G h a n a CO U N T RY ST U DY i Economics of Adaptation to Climate Change GHANA
ii E C O N O M I C S O F A D A P T AT I O N T O C L I M A T E C HAN G EEACC Publications and Reports1. Economics of Adaptation to Climate Change: Synthesis Report2. Economics of Adaptation to Climate Change: Social Synthesis Report3. The Cost to Developing Countries of Adapting to Climate Change: New Methods and EstimatesCountry Case Studies:1. Bangladesh: Economics of Adaptation to Climate Change2. Bolivia: Adaptation to Climate Change: Vulnerability Assessment and Economic Aspects3. Ethiopia : Economics of Adaptation to Climate Change4. Ghana: Economics of Adaptation to Climate Change5. Mozambique: Economics of Adaptation to Climate Change6. Samoa: Economics of Adaptation to Climate Change7. Vietnam: Economics of Adaptation to Climate ChangeDiscussion Papers:1. Economics of Adaptation to Extreme Weather Events in Developing Countries2. The Costs of Adapting to Climate Change for Infrastructure3. Adaptation of Forests to Climate Change4. Costs of Agriculture Adaptation to Climate Change5. Cost of Adapting Fisheries to Climate Change6. Costs of Adaptation Related to Industrial and Municipal Water Supply and Riverine Flood Protection7. Economics of Adaptation to Climate Change-Ecosystem Services8. Modeling the Impact of Climate Change on Global Hydrology and Water Availability9. Climate Change Scenarios and Climate Data10. Economics of Coastal Zone Adaptation to Climate Change11. Costs of Adapting to Climate Change for Human Health in Developing Countries12. Social Dimensions of Adaptation to Climate Change in Bangladesh13. Social Dimensions of Adaptation to Climate Change in Bolivia14. Social Dimensions of Adaptation to Climate Change in Ethiopia15. Social Dimensions of Adaptation to Climate Change in Ghana16. Social Dimensions of Adaptation to Climate Change in Mozambique17. Social Dimensions of Adaptation to Climate Change in Vietnam18. Participatory Scenario Development Approaches for Identifying Pro-Poor Adaptation Options19. Participatory Scenario Development Approaches for Pro-Poor Adaptation: Capacity Development Manual
G h a n a CO U N T RY ST U DY iEconomics of Adaptationto Climate ChangeG hana Ministry of Foreign Affairs Government of the Netherlands
G h a n a CO U N T RY ST U DY iiiContentsAbbreviations and Acronyms viiAcknowledgements ixCaveat xi Executive Summary xiiiImpacts of Climate Change xiiiAdaptation to Climate Change xivLessons and Policy Recommendations xiv1 Introduction 1Study Objectives 2Organization of Report 32 Overview of the EACC Global Track Study 53 Methodology 11Overall Approach and Key Assumptions 11Climate Forecasts 12Sector-Specific Approaches 144 Study Results 35Overview of the Ghanaian Economy 35Climate Change Projections 39Economic Impacts of Climate Change – CGE Model Results 42Economic Implications of Adaptation to Climate Change – CGE Model Results 54Adaptation Options 57Adaptation Costs 59Social Dimensions 62
iv E C O N O M I C S O F A D A P T AT I O N T O C L I M A T E C HAN G E5 Summary and Policy Implications 69Climate Change Impacts 69Adaptation to Climate Change Costs 70Looking forward 70Summary Matrix 73References 77Annexes (available on line at www.worldbank.org/eacc) Annex 1. Cli-Crop Modelling for AgricultureAnnex 2. Dose-Response Model for RoadsAnnex 3. IMPEND Model for Energy and WaterAnnex 4. DIVA Model for Coastal ZoneAnnex 5. Social Dimensions of Climate ChangeAnnex 6. Computable General Equilibrium (CGE) ModelingTablesTable 1. Total Annual Costs of Adaptation for All Sectors, by Region, 2010–50 7Table 2. Total Annual Costs of Adaptation for all Sectors, by Region and Period, 2010–50 7Table 3. A Comparison of Adaptation Cost Estimates ($ billions) 8Table 4. GCM Scenarios for Ghana Country Track Study 12Table 5. Trends in the Growth Rate of the Transport Sector 19Table 6. Share of the Transport Sector in Total GDP in Purchaser’s Value, 2002–2007 (%) 19Table 7. Road Sector Vulnerability to Potential Climate Change 20Table 8. Dose-Response Descriptions for Maintenance Costs 20Table 9. Electricity and Water Subsectors Growth Rates of Real GDP 23Table 10. Electricity and Water’s Share of GDP and Contribution to Overall GDP Growth 23Table 11. Projected Population of the Coastal Regions and Estimated Population at risk to Sea Level Rise 27Table 12. Land Area Distributions of the Ten Provinces of Ghana, divided into three zones 30Table 13. Economic Development Indicators in Ghana, 2005 to 2008 36Table 14. Temperature (Co) in Regional CC Scenarios, 2010–50 38Table 15. Precipitation Projections for Ghana’s 16 subbasins – Descriptive Statistics 41
G h a n a CO U N T RY ST U DY vTable 16. Standard Deviation of Annual Real Consumption Growth 45Table 17. Welfare Impact without Adaptation Investments 45Table 18. DIVA Annual Results for High Sea Level Rise Scenario 51Table 19. DIVA Annual Results for Low Sea Level Rise Scenario 52Table 20. Mean, Standard Deviation, and Extreme Values of Annual GDP Growth Rates by Region, 2006–50 53Table 21. Deviations of Welfare from Baseline under Alternative Adaptation Strategies 56Table 22. Average Annual Real GDP Growth Rates (2010–50) under Alternative Adaptation Strategies (%) 56Table 23. Regional Shares in Agricultural Production by Commodity 60Table 24. Commodity Composition of Agricultural Production by Region 61Table 25. Summary of Ghana Coastal Seal Level Rise (SLR) Annual Adaptations Costs 65Table 26. Summary recommendation on low-regret options and policy interventions in short and long term following the Ghana EACC Analysis 74FiguresFigure 1. Shares of the Total Annual Costs of Adaptation by Region 2010–50 7Figure 2. Flow Chart of Model Sequencing 14Figure 3. Trends in the Growth Rate of the Agricultural Sector, 2002–10 16Figure 4. Rural-Urban Potable Water Coverage by Region, 2006 and 2007 (%) 26Figure 5. Ghana, West Africa: (a) Geographical location, (b) Administrative units (termed provinces) and major coastal towns, and (c) The coastal zone 29Figure 6. Ghana Sector Contribution to the GDP 37Figure 7. Annual Real Growth Rate by Sector, 2002–09 37Figure 8. Ghana Dry Scenario Temperature Changes Compared to Base, 2010–50 39Figure 9. Temperature Variability Compared to Base 40Figure 10. Surface flow average difference from the no-climate change scenario, 2010–50 41Figure 11. Annual Deviations of Real GDP from Base, 2010–2050 (%) 42Figure 12. GDP Growth Path in Levels 2010–2050 43Figure 13. Terminal Period Real GDP (average annual GDP, 2046–50) 43Figure 14. Terminal Real Household Consumption Level (annual average, 2046–50) relative to 2005 Level 45Figure 15. Decomposition of Climate Change Impacts on Present Vale of Real Absorption (deviation from base in billion $) 46
vi E C O N O M I C S O F A D A P T AT I O N T O C L I M A T E C HAN G EFigure 16. Average Annual Agricultural Real GDP, terminal period 2046–50 46Figure 17a. Real GDP Deviation from Base for Maize, 2020–50 47Figure 17b. Real GDP Deviation from Base for Cocoa, 2020–50 47Figure 17c. Real GDP Deviation from Base for Cocoa, South Savannah 2020–50 47Figure 18. Climate Change Impacts of Cocoa Productivity in Ghana (deviations from baseline yields) 48Figure 19. Decadal Average Ratios of Future Livestock Net Revenues to Net Revenues under Baseline Conditions, Ghana Dry (on left) and Wet (on right) Scenarios, 2001–50 50Figure 20. Decadal Average Ratios of Future Livestock Net Revenues to Net Revenues under Baseline Conditions, Global Dry (on left) and Wet (on right) Scenarios, 2001–50 50Figure 21. Average Annual Water and Energy Sector Real GDP, 2046–50 51Figure 22. Deviations of Welfare from Baseline under Alternative Adaptation Strategies 56Figure 23. Annual Road Maintenance Costs, 2010–50 63Figure 24. Annual Average Road Maintenance Costs, 2010–50 63Figure 25. Total Energy Adaptation Costs 63
G h a n a CO U N T RY ST U DY viiAbbreviationsand AcronymsAR4 Fourth Assessment Report ITCZ Inter-Tropical Conversion ZoneBAU Business-as-usual LCA Latin America and Caribbean RegionCAADP Comprehensive Africa Agriculture MDGs Millennium Development Goals Development Program NCAR National Center forCGE Computable general equilibrium Atmospheric ResearchCO2 Carbon dioxide NAPA National adaptation plans of actionCMI Climate moisture index NCCAS National Climate ChangeCSIRO Commonwealth Scientific and Adaptation Strategy Industrial Organisation NGO Nongovernmental organizationDIVA Dynamic and interactive ODA Official development assistance vulnerability assessment PaMs Policies and measuresEACC Economics of Adaptation PET Potential evapotranspiration to Climate Change Ppm Parts per millionEAP East Asia and Pacific Region RD Research and developmentECA Europe and Central Asia Region SAS South Asia RegionENSO El Niño-Southern Oscillation SRES Special Report on EmissionsGCM General circulation model ScenariosGDP Gross domestic product SSA Sub-Saharan AfricaGHG Greenhouse gases SST Sea surface temperatureGIS Geographical information system TAR Third Assessment ReportGPRS Ghana Poverty Reduction Strategy UNDP United Nations DevelopmentGWCL Ghana Water Company Limited ProgrammeHDI Human Development Index UNFCCC United Nations FrameworkIFPRI International Food Policy Convention on Climate Change Research Institute VRA Volta River AuthorityIMPACT International model for policy analysis of agricultural commodities and tradeIPCC Intergovernmental Panel on Climate Change Note: Unless otherwise noted, all dollars are U.S. dollars.
G h a n a CO U N T RY ST U DY ixAcknowledgmentsThis study would not have been successfully the specific situation of Ghana. Particularly, wecompleted without the inputs of a large number gratefully acknowledge Dirk Willenbockel, Kenof organizations and individuals. Profound grat- Strzepek, Eihab Fathelrahman, Robert Nicholls,itude goes to officials from all the government Len Wright, Chas Fant, Paul Chinowsky, Chan-ministries, departments, and agencies, who con- ning Arndt, Sherman Robinson, Michelle Mini-tributed immensely to the success of the study by hane, William Farmer, Brent Boehlert, Alyssaproviding data and other information for the McClusky, and Jean-Marc Mayotte. Thanks alsoanalysis as well as the validation of methodology to the social scientist team that developed theand adaptation options. social dimensions of climate change, including Tony Dogbe, Joseph Yaro, David Pessey, EmiliaIn particular, we would like to recognize the Arthur, George Ahiable, Tia Yahaya, Kamilteams at the Environmental Protection Agency, Abdul Salam, Samantha Boardley, Simon Mead,Ministry of Environment, Science and Technol- and Livia Bizikova. In Ghana, consultants Danielogy, Ministry of Finance and Economic Plan- Sarpong, Dyson Jumpah, and Philip Acquahning, the National Development Planning reviewed sector strategies and adaptation options,Commission, and Ministry of Agriculture. In and Saadia Bobtoya supported the team withparticular, we would like to thank William Agye- information management and communications.mang-Bonsu, Jonathan Allotey, Alhassan Iddi- The technical writer for Ghana Case was Johnrisu, David Quist, Rudolph Kuuzegh, George Asafu-Adjaye.Scott, Winfred Nelson, and Regina Adutwumfor the overall guidance provided in the course The team would also like to thank developmentof the study. Many more contributed with ideas partners in Ghana for excellent coordination ofand technical input in July, August, and October work related to this study, including Sean Doolanof 2009 during workshops and meetings, and (United Kingdom Department for Internationalwell as during the final validation workshop in Development), Ton van der Zon (Royal Nether-September 2010. lands Embassy), Wagn Winkel (Royal Danish Embassy), Shigeki Komatsubara and StephenWe wish to also acknowledge the inputs of the Duah-Yentumi (United Nations Developmentglobal modeling team for their diligence in fitting Program), and Jannik Vaa (Europeanclimate change scenarios and economic models to Commission).
x E C O N O M I C S O F A D A P T AT I O N T O C L I M A T E C HAN G EThe study was kindly financed by the govern- Mearns, Sergio Margulis (team leader of thements of the United Kingdom, The Netherlands, overall EACC study) , Stephen Mink, Urvashiand Switzerland, as well as the governments of Narain, and Victoria Bruce-Goga. Several BankNorway and Finland through the Trust Fund for staff have commented and provided insight toEnvironmental and Social Sustainable Develop- sectors covered in this report, including Ajayment (TF-ESSD) and the World Bank. Kumar, Chris Jackson, Herbert Acquay, Ishac Diwan, John Richardson, Osman Kadir,The World Bank Task Team included Peter Sebastien Dessus, Shelley McMillan, and SunilKristensen (Task Team Leader), Aziz Bouzaher, Mathrani. Robert Livernash provided editorialAnne Kuriakose, John Fraser Stewart, Kiran services, Jim Cantrell contributed editorial inputPandey (Coordinator EACC country studies), and coordinated production, and Hugo MansillaRaffaello Cervigni, Robert Schneider, Robin provided editorial and production support.
G h a n a CO U N T RY ST U DY xiCaveatThis study is experimental and innovative innature. The CGE modeling has made use ofmany assumptions to estimate the economics ofadaptation to climate change in Ghana in a longtime horizon. The numbers and results in thereport should be used with caution, and consid-ered indicative. While the report suggests shortand long-term policy and investment options, theauthors believe that further review of the cost-benefit of adaptation options should beundertaken.
G h a n a CO U N T RY ST U DY xiiiExecutive SummaryImpacts of Climate Change fluctuations will increase the risk of floods and/or droughts in both rural and urban areas. Because mostClimate change is projected to have significant impacts of these changes are caused by upstream areas out-on Ghana. Although there will be fluctuations in both side the territory of Ghana, there is a need for dia-annual temperatures and precipitation, the trend for logue with Ghana’s neighbors on the management oftemperature over the period 2010–50 indicates warm- shared water resources.ing in all regions. The highest temperature increaseswill be in the Northern, Upper East, and Upper West Because Ghana’s economy is predominantly basedregions, while the lowest will be in the Brong Ahafo on agriculture, it will suffer severe economic conse-region. For example, under one of the climate scenar- quences from climate change. Although there will beios (Ghana Dry), temperatures in the three regions of considerable variation in real gross domestic productthe North will rise by 2.1–2.4°C by 2050. In compari- (GDP) growth, the overall trend over 2006–50 clearlyson, the predicted rise in the Ashanti, Western, East- indicates a downward trajectory in the absence ofern, Central, and Volta regions will be 1.7–2.0°C, and adaptation to climate change. Toward 2050, annualthe rise in the Brong Ahafo region will be 1.3–1.6°C. real GDP is projected to be 1.9 to 7.2 percent lower than in a dynamic baseline scenario without anthro-The forecast for precipitation indicates a cyclical pogenic climate change. Real household consump-pattern over the period 2010–50 for all regions, with tion also declines relative to the base scenario in all thehigh rainfall levels followed by a drought every four climate change scenarios analyzed in this study.decade or so. The wettest parts of the country areexpected to be the Forest agroecological zone Adverse agricultural productivity impacts become(Ashanti and Western regions) and Coastal agroeco- more pronounced over time. Relative to the baselinelogical zone (Volta, Eastern, Central, and Greater projection for the middle of the 21st century, agricul-Accra regions). The northern and southern Savan- tural GDP is estimated to decline by 3 to 8 percent.nah zones are expected to be relatively dry. The projections for cocoa pose serious socioeconomic implications in view of cocoa’s significant contribu-There will also be wide fluctuations in runoff and tion to national income and farmers’ livelihoods.stream flows, with areas in the Volta basin experienc-ing significant reductions in runoff, while the south- Damage to the coastal zone in the form of flooding,western area will experience increases. These land loss, and forced migration is estimated to be
xiv E C O N O M I C S O F A D A P T AT I O N T O C L I M A T E C HAN G E$4.8 million per annum by the 2020s, rising to Incomplete partial equilibrium modeling puts econo-$5.7 million per annum by the 2030s. mywide adaptation costs in a mid-range of $300– $400 million per annum. Partial equilibrium, asThe predicted climatic changes will have adverse opposed to the general equilibrium approach, consid-effects on human well-being and activities, food secu- ers each subsector of the economy in isolation fromrity, and water availability. In response to these climate the other sectors when it comes to prices and incomechanges, people will migrate in search of better land interactions among stakeholders.and environment. The migration and relocation ofpopulation from rural to urban areas will raisedemand and put pressure and on municipal ser- Lessons and Policyvices—including water supply and sanitation, public Recommendationshealth, energy, transportation, and housing services.Such higher demand coupled with weak infrastruc-ture and lack of services will slow economic growth Agricultureand development. Migration will occur not only There is a need to (a) increase investment in agri-within the country, but also from countries to the north cultural RD, backed by extension services, toof Ghana, which will also become hotter and drier. produce new crops and livestock, as well as early- maturing varieties; (b) improve water storage capacity to utilize excess water in wet years andAdaptation to Climate use it when it is needed during dry years; (c)Change improve agricultural and livestock extension ser- vices and marketing networks; (d) construct small to mid-size irrigation facilities; (e) improve entre-Adaptation in this study is aimed at restoring aggre- preneurial skills to generate off-farm incomegate national output to baseline, rather than restoring (alternative livelihoods); and (f) improve access toeach sector to the baseline. This suggests that even loans and microcredit.with adaptation, there will still be some residual dam-age at the sector level. Given the scarcity of resources Road transportat the government’s disposal, tough choices must be Recommended actions include proper timing of roadmade in the design and sectoral balance of the construction; for example, before the rainy season.national development strategy in light of the chal- There is also a need to ensure routine and timelylenges posed by climate change. maintenance; review overall road design criteria, including materials and drainage, road sizes, and pro-In the absence of adaptation, climate change tection of road shoulders; and reform road designcauses a decline in real output growth for all the standards to meet higher needs against extremeglobal circulation model (GCM) results. Planned events such as floods and droughts.adaptation can be effective in compensating theadverse impacts of climate change. Water and energy Recommended hard options for the water subsec-The general equilibrium modeling indicates that tor include increased water transfer from the Voltalosses in agriculture could be as much as $122 mil- basin to meet the needs of a growing urban popu-lion per annum, while losses in transport and hydro- lation; construction of efficient infrastructure; andpower could be up to $630 million and $70 million, blocking of dry-stream channels to harvest rainwa-respectively. Total economywide impacts are esti- ter to recharge the groundwater system, whichmated to range from $158–$765 million per annum. serves as an alternative water supply during dry
G h a n a CO U N T RY ST U DY xvyears. A number of soft options were also deemed level. The poorest are particularly vulnerable toto be of high priority: afforestation, improved land climate shocks, as they do not have stored assets touse practices, protection of river courses, and use during times of stress. A pro-poor approachdesedimentation of reservoirs. to climate change adaptation would look not only at reducing shocks to households, but also engageDiversification of the energy mix and development in transformative adaptation strategies thatof renewable sources—such as solar, wind, biomass, increase resilience and overcome past biases inwaste conversion, and mini-hydro dams—are priori- subnational investment.ties, as are soft options such as promoting policies andmeasures aimed at enhancing energy efficiency in all Geographically targeted, multisectoral interventionssectors. The government also should commit to a are needed to reduce the “development deficit” ofstrict infrastructure maintenance regime. vulnerable regions. Poverty and sensitivity to climate- related hazards are increasingly concentrated in par-Coastal zone ticular regions within the country. In many cases,The modeling results generally show that the poor communities—such as recent urban in-investment costs of coastal zone adaptation are migrants—are relegated to the most marginal areaslikely to be uneconomic because the costs are likely of the city. Adaptation policies at the national levelto far exceed any benefits, so defending the entire must take into account the diverse socioecological set-coastline by building dikes and sea defense walls is tings within the country, and devise area-specificnot a sensible strategy. A better strategy would be interventions that can support the livelihoods of theseto protect key investments and natural resources— vulnerable populations. Multisectoral interventionsports, harbours, beaches, and coastal mangroves— that aim to improve area resilience through reducingand to zone significant new infrastructure away the development gap are particularly effective formsfrom vulnerable areas to the greatest extent possi- of investment, including programming in education,ble. Emphasis must be placed on soft options such social protection and health, roads, market services,as enhancing capacity in early warning systems natural resource management, and skills training.and the use of GIS and satellite imagery for coastalzone management. New oil and gas development Regional integrationand related infrastructure and regional develop- It is important for Ghana to strengthen dialoguement in the Western region would need to be with neighboring countries to effectively deal withdesigned with climate change adaptation in mind. the challenges of climate change. Areas where negotiations and consultations would be requiredSocial dimensions are in the management of shared water resourcesComplementary investments in both hard and soft and regional migration of people.adaptation options are needed to ensure effectiveuse of infrastructure and to meet the needs of the Long-term planningpoorest. Adaptation investments in hard infrastruc- Given the development challenges and threats posedture without complementary investments in policy, by climate change and variability, Ghana needs aservice, and extension support will not operate in an long-term national plan that takes these factors intooptimally efficient manner. account. Currently, Ghana only has a medium-term development plan covering 2010–13. The long-termA policy shift is needed—from support for coping plan also needs to be integrated into the plans of thestrategies for climate shocks at the household regional coordinating councils and the district devel-level, to transformative adaptation strategies that opment plans to provide a coherent and integratedcan increase resilience at the household and area approach to development planning.
G h a n a CO U N T RY ST U DY 1IntroductionClimate change and variability is arguably one of example, regional climate systems such as the Elthe greatest challenges facing humankind this cen- Niño-Southern Oscillation phenomenon and thetury and into the next. Developing countries, in par- Asian monsoon will be altered.ticular those in Sub-Saharan Africa (SSA), areparticularly at risk because they are located in areas Even if GHGs are stabilized at 450ppm, thewhere temperatures will rise the fastest. They are annual mean global temperature will be aboutalso more vulnerable because they are mainly 2°C above preindustrial levels by the middle ofdependent on agriculture, which is the most climate this century due to the amount of gases alreadysensitive sector. Despite some uncertainty about the locked into the climate system. Therefore, theprecision of climate science, there is now general short-run option for both developed and develop-agreement among climate scientists on a number of ing countries is to adapt. However, without anyissues. Firstly, it has been firmly established that the mitigation, an adaptation-based strategy for deal-Earth is undergoing rapid changes due to significant ing with climate change is bound to be too costly.1increases in greenhouse gases (GHGs). For example, This is because a temperature increase far inglobal GHG emissions have roughly doubled since excess of 2°C (e.g., 4°C) is predicted to be associ-the early 1970s; if current policies continue, emis- ated with potentially catastrophic impacts whosesions could rise by over 70 percent during 2008–50. effects may be irreversible. Examples of suchAtmospheric concentrations of carbon dioxide impacts include extinction of half of all species(CO2) have increased by nearly 100 parts per million worldwide, inundation of 30 percent of coastal(ppm) compared to preindustrial levels, reaching wetlands, and increases in disease and malnutri-379 ppm in 2005, and the Earth has warmed 0.7°C tion. Although autonomous (or private) adapta-since 1900 (IPCC 2007; Brohan et al. 2006). Sec- tion is already occurring in various parts of theondly, human activities, particularly burning of fos- world, including SSA, the general view is that thissil fuels and deforestation, have been identified as approach will be incapable of dealing with warm-prime causes of the changes observed in the 20th ing in excess of 2°C. In such situations, plannedcentury and are likely to contribute to further adaptation would be required.changes in the 21st century (IPCC 2001). Thirdly,these atmospheric changes are highly likely to alter 1 While adaptation and mitigation are necessary responses totemperatures, rainfall patterns, sea level, extreme climate change, they need not be mutually exclusive. In fact it has been shown that there can be cobenefits and synergies betweenweather events, and other aspects of climate. For the two responses.
2 E C O N O M I C S O F A D A P T AT I O N T O C L I M A T E C HAN G EAt the 2007 Bali Conference, the developed coun- lacking for many developing countries. To closetries pledged among other things to provide “ade- this information gap, the World Bank initiated thequate, predictable, and sustainable financial Economics of Adaptation to Climate Changeresources and the provision of new and additional (EACC) study in early 2008, supported by fundsresources, including official and concessional fund- from the governments of the Netherlands, Switzer-ing for developing country parties” to assist them in land, and the United Kingdom. The objectives ofadapting to climate change (UNFCCC 2008). In the EACC are to develop an estimate of adapta-order to determine the order and magnitude of the tion costs for developing countries and to helpfinancial assistance required, it is necessary to know decision makers in developing countries under-how much adaptation would cost. Unfortunately, stand and assess the risks posed by climate changecurrent information on adaptation costs, particu- and design better strategies to adapt to climatelarly for developing countries, is not sufficiently change (World Bank 2010a). At the 2007 Balicomprehensive. For example, the World Bank pro- meetings, the Ghana delegation made a request toduced one of the first estimates of adaptation costs the World Bank for assistance to estimate the costfor developing countries in 2006, with estimates of climate change adaptation for planning andranging from $9–$45 billion a year (World Bank budgetary purposes. Ghana was therefore included2006). However, these estimates were restricted to among six other countries in which country-basedthe cost of climate-proofing only three categories of EACC studies would be undertaken. The otherinvestments: official development assistance (ODA) participating countries are Bangladesh, Bolivia,and concessional finance, foreign direct investment, Ethiopia, Mozambique, Samoa, and Vietnam.and gross domestic investment. The Stern Report(Stern 2007) estimated that adaptation costs would This report presents a synthesis of the findingsrange from $4–$37 billion per year by 2050, using from the Ghana EACC case study. The studythe World Bank (2006) approach, while the UNDP’s benefited from close collaboration and input fromestimates were $5–$67 billion a year by 2015. Oxfam various stakeholders, including government agen-International (2007), using national adaptation cies (Ministry of Environment, Science and Tech-action plans (NAPAs), estimated global adaptation nology; Environmental Protection Agency;to be at least $50 billion per year, while UNFCCC Ministry of Finance and Economic Planning; and(2007) estimated adaptation costs for five major sec- Ministry of Energy), civil society organizations,tors to range from $26–$67 billion per year by 2030. and development partners. As part of the GhanaOne of the latest estimates is by the Climate Works EACC study process, a series of participatory sce-Foundation; under their Project Catalyst Initiative, nario development (PSD) workshops highlightedthe costs of adaptation for developing countries are the impact of climate change on vulnerableestimated to lie between $15 and $30 billion for groups and also identified and vetted adaptation2010–20 and $30–$90 billion by 2030 (European strategies for further analyses.Climate Foundation 2009). A recent review of cur-rent climate change adaptation estimates (Parry etal. 2009) argues that the existing estimates are likely Study Objectivesto be gross underestimates due to the exclusion ofsome sectors or the incomplete accounting of cli- The main objectives of this study are to presentmatic effects. estimates of the impacts of climate change for key selected sectors for Ghana and to discuss theWhereas considerable work has been done in a implications for climate change adaptationlarge number of advanced countries on the cost of options and adaptation costs. This type of infor-climate change adaptation, such information is mation can assist policy makers in a number of
G h a n a CO U N T RY ST U DY 3areas. First, it would assist them to make appro- discussing the global EACC study and thepriate budgetary allocations for climate change EACC methodology, which was applied in thisadaptation and to inform the debate on the level study at a more disaggregated level. The sec-of assistance required for the development effort. tion highlights the differential impacts of cli-Secondly, given that scarce resources must be mate change among different regions of theallocated amongst competing needs, the informa- world, including Africa. Chapter 3 presents antion would enable them to make tough choices on overview of the methodology used, includingthe design and sectoral balance of the national the key assumptions. An effort has been madedevelopment strategy in light of the challenges to present this information in nontechnical lan-posed by climate change. The beneficiaries of this guage where possible. The more technicalreport will include not only the government, but aspects of the study can be found in the annexes.also the development partners, nongovernmental The sector results are contained in chapter 4.organizations, researchers, students, and citizens The chapter begins with an overview of theconcerned about the impacts of climate change. Ghanaian economy, followed by the climate projections for Ghana and the overall economic impacts. Next, the results for each sector areOrganization of the Report presented in three parts: climate change impacts, the adaptation options, and the adap-The report is organized as follows. The next tation costs. The final chapter concludes with asection puts the study into context by briefly summary and policy implications.
G h a n a CO U N T RY ST U DY 5Overview of the EACCGlobal Track StudyThe approach adopted in the global track study availability. Construction of the baselines alsowas to use country-level data sets to estimate involved the use of a consistent set of GDP andadaptation costs for all developing countries for population forecasts for 2010–50.2 Two climateseven key sectors of the economy — infrastruc- models were chosen to capture as large a range asture, coastal zones, water supply and flood pro- possible of model predictions, including modeltection, agriculture, fisheries and ecosystem extremes of dry and wet climate projections.services, human health, and forestry. In line with These were the National Center for Atmosphericthe Bali Action Plan’s call for “new and addi- Research (NCAR) CCSM3 and Commonwealthtional” resources to meet adaptation costs, the Scientific and Industrial Research Organizationstudy considered adaptation costs as additional (CSIRO) Mk3.0 models. There is not much dif-to the costs of development. Therefore, the costs ference in the model projections for warming byof measures that would have been undertaken 2050, with both models projecting increases ofeven in the absence of climate change were not about 2°C above pre-industrial levels. However,included. Adaptation cost was thus defined as the projections do vary substantially for precipita-the cost of appropriate capacity to deal with tion changes. Based on the climate moisture indexfuture climate change minus the cost of appro- (CMI), the NCAR model predicts the wettest sce-priate capacity to deal with current climate vari- nario globally (but not necessarily the wettest andation. The latter therefore includes the driest in every location), whereas the CSIRO“adaptation deficit,” which is defined here as the model predicts the driest scenario.lack of sufficient capacity to deal with currentclimate variation. The next step in the process was to predict what the world would look like with climate change. TheThe process of estimating the cost of adaptation 2050 time frame was chosen because of the manybegan with the establishment of a development uncertainties associated with forecasting climatebaseline for each sector. This is the growth path change beyond this period. This was done by esti-that would be followed in the absence of climate mating the impacts on agriculture, forestry, fisher-change to the year 2050 and which determines ies, consumption, human health, water availability,sector-level performance indicators—for exam-ple, productivity growth in agriculture, level of 2 The year 2050 was chosen due to the increasing error associatedinfrastructure assets, level of nutrition, and water with trying to make forecasts beyond this time period.
6 E C O N O M I C S O F A D A P T AT I O N T O C L I M A T E C HAN G Eand physical infrastructure. Adaptation cost was In general, the adaptation costs are dominated bythen calculated as the cost of climate-proofing the costs of infrastructure, coastal zones, andthese resources to enable them to withstand the water supply and flood protection in both scenar-impacts, as well as the cost of assisting people to ios. In terms of the sectoral breakdown, the high-deal with the impacts. Due to the complexity of est costs for East Asia and the Pacific are inmodeling different sectors at a global level, a zero infrastructure and coastal zones; for Sub-Saharandiscount rate was assumed with costs expressed in Africa, water supply and flood protection and2005 constant prices.3 A World Bank study— The agriculture; for Latin America and the Carib-Costs to Developing Countries of Adapting to Climate bean, water supply and flood protection andChange: New Methods and Estimates—offers a detailed coastal zones; and for South Asia, infrastructurediscussion on the logic behind the zero discount and agriculture.rate at the global level (World Bank 2010a). Table 2 indicates that under both climate scenar-The study used three different methods to aggre- ios, total annual adaptation costs rise over time.gate adaptation costs and benefits across sectors For example, for the NCAR model, annual adap-and countries. These were gross (no netting of tation costs are $73 billion during 2010–19, risingcosts), net (benefits are netted across sectors and 45 percent over the next 30 years to reach $106countries), and X-sums (positive and negative items billion in 2040–49. Similarly, for the CSIROare netted within countries but not across coun- model, costs also increase but more rapidly, risingtries). The study estimates that the global cost 67 percent over the entire period, from $57 bil-between 2010 and 2050 of adapting to an approxi- lion a year in 2010–19 to $95 billion bymately 2°C warmer world by 2050 lies between 2040–49.$75 billion and $100 billion a year (Table 1). Figure 1 Shares of the Total AnnualFigure 1 presents a chart of the share of the total Costs of Adaptation by Region, 2010–50costs by region using the CSIRO model and theX-sum cost aggregation method. The East Asia $7and Pacific Region has the highest share of the $4adaptation cost with 25 percent, followed by 7% 4% $25Sub-Saharan Africa and Latin America and the 25%Caribbean with 22 percent each, and then bySouth Asia with 20 percent. Europe and Central $22 22%Asia and the Middle East and North Africa havethe lowest shares of 8 percent and 4 percent,respectively. Although the NCAR model esti- 22%mates tend to be generally higher than the $22 20%CSIRO estimates, the rankings of the shares are $20similar in both models. Middle East Sub-Saharan Africa3 Discounting the time stream of investment costs would lower and North Africa the net present value of total investment or adaptation costs, but Europe and Latin America would not influence the choice of investments or the underlying Central Asia and Caribbean investment costs. South Asia East Asia and Pacific5 World Bank. 2010. The Costs to Developing Countries of Adapt- ing to Climate Change. http://beta.worldbank.org/content/ economics-adaptation-climate-change-study-homepage. Source: (World Bank 2009)
G h a n a CO U N T RY ST U DY 7 Table 1 Total Annual Costs of Adaptation for All Sectors by Region, 2010–50 ($ billions at 2005 prices, no discounting) Cost Middle East aggregation East Asia Europe and Latin America and North Sub-Saharan type and Pacific Central Asia and Caribbean Africa South Asia Africa Total National Centre for Atmospheric Research (NCAR), wettest scenario Gross sum 28.7 10.5 22.5 4.1 17.1 18.9 101.8 X-sum 25.0 9.4 21.5 3.0 12.6 18.1 89.6 Net sum 25.0 9.3 21.5 3.0 12.6 18.1 89.5 Commonwealth Scientific and Industrial Research Organization (CSIRO), driest scenario Gross sum 21.8 6.5 18.8 3.7 19.4 18.1 88.3 X-sum 19.6 5.6 16.9 3.0 15.6 16.9 77.6 Net sum 19.5 5.2 16.8 2.9 15.5 16.9 76.8 Source: (World Bank 2010a) Table 2 Total Annual Costs of Adaptation for all Sectors by Region and Period, 2010–50 (X-sums, $ billions at 2005 prices, no discounting) Middle East East Asia Europe and Latin America and North Sub-Saharan Period and Pacific Central Asia and Caribbean Africa South Asia Africa Total National Centre for Atmospheric Research (NCAR), wettest scenario 2010–19 22.7 6.5 18.9 1.9 10.1 12.8 72.9 2020–29 26.7 7.8 22.7 2.0 12.7 17.2 89.1 2030–39 23.3 10.8 20.7 3.0 13.5 19.2 90.5 2040–49 27.3 12.7 23.7 5.0 14.3 23.2 106.2 Commonwealth Scientific and Industrial Research Organization (CSIRO), driest scenario 2010–19 16.4 3.9 11.6 2.4 11.9 10.3 56.5 2020–29 20.1 4.7 13.1 2.6 17.5 13.3 71.3 2030–39 20.9 6.4 20.2 3.0 17.7 20.0 88.2 2040–49 21.0 7.6 22.8 3.9 15.3 24.1 94.7 Source: (World Bank 2010a).
8 E C O N O M I C S O F A D A P T AT I O N T O C L I M A T E C HAN G E Table 3 A Comparison of Adaptation Cost Estimates ($ billions) World Bank Economics of Adaptation to Climate Change (EACC) Study UNFCCC Parry et al. NCAR CSIRO Sector (2007) (2009) (wettest Scenario) (driest scenario) Infrastructure 2–41 18–104 29.5 13.5 Coastal zones 5 15 30.1 29.6 Water supply and flood 9 9 13.7 19.2 protection Agriculture, forestry, 7 7 7.6 7.3 fisheries Human health 5 5 2 1.6 Extreme weather events — — 6.7 6.5 Total 28–67 — 89.6 77.7 Source: (World Bank 2010a).Such a trend is to be expected as, under a busi- effects and refinements in the cost estimations,ness-as-usual (BAU) scenario, rising emissions adaptation costs tend to lie in the upper ranges ofresult in more than proportional environmental the UNFCCC estimates. In the area of coastalimpacts. Another important finding (not shown zone management and defense, the EACC esti-here) is that adaptation costs decline as a percent- mates actually represent a six-fold increase com-age of GDP over time. This suggests that coun- pared to the UNFCCC estimates.4tries become less vulnerable to climate change astheir economies grow if the countries considered The only area where the EACC estimates areadaptations to climate changes in their strategic lower is in human health; the UNFCCC studyplanning processes. Development enhances projects a cost of $5 billion per annum, whereashouseholds’ capacity to adapt by increasing levels the EACC projects $2 billion (NCAR model) andof incomes, health, and education. $1.6 billion (CSIRO model). This difference is partly explained by the inclusion of the develop-The study results indicate that there are consid- ment baseline in the EACC study, which reduceserable regional variations in the share of adapta- the number of additional cases of malaria, andtion costs as a percentage of GDP. The share is thereby adaptation costs, by some 50 percent byhighest in Sub-Saharan Africa, in large part 2030. With the exception of coastal zones, thebecause GDP is lower in the region. Percentages Parry et al. (2009) adaptation costs are muchremain stable in Europe and Central Asia and higher than the EACC study. Their estimate forthe Middle East and North Africa, and fall infrastructure, for example, ranges from $18 tosharply in all other regions. $104 billion per annum. They come up with higher estimates because they argue that low- andTable 3 compares adaptation costs derived fromthe EACC study with those of UNFCCC (2007) 4 This difference reflects the effects of the following refinements:and Parry et al. (2009). Given that the EACC better unit cost estimates, including maintenance costs, and the inclusion of the costs of port upgrading and risks from both sea-study uses a more comprehensive coverage of level rise and storm surges.
G h a n a CO U N T RY ST U DY 9middle-income countries have a large infrastruc- ability of governments to provide assistance.ture deficit and that the costs of climate-proofing Also, by its very nature, economic developmentthis additional infrastructure must be included in tends to shift resources away from agriculture,the adaptation cost. which is the most climate-sensitive sector, into less climate-sensitive areas such as services andFor Sub-Saharan Africa, as well as other devel- manufacturing.oping regions such as South Asia and East Asiaand the Pacific, the study results highlight a The global track study provides policy makersnumber of salient issues. First, for these regions with an indication of global adaptation costs.as a whole, the results indicate that adaptation to However, modeling of the climate scenarios andclimate change will be costly to implement and the climate change impacts are at a relatively highwould subject national budgets to further strain. degree of aggregation. It is highly likely that whenSecondly, given that the effects of climate change the models are downscaled to the country/localare already being felt in these regions, failure to level, the nature and pattern of the effects mighttake immediate action would even be costlier in be entirely different from those obtained at thethe future as the effects are bound to escalate regional level. For that reason, country-level stud-over time. Thirdly, economic development plays ies such as the Ghana EACC study are necessarya key role in enhancing adaptive capacity. By to complement the global track study.increasing levels of incomes, health, and educa-tion, economic development enhances thecapacity of households to adapt; and by improv- Overall Approach and Keying institutional infrastructure, it enhances the Assumptions
G h a n a CO U N T RY ST U DY 11MethodologyThe overall approach adopted in the study follows it is assumed that policy makers know what theclosely on the method used in the global track future climate will be and act to prevent its damages.study. Using a 2050 time frame, development base- Second, only four climate models (described below)lines are first developed for each sector. The base- are used in the Ghana case study; it is implicitlyline represents the growth path the economy would assumed that they cover the breadth of climatefollow in the absence of climate change. It is a rea- change impacts. Third, in costing the adaptationsonable trajectory for growth and structural change options, the study focuses on “hard options”—suchof the Ghanaian economy over a period of 40 as building dams and dikes—and ignores “soft”years that can be used as a basis of comparison options such as early warning systems, communitywith the climate change scenario. The baselines for preparedness programs, watershed management,each sector utilize a common set of GDP and pop- and urban and rural zoning. This approach wasulation forecasts for 2010–50. From the baselines, deliberately chosen because the former options aresector-level performance indicators—such as the easier to value and cost; it does not mean that thestock of infrastructure assets, level of nutrition, and latter are less important. Fourth, the adaptation costswater supply availability—are determined. Next, are based on current knowledge. This implicitlyGCM projections of climate change are used to assumes that there will be no innovation and techni-predict changes in various variables, including cal change in the future. However, we know thatagricultural output, consumption, water availabil- economic growth and hence development dependsity, and infrastructure such as roads and ports. The on technical change, which is likely to reduce thefinal steps involve identifying and costing adapta- real costs of adaptation over time. The only casetion options for the key economic sectors — agri- where technical change is considered is in the agri-culture, road transport, water and energy, and the cultural sector, where growth in total factor produc-coastal zone. For all sectors, the adaptation costs tivity is built into the model, and explicit investmentinclude the costs of planned, public policy adapta- in research is included in the costs. (We consider thetion measures and exclude the costs of private possible effects of these assumptions in the discus-(autonomous) adaptation. sion of the study’s limitations below.)Given the complexity of climate change and thenumber of variables and actors involved in the Climate Forecastsimpacts, a number of simplifying assumptions havebeen made in order to facilitate the modeling. First,
12 E C O N O M I C S O F A D A P T AT I O N T O C L I M A T E C HAN G EHistoric and future climate inputs specific to Ghana climate moisture index.and its river basins—such as monthly temperatureand precipitation—were used to drive the river In line with the global track, the climate projec-basin and water resource model and crop models tions from these two GCMs are used to generateoutlined below. Historic inputs were obtained using the “Global Wet” and “Global Dry” scenarios forthe University of East Anglia’s Climate Research the Ghana country-track study. In addition, theUnit’s global monthly precipitation and tempera- climate projections from the two GCM/SRESture data. Future inputs were taken from four combinations with the lowest and highest climateGCMs forced with different CO2 emission scenar- moisture index for Ghana are used to generate aios to represent the total possible variability in pre- “Ghana Dry” and a “Ghana Wet” scenario. Incipitation. In line with the approach taken in the the case of Ghana, the globally “wettest” GCMglobal track study, climate projections from the actually projects a drier future climate for GhanaNCAR and CSIRO models were used to generate than the globally “driest” GCM under emissionthe “Global Wet” and “Global Dry” scenarios for scenario A2.the Ghana case study. Four climate change scenarios are selected to rep-In the EACC global track study, the National resent the largest possible ranges of changes inCenter for Atmospheric Research (NCAR) temperature, precipitation, and water runoffs.CCSM3 and Commonwealth Scientific and The climate moisture index (CMI) is used as a cri-Industrial Research Organization (CSIRO) terion to select the Ghana climate change scenar-Mk3.0 models with SRES A2 emission forces ios. The index is a measure of the water balancewere used to model climate change for the analy- of an area in terms of changes in precipitation (P)sis of most sectors because they capture a full and losses of potential evapotranspiration (PET).spread of model predictions to represent inherent The moisture index (CMI) is calculated as CMI =uncertainty. In addition, they report specific cli- 100(P - PET)PET. The MI range in the variousmate variables—minimum and maximum tem- GCM scenarios is 115 percent—from -66 percentperature changes—needed for sector analyses. in the Ghana dry scenario to 49 percent in theThough the model predictions do not diverge Ghana wet scenario (Table 4).much for projected temperature increases by 2050(both projecting increases of approximately 2oC Precipitation and temperature data obtained fromabove preindustrial levels), they vary substantially these simulations were used to estimate the avail-for precipitation changes. Among the models ability of water at a subbasin scale. Historical cli-reporting minimum and maximum temperature mate data for each basin were gathered usingchanges, the NCAR was the wettest and the available precipitation and temperature dataCSIRO the driest scenario globally, based on the when available, along with the Climate Research Table 4 GCM Scenarios for Ghana Country Track Study Scenario GCM SRES CMI Deviation (%) Global Wet ncar_ccsm3_0 A2 -17 Global Dry csiro_mk3_0 A2 9 Ghana Wet ncar_pcm1 A1b 49 Ghana Dry ipsl_cm4 B1 -66 Source: Strzepek and Mccluskey (2010)
G h a n a CO U N T RY ST U DY 13Unit’s 0.5° by 0.5° global historical precipitation modified Hargreaves method was used. Actualand temperature database. evapotranspiration is a function of potential evapotranspiration and soil moisture state (follow-CLIRUN-II is used in this study to forecast runoffs ing the FAO method). Soil water is modeled as ain Ghana. CLIRUN-II is the latest model in a two-layer system: a soil layer and a groundwaterfamily of hydrologic models developed specifically layer. These two components correspond to afor the analysis of the impact of climate change quick and slow runoff response to effectiveon runoff. Kaczmarek (1993) presents the theo- precipitation.retical development for a single-layer lumpedwatershed rainfall runoff model-CLIRUN. Kacz- The soil layer generates runoff in two ways. Firstmarek (1996) presents the application of CLIRUN there is a direct runoff component, which is theto Warta River catchment, Poland. Another cor- portion of the effective precipitation (precipita-nerstone publication on the family of hydrologic tion plus snowmelt) that directly enters the streammodels and water balance components is pre- systems. The remaining effective precipitation issented in Cohen et al. (1999). CLIRUN-II (Strze- infiltration to the soil layer. The direct runoff is apek et al. 2008) is the latest in the “Kaczmarek function of the soil surface and modeled differ-School” of hydrologic models applied to the anal- ently for frozen soil and non-frozen soil. The infil-ysis of water flow and economic impacts of the tration then enters the soil layer. A nonlinear setHigh Dam in Egypt. It incorporates most of the of equations determines how much water leavesfeatures of the water balance module WATBAL the soil as runoff, how much is percolated to theand CLIRUN, but was developed specifically to groundwater, and how much goes into soil stor-address extreme events at the annual level, model- age. The runoff is a linear relation of soil watering low and high flows. CLIRUN and WATBAL storage and percolation is a nonlinear relation-did very well in modeling mean monthly and ship of both soil and groundwater storages. Theannual runoff, important for water supply studies, groundwater receives percolation from the soilbut was not able to accurately model the tails of layer, and runoff is generated as a linear functionrunoff distribution. CLIRUN-II has adopted a of groundwater storage.two-layer approach following the framework ofthe SIXPAR hydrologic model (Gupta and Soil water processes have six parameters simi-Sorooshian 1985) and a unique conditional lar to the SIXPAR model (Gupta and Sorooshianparameter estimation procedure was used. In the 1983) that are determined via the calibrationfollowing section a brief description of the com- of each watershed. When CLIRUN-II is cali-ponents of the model will be presented. brated in a classical rainfall-runoff framework, the results are very good for the 25th to 75thCLIRUN-II models runoff as a lumped water- percentile of the observed streamflows, produc-shed with climate inputs and soil characteristics ing an R2 value of 0.3 to 0.7 However, for mostaveraged over the watershed, simulating runoff at water resource systems, the tails of the stream-a gauged location at the mouth of the catchment. flow distribution are important for design andCLIRUN can run on a daily or monthly time operation planning. To address these issues, astep. In the CLIRUN-II system, water enters via concept know as localized polynomial—devel-precipitation and leaves via evapotranspiration oped by Block and Rajagopalan (2008) forand runoff. The difference between inflow hydrologic modeling of the Nile River—wasand outflow is reflected as change in storage extended to calibration of rainfall runoff mod-in the soil or groundwater. A suite of potential eling in CLIRUN-II (Strzepek et al. 2008).evapotranspiration models are available for use in When calibrating, each observed year is catego-CLIRUN-II. For this study, the rized as to whether it falls into a dry year (0–25
14 E C O N O M I C S O F A D A P T AT I O N T O C L I M A T E C HAN G E Figure 2 Flow Chart of Model Sequencing Location GCM GENERAL CIRCULATION MODEL TEMPERATURE PRECIPITATION Surface Slope CliRun CLIMATE RUNOFF TEMPERATURE PRECIPITATION TEMPERATURE RAINFALL RUNOFF Soil Composition Reserve Specifications Crop Type Discount Rate IMPEND INVESTMENT MODEL FOR CliCrop PLANNING ETHIOPIAN CLIMATE CROP AND NILE DEVELOPMENT WATER RESOURCE ALLOCATIONS IRRIGATION DEMAND CROP YIELD Reservoir Specifications River Basin Management Municipal and Industrial Demand WEAP WATER EVALUATION AND PLANNING RESOURCE ACCOUNTING Discount Rate CGE COMPUTABLE GENERAL EQUILIBRIUMpercent of the distribution), a normal year (25– data when available, along with the Climate75 percent), or a wet year (greater than 75 per- Research Unit’s 0.5° by 0.5° global historicalcent). Separate model parameters were estimated precipitation and temperature database. CLI-for the three different classes of annual stream- RUN-II is a two-layer, one-dimensional infiltra-flow. The Climate Research Unit (CRU) and tion and runoff estimation tool that uses historicGlobal Runoff Data Center (GRDC) are the surfaces. A 0.5° by 0.5° historic global surfacetwo major data sources for the CLIRUN-I. Pre- flow database generated by the Global Runoffcipitation and temperature data obtained for the Data Center (GRDC) is used for modeling theCLIRUN-II simulations were used to estimate surface flow, as explained above.the availability of water at a subbasin scale. His-torical climate data for each basin were gatheredusing available precipitation and temperature
G h a n a CO U N T RY ST U DY 15Sector-Specific Approaches shocks simultaneously on all sectors of the economy. Third, CGE models are able to take into consider- ation secondary or feedback effects caused by aThe modeling of the impacts of climate change given shock, and are therefore suitable for analyzingin the selected sectors was carried out using a climate-related issues.5suite of models (CLIRUN, CLICROP, IMPEND,WEAP, DIVA) that are briefly described below. Assumptions about the behavior of economicFigure 2 depicts the modeling process, starting agents in the CGE model are grounded in eco-with the climate forecasts. Climate data from the nomic theory and the magnitudes of some modelGCMs are entered into CLIRUN and CLICROP parameters are determined by resort to second-in order to produce streamflow runoff estimates ary econometric studies. Producers maximizeand crop irrigation demand estimates, respec- profits (and thus minimize costs) under constanttively. Inflows calculated using CLIRUN are then returns to scale and consumers maximize utilityfed into IMPEND, where storage capacity and subject to their budget constraints. It wasirrigation flows are optimized to maximize net assumed that the economy is perfectly competi-benefits. The outputs from IMPEND along with tive and that markets clear. The CGE model wasthe irrigation demands estimated from CLICROP calibrated to a regional 2005 social accountingare then entered into the Water Evaluation and matrix (SAM) of Ghana jointly constructed byPlanning System (WEAP), where water storage the International Food Policy Research Instituteand hydropower potential are modeled based on and the Ghana Statistical Service (GSS) usingtheir interaction with the climate and socioeco- national accounts, trade and tax data, andnomics of the river basins. household income and expenditure survey data. Further details on the features of the GhanaFinally, this information is fed into a dynamic com- CGE model are provided in Annex 6.putable general equilibrium (CGE) model wherethe economic implications of the modeled data are The CGE modeling approach captures threeassessed. Within the river basin model there is, main mechanisms by which climate change ishowever, one interaction with the potential for expected to influence Ghana’s economic growthnonlinearity. The interaction between IMPEND and development. First, it estimates the economy-and WEAP is an iterative process depending on wide impacts of productivity changes in dry-landthe scenario. Reservoir flow calculated in WEAP agriculture, using the CLICROP inputs. Second,may change previous inputs into IMPEND, thus it incorporates the fluctuations in hydropowerrequiring the net benefits to be re-calculated and production due to variation in river flow. Rivertheir implications re-modeled in WEAP. flow will only affect agricultural production if the irrigated area available for planting is greaterThe CGE modeling approach was chosen to model than the maximum potential area that could bethe impacts of climate change because it has a num- irrigated given water availability constraints.ber of features that make it attractive for analyzing Third, it will account for changes in temperaturesuch issues. First, these models portray the function- and precipitation, which in turn influence main-ing of a market economy, including markets for tenance requirements for infrastructure, particu-labor, capital, and commodities, and account for the larly roads. Rainfall or temperature realizationsrole of relative prices and market mechanisms in thedecisions of economic agents. Second, CGE models 5 An alternative approach is to use partial equilibrium (i.e.belong to the class of general equilibrium models econometric) models, which are limited in the sense that they can consider the impact of only one variable at a time in a singlethat are able to determine the impacts of exogenous sector.
16 E C O N O M I C S O F A D A P T AT I O N T O C L I M A T E C HAN G Eoutside of the band of design tolerances are likely (2) cocoa, (3) forestry and logging, and (4) fish-to require more frequent or more expensive main- ing. Agriculture contributes to 40 percent of realtenance costs. In the CGE model, these greater GDP, with the cocoa sector accounting for 32 per-maintenance requirements result in either less cent of exports. Overall, over 50 percent of therapid expansion in the road network for a given population derives their livelihood from agricul-level of spending on roads, or an actual shrinkage ture. Growth in the sector has been variable in thein the network if the resources necessary to main- past few years. Starting from a low of 4.4 percenttain the network are unavailable. in 2002, the sector’s growth rate rose to a high of 7 percent in 2004 before declining to anotherWe now turn to the specific approaches used to low of 3.1 percent in 2007 (Figure 3). The growthmeasure the impacts of climate change in the decline in 2007 was due to drought, particularlyselected sectors—agriculture, road transport, in the forest zone where cocoa is cultivated. Thewater and energy, and coastal zone. For each 2009 budget projected growth of 5.3 percent insector, we briefly describe the sector’s contribu- 2009 and 5.9 percent in 2010.tion to the economy, its vulnerability to climatechange, the baseline (BAU) scenario, and the Vulnerability to Climate Change. Across Ghana’smethodology used. agroecological zones, there are some significant differences in the regional distribution of agri-Agriculture cultural GDP. The forest zone accounts for 43Contribution to the Economy. The Ghanaian economy, percent of agricultural GDP, compared to aboutlike that of most developing countries, is based on 10 percent in the coastal zone, and 26.5 and 20.5agriculture. The agricultural sector is composed percent in the southern and northern savannahof four subsectors: (1) food crops and livestock, zones, respectively. The northern savannah zone Figure 3 Trends in Agricultural Growth 2002 to 2010 35.0 30.0 25.0 20.0GROWTH RATE (% P.A.) 15.0 10.0 5.0 0.0 2002 2003 2004 2005 2006 2007 2008 2009 2010 -5.0 -10.0 AGRICULTURE CROPS AND LIVESTOCK COCOA FORESTRY AND LOGGING Source: (World Bank 2009)
G h a n a CO U N T RY ST U DY 17is the main producer of cereals, accounting for The plan has been developed using a largely par-more than 70 percent of the country’s sorghum, ticipatory process and based on food and agricul-millet, cowpeas, groundnuts, beef and soybeans. ture development policy II (FASDEP II) objectives,On the other hand, the forest zone supplies a large with a target for agricultural GDP growth of atshare of higher-value products such as cocoa and least 6 percent annually and government expen-livestock (mainly commercial poultry) (Breisinger diture allocation of at least 10 percent within theet al. 2008). Ghana’s agricultural sector is highly plan period. These targets are in conformity withvulnerable to climate change and variability agricultural performance targets of the country’sbecause it is predominantly rainfed and is charac- National Development Planning Commissionterized by low levels of productivity. (NDPC) and other relevant government develop- ment policies. Ghana’s agriculture and irrigationBaseline. The current development strategy for policies are expected to contribute significantly toagriculture is to ensure sustainable utilization the achievement of the MDGs.of resources and commercialization of activitieswith market-driven growth. Commodity target- Irrigation in Ghana contributes only about 0.5ing for food security and income diversification percent of the country’s agricultural production.of resource-poor farmers is given a high priority. About 11,000 hectares (out of a potential irrigableThe strategy seeks to enhance the commodity area of 500,000 hectares) have been developed forvalue chain using science and technology. There irrigation, and even the developed area is largelyis also an emphasis on environmental sustain- underutilized due to institutional, management,ability and greater engagement with the private input, and other constraints. The investment plansector and other partners (GoG/NDPC 2009). concluded that: “It is necessary that the Govern-As stated in the Ghana Poverty Reduction Strat- ment regards irrigated agricultural infrastructureegy (GPRS, GoG 2003), Ghana’s agricultural as a public good, which can be leased to waterdevelopment strategy is to ensure a modernized users’ associations and/or private managementagriculture culminating in a structurally trans- bodies to ensure efficiency through better manage-formed economy that will provide food security, ment practices.” METASIP estimated an irriga-employment opportunities, and reduced poverty tion funding gap of $423 million in 2009, rising toin line with the goal set for the sector in GPRS about $1.6 billion in 2015 (GoG 2009). METASIPI. The strategy emphasizes the sustainable utili- noted that climate change— which has had a sig-zation of all resources and commercialization of nificant adverse impact on the nation’s agricultureactivities in the sector based on market-driven over the years—added uncertainties to the agricul-growth. Climate change impacts and national ture sector. The report also said that even thoughplans to deal with these changes are not explicitly irrigated agriculture is well-known to be important,stated in national and agricultural sector goals, it is yet to be significant in Ghana.although there is provision for irrigation develop-ment in various parts of the country. The policy Methodology. As indicated earlier, the impact ofdocument emphasizes that small- and large-scale climate change on the agricultural sector wasirrigation systems and efficient water harvesting estimated using CLICROP. CLICROP is aand management systems are required to reduce generic crop model used to calculate the effectreliance on rainfed agriculture (GoG 2003--). of changing daily precipitation patterns caused by increased CO2 on crop yields and irrigationRecently the government of Ghana issued vol- water demand. It was developed in response toume 1 of the Medium Term Agriculture Sector Invest- the available crop models that use monthly aver-ment Plan (METASIP) 2009–2015 (GoG 2009). age rainfall and temperature to produce crop
18 E C O N O M I C S O F A D A P T AT I O N T O C L I M A T E C HAN G Eoutputs. These monthly models do not capture and the fraction already under irrigation; irriga-the effects of changes in precipitation patterns, tion investment and maintenance cost per ha ofwhich greatly impact crop production. For exam- irrigated land; and the current level of provisionple, most of the IPCC GCMs predict that total of extension services. These pieces of informa-annual precipitation will decrease in Africa, but tion were then fed as inputs into the CGE modelrain will be more intense and therefore less fre- as shocks/stressors caused by the predictedquent. Currently, CLICROP is able to produce weather changes from the GCMs. The modelpredicted changes in crop yields due to climate then computes the values of the key economicchange for both rainfed and irrigated agriculture, variables based on the response of economicas well as changes in irrigation demand. agents to these climate-related shocks. A detailed description of the CLICROP methodology isFive yield estimates (one for each of the four presented in Annex 1.development stages, and one for the whole sea-son) were computed using Equation 1. A specific module on the impact of climate change on livestock productivity was created for this study. To model the effect of climate on live- [1 – Y ] = K [1 – ETC ] Y d d Equation 1: a m y ETA d stock, this analysis relies on the approach and results of a structural Ricardian model of Afri- Where Ya = predicted actual yield can livestock developed by Seo and Mendelsohn Ym = maximum yield (2006). This approach measures the interaction Ya / Ym = % Yield d d between climate and livestock and considers the Ky = yield coefficient, different for development stage d to y adaptive responses of farmers by evaluating ETCd = sum of daily ET crop demand for which species are selected, the number of ani- development stage d mals per farm, and the net revenue per animal ETAd = sum of daily actual ET for under changes in climate. The current analysis development stage d transfers the findings from Seo and Mendelsohn %Yieldd = ratio of actual yield over maximum yield, value reported by to the Ghana-specific context. Seo and Mendel- CLICROP sohn rely on a survey of over 5,000 livestock farmers in ten African countries. In this data set,The inputs into CLICROP include weather the variation in livestock productivity and(temperature and precipitation), soil parameters expected incomes in different regions demon-(field capacity, wilting point, saturated hydraulic strates a clear relationship to regional climate,conductivity, and saturation capacity), historic which provides a mechanism—through spatialyields for each crop by ecological zone, crop dis- analogue—to statistically analyze how climatetribution by ecological zone, and current irriga- change may affect livestock incomes acrosstion distribution estimates by crop. These were Africa. The authors develop a three-equationused to compute estimates for changes in annual farm-level model. The first equation predicts theproduction (yield) for both irrigated and rainfed probability of selecting each livestock type as thecrops as well as irrigation demand (mm/ha) for primary animal for the farm, the second predictsthree industrial crops and four food crops (See the net income of each animal, and the finalAnnex 1). The estimated yields reflect the reduc- equation predicts the number of animals ontions in yield both due to the lack of available each farm. Farm net revenues are the sum prod-water and due to the overabundance of water uct of these three outputs; that is, the probabilitythat causes waterlogging. Additional data of selecting each type of animal multiplied byobtained included total area of irrigable land the number of animals and then the expected