. In the transport sector, some signif- change and appropriate adaptation measures. But icant soft approaches include proper timing of road even more remains uncertain, so that the essence of construction, routine and timely road maintenance, adaptation is learning how to cope with greater lev- and upgraded road design speci�cations (including els of uncertainty. Shifting resources toward more the choice of materials). productive uses and away from less productive ones in the context of uncertainty is already a principal In Ghana, a number of soft measures were given prior- aim of development. Climate change increases the ity over hard measures, including (a) upgrading peri- importance of achieving this aim, but it makes the urban slums and controlling the development of new task more complex. Time is also crucial. On some ones; (b) the protection, management, and sustainable issues it is possible—and necessary—to implement use of coastal wetlands; and (c) a review of Ghana’s adaptation measures within the next 5–10 years, but coastal development plans to take into consideration the whole process will extend over many decades. It is climate change adaptation, including strengthened trivial to note that investments designed for a future protection of coastlines and ports; additional flood that never materializes should be avoided. It is much protection measures; and greater attention to the pro- less trivial to identify what those investments are. It tection of coastal communities and �shery industry. may be wise to undertake some forms of adaptation now, especially those that provide resilience to a wide Several of the country studies—e.g. Bolivia, Mozam- range of climate outcomes. But, equally, it may be bique and Vietnam—emphasise the links between better to postpone expensive investments in adapta- autonomous and planned adaptation in agriculture, tion until there is less uncertainty about whether they forestry and aquaculture in responding to the effects will be justi�ed. of climate change. Usually the planned adaptation involves some combination of expenditures on agricul- tural research and extension and investment in irriga- tion development. At the same time, it is important to ensure that market structures and incentives sup- port the adoption of improved methods of production. This was particularly evident for cat�sh cultivation in Vietnam for which autonomous adaptation is likely to be dominant, provided that producer margins are not squeezed by processors and export companies. LESSON 7: Hard and soft approaches to adaptation are two sides of the same coin. Good policies, planning, and institu- tions are essential to ensure that more capital-inten- sive measures are used in the right circumstances and yield the expected bene�ts. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 97 References Anthoff, D., and R.S.J. Tol. 2008. “The Impact of Climate Change on the Balanced Growth Equivalent.� Working Paper 228. Dublin, Ireland: Economic and Social Research Institute. Barnett, J., and M. Webber. 2010. “Accommodating Migration to Promote Adaptation to Climate Change.� A policy brief prepared for the Secretariat of the Swedish Commission on Climate Change and Development and the World Bank World Development Report 2010 team. Melbourne: Department of Resource Management and Geography, The University of Melbourne. Blankenspoor, B., S. Dasgupta, B. Laplante, D. Wheeler. 2010. Economics of Adaptation to Extreme Weather Events in Developing Countries. Discussion Paper Number 1. Washington, DC: World Bank. Ghana Statistical Service, GSS (2008). Ghana Demographic and Health Survey Preliminary Report. Calverton, Maryland USA: MEASURE DHS, Macro International Inc Hope, C. 2006. “The Marginal Impact of CO2 from PAGE2002: An Integrated Assessment Model Incorporating the IPCC’s Five Reasons for Concern.� Integrated Assessment 6 (1): 19–56. Hughes, G., P. Chinowsky, K. Strzepek. 2010. The Costs of Adapting to Climate Change for Infrastructure. Discussion Paper Number 2. Washington, DC: World Bank. IPCC (Intergovernmental Panel for Climate Change). 2007. “Summary for Policymakers.� In M.L. Parry, O.F. Canziani, J.P. Palutikof, P.J. van der Linden, and C.E. Hanson, eds. Climate Change 2007: Impacts, Adaptation, and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, UK: Cambridge University Press. IWTC. 2006. “Statement on Tropical Cyclones and Climate Change.� WMO International Workshop on Tropical Cyclones, San Jose, Costa Rica, November 2006. Kellerer, Hans, Ulrich Pferschy, and D. Pisinger. 2004. Knapsack Problems. Berlin: Springer. Lange. G.M and Dasgupta. S, Thomas. T, Murray. S, Blankespoor. B, Sander. K, Essam. T. 2010. Economics of Adaptation to Climate Change-Ecosystem Services. Discussion Paper Number 7. Washington, DC: World Bank. 98 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT Nelson, G.C, M.W. Rosegrant, K. Koo, R. Robertson. T. Sulser, T. Zhu, C. Ringler, S. Msangi, A. Palazzo, M. Batka, M. Magalhaes, R. Valmonte-Santos, M. Ewing, D. Lee. 2010. Costs of Agriculture Adaptation to Climate Change. Discussion Paper Number 4. Washington, DC: World Bank. Nicholls. R, S. Brown, S. Hanson, J. Hinkel. 2010. Economics of Coastal Zone Adaptation to Climate Change. Discussion Paper Number 10. Washington, DC: World Bank. Nordhaus, W.D. 2002. “Modelling Induced Innovation in Climate-Change Policy.� In A. Grübler, N. Nakicenovic, and W.D. Nordhaus, eds. Technolgical Change and the Environment. Washington, DC: Resources for the Future. Pandey, K. 2010. Costs of Adapting to Climate Change for Human Health in Developing Countries. Washington, Discussion Paper Number 11. DC: World Bank. Parry, M., N. Arnell, P. Berry, D. Dodman, S. Fankhauser, C. Hope, S. Kovats, R. Nicholls, D. Satterthwaite, R. Tiffin, and T. Wheeler. 2009. Assessing the Costs of Adaptation to Climate Change: A review of the UNFCCC and Other Recent Estimates. London: International Institute for Environment and Development and the Grantham Institute for Climate Change, Imperial College. RMSI. 2009. Malawi: Economic vulnerability and disaster risk assessment. Draft �nal report prepared for the World Bank, August. Mimeo. Sedjo, R.A. 2010. Adaptation of Forests to Climate Change. Discussion Paper Number 3. Washington, DC: World Bank. Strzepek, K.M, A.L. Mccluskey. 2010. Modeling the Impact of Climate Change on Global Hydrology and Water Availability. Discussion Paper Number 8. Washington, DC: World Bank. Strzepek, K.M and A.C. Schlosser. 2010. Climate Change Scenarios and Climate Data. Discussion Paper Number 9. Washington, DC: World Bank. Sumaila, U.R, W.W.L. Cheung. 2010. Cost of Adapting Fisheries to Climate Change. Discussion Paper Number 5. Washington, DC: World Bank. UNFCCC (United Nations Framework Convention on Climate Change). 2007. Climate Change: Impacts, Vulnerabilities, and Adaptation in Developing Countries. Bonn, Germany: United Nations Framework Convention on Climate Change. Ward, P.J, P. Pauw, L.M. Brander, C.J. Jeroen, H. Aerts, K.M. Strzepek. 2010. Costs of Adaptation Related to Industrial and Municipal Water Supply and Riverine Flood Protection. Discussion Paper Number 6. Washington, DC: World Bank. Weingartner, H.M. 1963. Mathematical Programming and the Analysis of Capital Budgeting Problems. Englewood Cliffs, NJ: Prentice-Hall. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 99 World Bank. 2010a. The Cost to Developing Countries of Adapting to Climate Change: New Methods and Estimates. Washington, DC: World Bank. World Bank. 2010b. Economics of Adaptation to Climate Change: Social Synthesis Report. Washington, DC: World Bank. World Bank. 2010c. Bangladesh: Economics of Adaptation to Climate Change. Washington, DC: World Bank. World Bank. 2010d. Bolivia: Adaptation to Climate Change: Vulnerability Assessment and Economic Aspects. Washington, DC: World Bank (forthcoming). World Bank. 2010e. Ethiopia: Economics of Adaptation to Climate Change. Washington, DC: World Bank. World Bank. 2010f. Ghana: Economics of Adaptation to Climate Change. Washington, DC: World Bank. World Bank. 2010g. Mozambique: Economics of Adaptation to Climate Change. Washington, DC: World Bank. World Bank. 2010h. Samoa: Economics of Adaptation to Climate Change. Washington, DC: World Bank. World Bank. 2010i. Vietnam: Economics of Adaptation to Climate Change. Washington, DC: World Bank. WRI (World Resources Institute). 2007. Weathering the Storm: Options for Framing Adaptation and Development. H. McGray, A. Hammill, R. Bradley, with E.L. Schipper and J-E. Parry, eds. Washington, DC: WRI. Yu, W.H., et al. 2010. Climate Change Risks and Food Security in Bangladesh. London: Earthscan Publishers. 100 ECONOMI C S OF ADAP TAT I ON TO C LI MAT E C HAN G E: SYNT HE S IS R EP O RT Notes 1. UNFCCC (United Nations Framework Convention on Climate Change), 2007. Climate Change: Impacts, Vulnerabilities, and Adaptation in Developing Countries. Bonn, Germany. 2. This number is slightly smaller than the original (US$75 billion) reported in the World Bank 2010a, reflecting our revised estimates in the agricultural sector. 3. This number is slightly smaller than the original (US$75 Billion) reported in the World Bank 2010a, reflecting our revised estimates in the agricultural sector. 4 Hughes et al 2010. 5 Nicholls et al 2010. 6 Ward et al 2010. 7 Nelson et al 2010. 8 Pandey 2010. 9. The UNFCCC study only works with our equivalent gross-sum, but we still use our X-sum as the best estimate of the costs. This eventually narrows the difference between the two study results. 10 Climate change after 2050 is not ignored in the analysis. It is assumed that major investment decisions for coastal protection and infrastructure look at climate risks 50 years ahead of the date of investment. Hence, climate conditions projected for 2100 are taken into account in designing and costing sea defenses, roads, buildings, etc that are constructed in 2050. 11 See Kellerer et al. (2004). A widely used approach is to formulate the problem of selecting and schedul- ing climate-resilient investment alternatives as a project selection model related to the knapsack problem (Weingartner, 1963) and solved as a mixed integer programming problem. A more practical approach is to use real option analysis, where uncertainty (or the risk) is incorporated into the business decision of undertaking or not a certain investment – in this case alternative adaptation actions. Given the paucity of data and the levels of uncertainty, simpler approaches may be called for. 12 In the global analyses, investments in coastal protection and infrastructure have a time horizon of 2100, so that investment decisions can be made 50 years ahead, i.e., in 2050.bb 13 It is important to note that this is the result of both government’s desire to work with research teams they have experience with and the project’s desire to produce results. 14 Institutional issues were only to be looked at in the context of the social work, and are more widely dis- cussed in the EACC-Social Synthesis Report. 15 These additional costs for the provision of public goods must not be confused with overall economic damages and cannot be usefully compared with mitigation costs. 16 There are a number of reasons why it has not been possible to optimize cross-sectorally and inter-tempo- rally; for example, the CGE models and the sectoral models have been developed separately and generally do permit resources to flow across sectors for which adaptation strategies have been developed. 17 See footnote to Table 2 below. This number is slightly smaller than the original ($75 billion) reported in the EACC global report (World Bank 2010), reflecting our revised estimates in the agricultural sector. 18 Hughes et al 2010. 19 Nicholls et al 2010. 20 Ward et al 2010g. 21 Nelson et al 2010. ECON O MI CS OF ADAPTATION TO CL IM ATE C HAN G E: SYNT HE S IS R E PO RT 101 22 Pandey 2010. 23 Blankespoor et. al 2010. 24 The UNFCCC study only works with our equivalent gross-sum, but we still use our X-sum as the best estimate of the costs. This eventually narrows the difference between the two study results. 25 World Bank 2010g. 26 The study of sea level rise in Mozambique considers three sea-level rise scenarios—low, medium, and high, ranging between 40cm and 126cm by 2100—following the approach used in the global study. 27 The CGE model takes into account the full transportation sector, including coastal infrastructure. Coastal adaptation options are studied and presented separately. 28 Options include both hard and soft infrastructural components; for example, changes in transportation operation and maintenance, new design standards, transfer of relevant technology to stakeholders, and safety measures. 29 Welfare is measured by aggregate �nal demand (sum of consumption, investment, and government expenditure). 30 Coefficient of variation (CV) is the standard deviation (SD) divided by the mean of the year-to-year growth rates. 31 World Bank 2010e. 32 World Bank 2010f. 33 It makes no economic sense to invest more than this amount in adaptation measures aimed at making Ghana as well off as it would be in the absence of climate change. If the costs of adaptation policy measures aimed at restoring aggregate welfare to the baseline are higher than the welfare loss from climate change, it would be cheaper to restore welfare through lump-sum compensation payments. 34 World Bank 2010c. 35 The inundation depths and potential vulnerable zones are estimated based on a hydrodynamic modeling system of the Bay of Bengal combined with historical data of inundation depths of all 19 cyclones for the base case, and 5 potential cyclone tracks consisting of the 4 large cyclones of 1974, 1988, 1991, and 2007 for the second scenario. 36 World Bank 2010d. 37 The global wet scenario was drier than the Bolivia Wet Scenario, so it was not considered in the study. 38 1146 mm reported by Aquastat, 1459 mm from PNCC (2007), 1189 mm own estimations from CRU data. 39 Ministry of Environment and Water, 2008. 40 The water supply system of La Paz - El Alto, had suffered a scarcity alert in the wet season of 2008, which was repeated on the fall of 2009. Emergency measures, such as drilling emergency wells were implemented to be able to meet demand levels in those periods. 41 A wet climate scenario was not available at the time of the analysis. Only the Bolivia dry scenario was used for this exercise. 42 Seventy four potential projects have been identi�ed in 16 of the 22 sub-basins. Of these 16 sub-basins, only 3 experience water scarcity prior to 2050, even under the “dry� scenario. 43 The budgetary decentralization rule at the sub-watershed level that was investigated was to provide equal per capita investment resources across all sub-basins and to allow them to optimize independently. 44 World Bank 2010i. 45 World Bank 2010h. Ministry of Foreign Affairs Government of the Netherlands UNITED KINGDOM The World Bank 1818 H Street NW Washington DC, 20433 USA www.worldbank.org/eacc www.worldbank.org/sdcc