EACC Mozambique

d i s c U s s i o n pA p E R n U m B E R 1 6 d E c E m B E R 2 0 1 0E c o n o m i c s o f A d A p t A t i o n t o c l i m A t E c h A n g EThe Social Dimensions of Adaptationto Climate Change in Mozambiqueprepared by Anthony Patt(international institute for Applied systems Analysis),with Raul Varela (independent consultant),Isilda Nhantumbo (independent consultant), andLivia Bizikova (international institute for sustainable development)Papers in this series are not formal publications of the World Bank. They are circulated to encourage thought and discussion. The use andcitation of this paper should take this into account. The views expressed are those of the authors and should not be attributed to the WorldBank. Copies are available from the Environment Department of the World Bank by calling 202-473-3641.

© 2010 The International Bank for Reconstruction and Development/The World Bank1818 H St., NWWashington, DC 20433 U.S.A.Telephone: 202-473-1000Internet: www.worldbank.org/climatechangeE-mail: feedback@worldbank.orgAll rights reserved.December 2010This paper is based upon work that has been commissioned by the World Bank as part of the Economics of Adaptation to Climate Change study. Theresults reported in the paper are preliminary and subject to revision. The analysis, results, and views expressed in the paper are those of the authors aloneand do not represent the position of the World Bank or any of its member countries.Papers in this series are not formal publications of the World Bank. They are circulated to encourage thought and discussion. Copies are available fromthe World Bank Environment Department by calling 202-473-3641.The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shownor any map in this work do not imply any judgment on the part of the World Bank concerning the legal status of any territory or the endorsement oracceptances of such boundaries.R I G H T S A N D P E R M I S S I O N SThe material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicablelaw. The International Bank for Reconstruction and Development/The World Bank encourages dissemination of its work and will normally grantpermission 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 978-750-4470; Internet: www.copyright.comCover photo courtesy of Jonathan Ernst/World Bank.

iiiTable of ConTenTsAcRonYms And ABBREViAtionsAcknowlEdgmEntsviiviii1. intRodUction And oVERViEw 1The EACC study 1The Process of the Study 3The Outline of this Report 32. sociAl dimEnsions of climAtE chAngE in mozAmBiqUE 5Expected Physical Impacts of Climate Change in Mozambique 5Current Hazards 5How the Hazards Will Change Because of Climate Change. 7National Planning for Climate Change: NAPA and Other Strategies 8Overview of Social Vulnerability in Mozambique 11Bottom-up Research Studies. 11Studies Using the Household Food Economy Approach 14Top-down Approaches 153. REsEARch mEthodologY 17Research Strategy and Questions 17Site Selection and Sampling 174. pARticipAtoRY scEnARio dEVElopmEnt woRkshop REsUlts 21PSD Workshops: Design Overview 21Overview of Results from Local and National Workshops 22Xai Xai Key Insights. 22Beira Key Insights. 24Maputo Key Insights. 25Synthesis. 25

viiaCRonYMs anD abbReVIaTIonsCCGC Coordination Council for Disaster ManagementCRED Center for Research on the Epidemiology of DisastersEACC Economics of Adaptation to Climate ChangeFEWS-NET Famine Early Warning System NetworkGEF Global Environment FacilityHEA Household Economy ApproachIISD International Institute for Sustainable DevelopmentIPCC Intergovernmental Panel on Climate ChangeNAPA National Adaptation Program of ActionINGC National Disaster Management AgencyMDGs Millennium Development GoalsMICOA Ministry for the Coordination of Environmental AffairsNAPA National Adaptation Program of ActionPSD Participatory Scenario DevelopmentSRES Special Report on Emissions ScenariosUNFCCC United Nations Framework Convention on Climate Change

viiiaCknowleDgMenTsThe social component of the Economics of Adaptationto Climate Change (EACC) study was developedthrough the joint efforts of the World Bank SocialDevelopment (SDV) and Environment Departments(ENV) and the LCR, AFR, EAP, and SAR Regions ofthe World Bank; ESSA Technologies Ltd and theInternational Institute for Sustainable Development(IISD), Canada; and research institutions in Bolivia,Bangladesh, Ethiopia, Ghana, Mozambique, andVietnam. The study was undertaken by a core team ledby Robin Mearns and comprising Anne Kuriakose,Carina Bachofen, Nilufar Ahmad, and Minna Kononen(all SDV). An overall synthesis report of all six countryreports has been prepared and is available at www.worldbank.org/eacc. The study was designed and implementedin close collaboration with the team leader forthe overall EACC study (Sergio Margulis), and AnaBucher, Laurent Cretegny, Urvashi Narain, KiranPandey, Robert Schneider (all ENV), and sectorconsultants.The Mozambique country study for the social componentwas led by Carina Bachofen (SDV), under theoverall guidance of Robin Mearns, in collaboration withthe EACC country team leader Jean-Christophe Carret(AFTEN). Research partners in Mozambique wereAnthony Patt of IIASA, who led the study teamcomprising Raul Varela and Isilda Nhantumbo. Supportin methodology and workshop design and for trainingof trainers for the Participatory Scenario Development(PSD) workshops was provided by Livia Bizikova(IISD), Philip Bailey, and Samantha Boardley (allESSA). A separate report is available summarizing thisPSD work, along with a training of trainers manual atwww.worldbank.org/eacc.We could not have completed this work without thecontinuous logistical support provided by HawantyPage (ENV), Grace Aguilar (ENV), MustafaPajazetovic, and Carmen Martinel (both SDV). Wegratefully acknowledge editorial services provided byRobert Livernash, and production managementprovided by Danielle Christophe (SDV) and RobertReinecke.The Mozambique country study has benefited greatlyfrom peer review comments and other feedback fromWorld Bank staff, including Carolyn Winter (AFTCS),and from external peer reviewers, including KatrinaBrown and Richard Klein.Any remaining errors and omissions are entirely thoseof the country study team.

11. InTRoDuCTIon anD oVeRVIewThe eaCC sTuDYThe purpose of the Economics of Adaptation to ClimateChange (EACC) study is to better understand and estimatethe true costs of adapting to climate change in lessdeveloped countries. The study is made up of threecomponents. At a global level, there is an analysis of costsacross different economic sectors. At a country level,there is an economic component and a social component,taking place in a set of representative case study countries.Mozambique is one of these countries.The economic component of the Mozambique countrystudy has the objective of identifying a set of robustadaptation options for the country, then comparing thedirect costs and benefits of those options. To calculatethe costs, the team has utilized a computable generalequilibrium method. This method is data intensive—itrequires a good model of the national economy—butcan generate an estimate of the costs of targetedgovernment interventions, in terms of reduced overalleconomic growth, once those effects have trickledthrough the labor and capital markets and the economyhas returned to equilibrium. The economic teamconsidered a range of adaptation options, which weregathered from the literature and from interactions withnational level policy makers and other stakeholders.The social component of the Mozambique countrystudy has a less precisely defined purpose, and thisreflects the relative ambiguity of the term “social vulnerability”in the first place. Crudely stated, the concept ofsocial vulnerability rests on the premise that the extentto which climate change harms people depends on abroad set of factors having to do with individualempowerment to weather storms and to make changes,and the extent to which social interactions contribute toor detract from that empowerment (Adger 1999; Brookset al. 2005). So one purpose of this study is to find outwhat individual and social factors make people more orless vulnerable, and to identify adaptation strategies thatwould reduce that vulnerability (Cutter 2001). The ideaof social vulnerability also rests on the premise thatwithin any community there are some people who areespecially vulnerable, and a just society should take stepsto help them in particular. So a second purpose of thisstudy of social vulnerability and adaptation is to identifythose livelihood activities and people likely to be mostharmed by climate change, and then identify realisticways of improving their situations (Osbahr et al. 2008).This study builds on a recent assessment of socialvulnerability in Mozambique, conducted as part of astudy led by the National Disaster Management Agency(INGC). Patrick Nussbaumer took a standard socialvulnerability framework from the literature—previouslyused to rank African countries (Vincent 2004; Vincent2007)—that relied on a set of a dozen indicators. Heestimated future trends in these indicators consistentwith two development scenarios—the A2 and B1Special Report on Emissions Scenarios (SRES) of theIntergovernmental Panel on Climate Change(Nakicenovic and Swart 2000)—to see whereMozambique will be in 50 years compared to now andcompared to other African countries behind which itcurrently lags. Figure 1 shows the results. As is evident,the social vulnerability of Mozambique by 2060 looks

2 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U Ebetter than that of South Africa today. That is goodnews for people who think that South Africa is in relativelygood shape. It would suggest, perhaps, that toestimate future climate change impacts on people inMozambique in the future we should think about theeffects on people in South Africa today.That study, however, was relatively silent on the issue ofadaptation. A key issue for this study, then, is to moveforward and reach conclusions about adaptation basedon an appraisal of social vulnerability.In this study, we have taken the deliberate approach ofconsidering development first, and then putting climatechange on top of that, rather than the other way around.This means identifying where the country and its peopleare headed and would like to go, and associating numerousactivities and steps with the achievement of thatprogress. Then, we have looked at where climate changemay interfere with those activities and steps, and hencethreaten the development vision. Adaptations are thingsthat allow the development vision to proceed. The alternativeapproach, which is what has arguably guided mostresearch efforts to date, is to start with the identificationof climate impacts, and see how these might negativelyaffect society as it is structured now. After having doneso, one then sees how future development pathways mayexacerbate or ameliorate this situation.If one could be completely “rational” or “logical” in ouranalysis, the two approaches would end up in the sameplace. But we are not that perfect, and as researchersand as stakeholders tend to become anchored inparticular visions of the world. By considering developmentfirst, we try to free our minds to consider afuture for the country that is quite different than thepresent. Perhaps there will be no subsistence farmersin 50 years, just like subsistence farming vanishedfrom North America and Europe in the 20 th Century.If that is the case, then we free our minds fromconsidering the impacts on subsistence farmers, andinstead consider the impacts of people’s transitionfrom being subsistence farmers into being somethingelse.Another purpose of this study, then, is to inform theprocess of vulnerability assessment, which continues toevolve to suit the needs of policy makers morecompletely (Patt et al. 2008). We are interested in seeingwhether taking a development-first approach is evenpossible: It seems nice when written on paper, but canpeople really think this way when called to work on astudy with climate change in its title? If we can take theapproach, and stick to it, do we in fact arrive at resultsthat are different from the more traditional impacts-ledapproach? This is a crucial issue to evaluate in theconclusion of this report.fIguRe 1. soCIal VulneRabIlITY In MozaMbIque PRojeCTeD To 2060Social Vulnerability Index0.6A20.50.40.3ZimbabweSouth AfricaMozambique2000 2010 2020 2030 2040 2050 2060PovertyUrbanizationDependentpopulationTelephonesRuralpopulationTotal SVISocial Vulnerability Index0.60.50.40.3B1South AfricaMozambiqueZimbabweTotal SVI2000 2010 2020 2030 2040 2050 2060PovertyUrbanizationDependentpopulationTelephonesRuralpopulationSource: Patt and nussbaumer (2009).note: The two graphs show the relative contribution of different factors to the reduction of social vulnerability, relying on a theoretically informed model.

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s3The PRoCess of The sTuDYIn planning this study, the World Bank staff designed afive-step methodology, following an ambitious schedule.Phase 1 would summarize existing literature on differentlivelihood groups, including their relative vulnerabilities.It would then engage in discussions with a widerange of stakeholders to provide details on hotspots ofvulnerability (i.e., by livelihood groups and by location),including a detailed description of that vulnerability andexisting adaptive capacity. Phase 2 would conduct anumber of workshops in order to develop adaptationscenarios. Phase 3 would analyze the social effects ofthe different scenarios. Phase 4 would add to the socialanalysis by incorporating economic analysis, using theresults of computable general equilibrium models toidentify distributions of costs and benefits. Phase 5,taking place contemporaneously, would be the preparationof a final country case study report so the resultscould be evaluated in conjunction with the othercountries.Actual work on the study followed the same generaloutline to a large extent, but the scheduling changed agreat deal, due to unanticipated difficulties finalizingcontracts, scheduling meetings, and resolving paymentissues. The only structural result of the change in schedulingwas that the Phase 2 workshops preceded thestakeholder discussions designed for Phase 1; this hadlittle practical significance. Thus, the first part of Phase1 involved conducting an institutional analysis, an identificationof socio-geographic zones, and a review ofexisting livelihood assessments in those zones. All ofthis was included in an inception report, submitted bythe lead consultant in mid-February 2009. Phase 2consisted of a set of “participatory scenario development”(PSD) workshops, under the guidance of theMozambique country team and in coordination with aconsultant team from Canada, ESSA Technologies, andthe International Institute for Sustainable Development(IISD). ESSA Technologies and IISD organized a trialPSD workshop in Ghana in June 2009, at which finalscheduling decisions were made, namely to hold threeworkshops in Mozambique in July and August.Following these workshops, in September, the countryteam conducted interviews and surveys in a targeted setof field sites, in order to supplement the existing literature.Results from this fieldwork were availablefor analysis in November. Phases 3, 4, and 5 took placeiteratively, commencing in mid-November andcompleted in January 2010. This report is a result ofthat work.The ouTlIne of ThIs RePoRTThis report follows the chronology of the work in thecountry study as that work actually took place.• The first section contains a review of existingknowledge about vulnerability and adaptation inMozambique, including the institutional landscapewithin which adaptation planning has taken place.Much of the material in section 2 has been copiedfrom the inception report.• The second section describes the general approachto conducting original research within this project.This methodology was agreed upon over the courseof several months, after the inception report hadbeen submitted, and concluded with discussions inGhana in June. It includes a map of the locationswhere fieldwork ultimately took place and the guidingquestions for the work.• The third section describes the results from thethree PSD workshops. Results from the first ofthese workshops—which took place in Xai Xai—had already been submitted to the World Bank bythis consultant, while the results from the thirdworkshop—in Maputo—had been submitted byESSA and IISD. The second workshop, in Beira,took place under the direction of the localMozambique project team, and they supplieddetailed notes about that workshop to this consultant.Hence, section 4 pulls these three sets ofresults together into one place, allowing for comparingacross the three workshops.• The fourth section describes the results from thefieldwork, which took place in late August andearly September 2009. That fieldwork consisted ofthree activities: a set of institutional interviews, aset of focus group discussions with representativestakeholder groups, and a household survey. Theresults here are based on the field notes from thefirst two activities, translated from Portuguese intoEnglish by an additional consultant in WashingtonDC, and on a data file containing the results fromthe household survey.

4 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U E• Finally, the fifth section synthesizes the findingsand draws out the key conclusions that are relevantfor policy.There are three annexes. Annex 1 consists of thisconsultant’s identification of the important lessons fromeach of the focus group discussions. These help to painta more complete picture of how participants conceptualizedclimate change vulnerability and adaptation.Annex 2 consists of the household survey instrument. Itis in Portuguese, which is the language in which it waswritten, and from which it has not been translated.Annex 3 consists of the field report from the lead localconsultant, Raul Varela.

52. soCIal DIMensIons of ClIMaTeChange In MozaMbIqueexPeCTeD PhYsICal IMPaCTs ofClIMaTe Change In MozaMbIqueCurrent hazardsAs with many other countries in Africa (Boko et al.2007), Mozambique's vulnerability to climate change isin large part defined by its vulnerability to naturalhazards (República de Moçambique 2007; Ehrhart andTwena 2006; Comité de Conselheiros 2003; Nkomo etal. 2006; and Mavie 2003). The three main hazards thatMozambique faces are droughts, floods, and tropicalcyclones. Figure 2 plots out the numbers of lives lost,and the number of people otherwise affected (needingsome sort of assistance) from these three hazards overthe period 1990–2007. It is taken from the database ofthe Center for Research on the Epidemiology ofDisasters (CRED). Each of the dots represents a singlecountry in Africa, while Mozambique and its immediateneighbors are labeled. As can be seen, Mozambiquehas suffered among the greatest effects from naturaldisasters in Africa.Table 1 lists the most important disasters recorded inMozambique, sorted according to the number of peopleaffected and requiring some sort of assistance duringand after the event. Droughts have affected the mostpeople and caused the most deaths. However, one needsto be cautious about how many people have been killedby droughts; for example, the drought of 1981 isfIguRe 2. nuMbeR of PeoPle affeCTeD oRkIlleD bY naTuRal DIsasTeRs In afRICanCounTRIes, 1990–2007People killed,per millon annually100.1MadagascarMozambiqueTanzaniasouth africazambia0.010.001 0.01 0.1 1 10 100People affected, per thousand annuallyMalawizimbabweswazilandSource: Center for Research on the epidemiology of Disasters, The ofDa/CReD international disaster database.note: Mozambique has suffered among the highest losses of african countriesin both cases.1,000attributed with 100,000 deaths, but it is unclear howmany of these deaths were due in large part to theongoing civil war, which made relief efforts problematic.Floods, while not typically affecting as many people,typically do cause loss of life but do cause even greaterlosses in terms of infrastructure. The flood of 2000, forexample, caused an estimated $419 million worth ofdamage and set back the country’s development byyears. Storms, most prominently tropical cyclones, arethe third major hazard, and also cause a great deal ofinfrastructure loss. Epidemics are a secondary hazardoften brought on by one of the other three, either

6 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U ETable 1. MosT IMPoRTanT DIsasTeR TYPesIn TeRMs of nuMbeRs affeCTeD anDkIlleDfIguRes 3a anD 3b. PReCIPITaTIon anDDRoughT3a.DisasterNumber ofeventsNumberaffectedNumber killeddrought 10 16,444,000 100,200flood 20 9,039,000 1,900storms 17 3,002,400 700Epidemic 18 314,000 2,500Source: IngC.because of a deterioration in safe drinking water provisionor the spread of tropical disease vectors.One of the most important determinants of agriculturalsuitability is rainfall. Figure 3a is taken from theNational Adaptation Program of Action (NAPA, p. 68)and shows average annual precipitation in millimeters ofrainfall. It shows that most of the country receives anaverage of between 600 and 1,200 mm of rainfall annually,which is generally sufficient for maize or cassavacultivation. Some areas—mountainous regions in thenorth—receive more than this, while the low-lyinginland region in the south receives substantially less.3bCorresponding to the variability in rainfall is the risk ofdrought. Figure 3b shows the risk of drought throughoutthe country. The highest risk levels are, not surprisingly,where the average annual rainfall is also thelowest, and hence the chances of receiving insufficientrainfall to support crop growth is quite high. Also highin risk are several districts in the inland central region.These also receive relatively little rainfall.While usually costing fewer lives than droughts, floodsare often the most visible hazard hitting the country.The most notable recent example was in 2000. In earlyFebruary 2000, heavy rains started to fall across much ofsouthern Africa, hitting southern Mozambique thehardest. On February 9 the capital of Mozambique,Maputo, was flooded, with slums in the peri-urban areashardest hit, and the road north to Beira underwater.The rains continued, and on February 11 the LimpopoRiver, north of Maputo, broke its banks, contaminatingSources: República de Moçambique, National Adaptation Program of Action(NAPA) 2007, p. 39, and INGC.note: The hash-marked areas in figure 3a are those where stakeholder interviewswere conducted in support of the naPa, which is not relevant for thisstudy. figure 3b shows the drought risk mapped throughout the country, indicatingthe highest risk levels in the inland south—where average precipitationis also the lowest—as well as the inland central region of the country, whereaverage precipitation is higher.

8 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U Egeneral trend to be expected with climate change is forwetter conditions. This is the case for Kenya andTanzania. Mozambique, of course, borders bothregions, and given uncertainty about where the linebetween the two directions of change will fall, it is verydifficult to say what direction of change any part of thecountry will experience. That being said, it is morelikely that the southern part of the country will experiencea trend toward drier conditions, while the northernpart of the country will experience a trend towardwetter conditions. But there is a great amount ofuncertainty. Another reason for the uncertainty lies inthe paucity of long-time series of historical climatedata for Mozambique and the neighboring countries;this makes it hard to calibrate climate models, andhence to obtain reliable estimates for the future.The best assessment to date of these changes was in arecent report carried out by INGC in 2008 and 2009(Asante et al. 2009). The climate modeling team fromthe University of Cape Town in South Africaconducted a comprehensive downscaling of data fromclimate models across the region, and had access to thefull set of Mozambique weather station data. Thisreport made projections for changes in temperature andaverage precipitation, as primary impacts. As secondaryimpacts, the report made projections for changes in soilmoisture availability (influenced by both temperatureand precipitation) and flooding (influenced by precipitationand mapped out onto a statistical floodingmodel).Figure 5 shows these projected changes that could affectthe risk of drought. It is quite clear that temperaturewill rise. It also seems reasonably clear that most of thecountry—with the exception of the south—will experiencea rise in average precipitation. These two effectsoppose each other in terms of how they influencepotential soil moisture, however, and so it is less clearwhat the combined effects will be. Except in the south,the median estimates are for slight increases in soilmoisture, while in all cases the range of estimates fromthe ensemble of seven models indicates that the directionof change could go either way. What will this meanfor drought risk? It is difficult to say, because droughtrisk depends not just on average rainfall and soil moisture,but also on the frequency with which there aresevere negative deviations. A secondary analysis of theINGC data suggested a slight increase, perhaps as muchas 15 percent, in the frequency of drought, especiallyover the southern part of the country. Again, however,the range of estimates (derived from the range ofmodels used) crossed into the range of a possibledecrease in drought risk.Figure 6 shows results from the INGC flood mappingwork. The team used a statistical flood model—basednot on exact hydrological characteristics and digitalelevation map, but rather on the correlation betweenpast climatic conditions and reported flooding—toestimate how the risk would change in the future. Asthe figure caption indicates, there is a great amount ofuncertainty about the direction of change: Flood riskcould increase in the future, or it could decrease. It isvery difficult to say more. It is also difficult to identifyclear regional trends within the country.Finally, the INGC study attempted to characterize thechanging risk levels due to tropical cyclones. Here, it ispossible to identify the likely direction of change, but dueto a lack of data it is very difficult to quantify thatchange. Both models and empirical data suggest a positivecorrelation between sea surface temperatures andcyclone intensity but no obvious correlation between seasurface temperatures and cyclone frequency. Given thatsea surface temperatures are likely to increase (McDonaldet al. 2005.; Bengtsson et al. 2007; and Emanuel et al.2008), one can be confident that there will be a shifttoward stronger cyclones, but not necessarily any changein their overall frequency. However, since damages arerelated to the cube of the wind speed, strong cyclonescause a much higher amount of damage than do weakercyclones. Overall, one can expect more damages due tocyclones in the future.naTIonal PlannIng foR ClIMaTeChange: naPa anD oTheR sTRaTegIesTo understand the policy and institutional landscapein Mozambique related to climate change vulnerabilityand adaptation, it is essential to pay attention tothe degree of activity that is driven by external fundersand donors on the one hand, and Mozambique’sparticipation in the United Nations FrameworkConvention on Climate Change (UNFCCC) andcommitments to achieve the Millennium

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s9fIguRes 5a–5D. PRojeCTIons MaDe bY The IngC sTuDYSource: asante et al.( 2009).note: each chart shows projected monthly changes from the baseline period (1960–2000) until 2050 (green) and 2100 (blue). The shaded region represents therange captured by an ensemble of the seven climate models, while the solid line represents the median estimate of those models. figure 5a shows projections forrainfall, in terms of mm per day. figure 5b shows changes in maximum temperature, in terms of °C. figure 5c shows changes in potential evapotranspiration, interms of mm per day, and derived from temperature estimates and land cover data. figure 5d shows the potential moisture index (indication soil water availability)derived from the other three estimates.Development Goals (MDGs) on the other. In thissection, we provide an overview of the activities thathave been carried out to date.One of the main drivers of planning activities inMozambique in the area of climate vulnerabilityreductions is the country’s adoption of the MDGs andUNFCCC. In September 2000, Mozambique was one of189 countries to formally adopt the MDGs. As part ofits commitment to cut absolute poverty in half by 2015(MDG 1), the government committed to developing astrategy for attainment, known generally as the Poverty

10 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U EfIguRe 6. Changes In flooD RIskSource: asante et al. 2009.note: Three different global climate models were used, and their results fedinto a statistical flood model. The wettest of the climate models was the IPslmodel, and this resulted in estimates of increased flood risks in most of theriver basins in Mozambique. The driest of the models, gfDl, indicated adecrease in the frequency of flooding events across most river basins. Themedian model, eChaM, indicated a mix, with many areas experiencing aslight increase in risk levels, and some experiencing a slight decrease.Reduction Strategy Papers (PRSPs) and in Mozambiqueby the Portuguese equivalent of PARPA. Preparation andapproval of PRSPs on an ongoing basis is a preconditionfor continued support from the International MonetaryFund (IMF) and the International DevelopmentAssociation (IDA), the branch of the World Bankresponsible for helping the world’s poorest countries; thetwo organizations are involved in funding and approvingthe PARPA and monitoring its implementation. Thegovernment completed the first PARPA in 2001(República de Moçambique 2001–2005) and the second,PARPA II, in 2006 (República de Moçambique 2006–2009). The Ministry of Planning and Finance (MPF)coordinated the preparation of PARPA material, whichoccurred across numerous government ministries andagencies and received assistance from a long list ofnational, international, and nongovernmental donors.IMF and IDA commented on the PARPA, noting that itadvanced planning on poverty in Mozambique significantly,and yet left room for improvement in the area ofvulnerability, and approved it as a satisfactory prerequisitefor continued IMF and IDA support (IMF and IDA2001). A similar approval process occurred for PARPAII, noting that it represented a significant improvementover the original PARPA (IMF and Republic ofMozambique 2007). Related documents suggest that theimplementation of activities under the PARPA isprogressing, but the extent to which this is taking place isunclear.Mozambique ratified the UNFCCC in 1995, and as partof this committed to preparing NationalCommunications, led by the Ministry for theCoordination of Environmental Affairs (MICOA), theUNFCCC focal point for the government ofMozambique. The first national communication wascompleted in 2003 (MICOA 2003), but due to aprotracted period of governmental approval, submittedonly in 2006. The initial national communicationcontains required information on greenhouse gas emissions,mitigation options, climate change vulnerability,and adaptation options. Under the UNFCCC, leastdeveloped countries (LDCs) receive financial assistancefrom the Global Environment Facility (GEF) to prepareNational Adaptation Programs of Action (NAPAs).MICOA began this process in 2003, at the time itcompleted its initial national communication. Itcompleted work on the NAPA in 2007 and submitted itto the UNFCCC in 2008 (República de Moçambique2007). Once the NAPA has been submitted to theUNFCCC, the country becomes eligible to draw fromthe LDC fund, also managed by the GEF, to implementthe planned activities. It is unclear whether any stepshave been taken toward implementation of the NAPA.The NAPA represents the main area of planning forclimate adaptation within the country (República deMoçambique 2007). Its conclusions are entirely consistentwith a view of Mozambique’s adaptation needs beingdriven be the three hazards of drought, flood, and cyclone.The four strategic options listed in the NAPA were:

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s11• Strengthening the early warning system for floods,droughts, and tropical cyclones;• Strengthening the capacities of agricultural producersto cope with climate change;• Reduction of climate change impacts in coastalzones; and• Management of water resources under climatechange.There has also been nationally driven planning. In 1999,the government adopted the National DisasterManagement Policy. This creates a set of priorities andobjectives, including better coordination between disasterplanning and other government objectives, such aseconomic development. The policy established theNational Institute for Disaster Management (INGC)and the Coordination Council for DisasterManagement (CCGC). INGC is an independent legalentity, originally located within the Ministry of ForeignAffairs but since 2005 in the Ministry of StateAdministration. Its general competencies are to coordinatedisaster management planning, information, andactivities; to deal with donors in the area of disastermitigation, planning, or management; and to increasethe flow of information, both with neighboring countriesand with the Mozambique public, related to disastermanagement. It has offices in each of the provinces,with the exception of the Maputo City Province. TheCCGC is a council made up of representatives from anumber of ministries, and with one representative fromINGC. In addition to making sure that all necessaryministries are involved in disaster management andrecovery efforts, the CCGC is charged with theapproval of regularly updated disaster managementplans. The first of these was completed in 1999, but asecond one has not yet been developed. A morecomplete picture of the institutional history and frameworkfor disaster planning and management can befound in a report developed and funded by the UnitedNations Human Settlements Program, UNEP, and theGEF (Muianga 2007).To a large extent, the 2009 INGC study alreadymentioned (Asante et al. 2009) represents the mostextensive nationally driven assessment of climate changehazards in Mozambique and can serve as the basis foradditional planning. That study did not contain work ondeveloping policy guidelines, however. In addition tothese planning activities of INGC, there have been anumber of initiatives carried out by other agencies andministries. These include contingency plans as a resultof seasonal climate forecasts issued at the SouthernAfrican Regional Climate Outlook Forum (Muianga2007; Lucio et al. 2007); mapping activities such as theLimpopo Atlas completed by CENACARTA withassistance from FEWS-NET; and others.oVeRVIew of soCIal VulneRabIlITYIn MozaMbIqueThere have been a large number of place-based studieslooking at social vulnerability in Mozambique, employingboth bottom-up and top-down strategies.bottom-up Research studies.Reporting on work conducted by the German Agencyfor Technical Cooperation (GTZ), Ferguson (2005)analyzed the natural disaster risk in the Búzi District inSofala Province. In collaboration with the CatholicUniversity of Mozambique, a participatory methodologywas used to identify the population at risk fromdifferent disaster types. The analysis suggested thatdifferent types of natural hazards threatened the safetyand livelihoods of approximately one-third of the population.The study found that human activities—agricultureand deforestation—had degraded the study area’snatural resources (forest and savannah). Fergusonargued that the population is particularly vulnerable dueto a combination of factors, some related to the locationof the area and its topography, and others related to theculture and socioeconomic conditions, and that it is thisvulnerability that turns a hazard into a disaster.Table 2 summarizes the factors leading to an increasedhazard on the one hand and the vulnerability of thepopulation on the other. The study chronicled adaptationmeasures that had reduced vulnerability: theconstruction of new settlements on higher ground andaway from the river; the rebuilding of damaged infrastructurewith due account of the need for being moreresistant to cyclones; and the improvement of disasterpreparedness through simulation exercises that practicedthe implementation of early warning systems. As morefrequent extreme climate events can be expected in thefuture, and considering the fact that relatively little can

12 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U Ebe done to reduce the hazard, vulnerability reduction isof utmost importance to minimize casualties and materiallosses. In this regard, Ferguson saw an importantrole for the government to play at different levels, andargued that disaster risk management should be an integralpart of rural development strategies.Carmo Vaz (2000) reviewed the major flood episodesthat occurred in Mozambique since independence in1975 in the Maputo, Umbeluzi, Incomati, Limpopo,Save, Buzi, Pungoé, and Zambezi river basins, andanalyzed the measures taken to mitigate floods inMozambique. The author classified mitigation strategiesinto two categories: structural and non-structural.Structural flood mitigation measures included dams,levees, flooding areas, and river training, whereas nonstructuralmeasures comprised flood zoning, floodmanagement, flood warning systems, emergency plans,raising awareness, and insurance. In regard to dams, thereview noted that dams with sufficient storage capacitycan play a significant role in attenuating floods. Whileall major reservoirs in Mozambique incorporate a floodreserve in their operating, Carmo Vaz underlined thepotential perverse effect of dams as a means ofmitigating floods. Indeed, for small floods that areabsorbed by the dams, the preparedness of socioeconomicsystems might decrease and thus be hit evenharder in the case of a major flood, since large floodscan exceed by far the storage capacity of the reservoirs.Nevertheless, Carmo Vaz argued for including floodcontrol in the planning, design, and construction stagesof all new dams.As part of a more general inquiry into adaptation toclimate extreme events, Mirza (2003) reviewed the causesof high vulnerability in Mozambique and characterizedextreme weather events in two categories: (1) primaryclimatic events such as floods, droughts, tropical cyclones,heat waves, or cold waves, as well as coastal storms andstorm-generated surges; and (2) secondary events such asmalnutrition or under-nutrition and hunger, outbreaks ofdiseases or epidemics, rural and urban water shortages,crop plantation failure or harvest failure, and landslides,mudflows, and saline water intrusion.The author highlighted nine dimensions and areas forpotential improvement. First, socioeconomic conditionswere judged important factors to vulnerability. FortyTable 2. RIsk anD VulneRabIlITY faCToRs aCCoRDIng To naTuRal DIsasTeRsIn The búzI DIsTRICTFloodCyclones andtropical stormsDroughtHazard• high precipitations (either in district or upstream)• high tide can temporarily obstruct the flow of the RioBúzi to the sea• cyclones develop over the indian ocean or themozambique channel when the water temperature iswarm; wind speeds can reach up to 300km/h, andare usually accompanied by heavy rain (which couldsimultaneously cause flooding)• threat especially for settlements close to coast• climatic variation in mozambique can lead to one ormore years of precipitation below average• influence of El niño-southern oscillations, whichcause high temperature and low precipitation inEastern Africa, while la niña causes heavy rains andfloods• intrusion of saline seawater into the groundwater andthe soils during high tide when the water level of theBúzi River is lowVulnerability• low level of education• proximity of towns to the river• Absence of high elevation point for escape• most people cannot swim and do not have enoughboats• non-availability of warning system(same as above)• traditional houses are not designed and built to resistcyclones• population living on subsistence agriculture; changesin climatic conditions exert significant impacts onsocioeconomic systems• incapacity of storing supplies as seed for the followingyear in case of insufficient yields• Alternative sources of food, such as fishing, smalllivestock, honey, do not allow for substitution• Almost no irrigation system is in place (comité deconselheiros 2003), neither traditional nor modernSource: ferguson (2005).

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s13percent of the population live under the abject povertyline (less than $1 per day), and another 40 percent live onless than $2 per day. Second, Mozambique’s debt repaymentis drawing substantial financial resources from thenational budget, which could otherwise be allocatedtoward promoting development. Third, most of the watercausing devastating floods actually originates fromabroad. Fourth, the primary objective of water dams iselectricity generation, while a multipurpose managementalso would consider flood prevention. Fifth, design criteriafor embankment construction typically consider 5–10year flood return periods, rather than floods of highermagnitude and greater return periods. Sixth, rural areasare generally more affected by flooding than urban areasbecause they are more dependent on agriculture. Seventh,communication was poor during past natural disasters,thus preventing quick and effective humanitarian interventions.Eighth, although Mozambique had a rapidlygrowing economy, living conditions for the majority ofthe population had not significantly improved due tomarked inequality in the distribution of resources. Ninth,human and material resources proved to be inadequate todeal with the major flood in 2000.Patt and Schröter (2008) conducted a study funded bythe World Bank on perceptions of changing vulnerabilityand risk levels. They analyzed three separate sets ofdata. First, they held workshops with farmers and localleaders in villages that had suffered from the LimpopoRiver flood of 2000, and with national level policymakers in Maputo. What they observed in the two setsof discussions was an apparent mismatch in concernover the primary risks: Farmers seemed less worriedabout the risks from future flooding events than theywere about how the shortcomings in the previous recoveryefforts remained apparent, while policy makersseemed unconcerned with the potential unintendedconsequences of their adaptation strategies. For example,in the policy-makers’ workshop, the researchersdivided participants into a number of working groupsand asked each to identify several adaptation strategies,the barriers to successful implementation of those strategies,and the potential negative consequences of thosestrategies were they to be fully implemented. All of thegroups were able to come up with the first and secondlists. However, it was only the working group thatcontained the least expertise on climate adaptation thatwas actually able to envision negative consequences ofclimate adaptation; the other groups simply listed additionalbarriers to implementation.Second, the researchers conducted a survey askingpolicy makers and farmers to indicate the risk levelsfrom climate-related and non-climate-related events. Ingeneral, policy makers saw the climate-related events asmost risky, while the farmers saw the non-climaterelatedevents as riskier. For example, there are differentperceptions between farmers and policy makers as towhich risks are presently becoming more severe. This inturn could explain why some adaptation measures, suchas resettlement into villages outside of the floodplaincould thus be viewed as unattractive because they couldworsen the non-climate risks (crime, for example).Finally, the researchers conducted a household survey,again in two villages in the Limpopo floodplain, in whichthey explored people’s perception of climate change andthe causes for the changes they observed. They foundthat while most people had observed changes, they didnot attribute the changes to issues of pollution comingfrom outside their community or country. Rather, thosesurveyed believed that they had caused some of thechanges by ignoring traditional practices. The researcherssuggested that this could lead to an unwillingness toengage in adaptive behavior, since adaptive behaviorwould represent yet another departure from tradition andhence result in even more unwanted changes.Eriksen et al. (2008) conducted a comparative analysis ofthree bottom-up vulnerability assessments inMozambique and South Africa. The results are interestingboth for the findings of the underlying assessmentsand for the added analysis comparing them. First, anumber of agencies throughout southern Africa—includingthe Southern African Development Community(SADC), the UK Department for InternationalDevelopment (DFID), and the Famine Early WarningSystem Network (FEWS-NET)—formed vulnerabilityassessment committees (VACs) for Mozambique,Malawi, and several other SADC countries. The objectivewas to develop a coordinated system to monitorongoing food insecurity, allowing for cross-nationalcomparisons and the prioritization of relief aid. Theassessments started in 2002 with the proposition thatdroughts were the primary trigger for food insecurity,which implied that rainfall and crop monitoring were the

14 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U Emost important activities to engage in as part of a foodinsecurity monitoring effort. Later, the VACs came tothe conclusion that indeed there were multiple triggeringfactors for food insecurity, and hence it was essential tomonitor a wider variety of indicators and take responsemeasures that consider not just immediate hunger, butalso the patterns of development that were the precursorsto food insecurity.The second assessment considered was the EconomicImpacts of Climate Change vulnerability assessment,led by the World Bank in cooperation with theNorwegian climate research institute CICERO, theUniversity of Oslo, Rutgers University, and theUniversity Eduardo Mondlane. A key focus of thisassessment was on the potential interaction betweenclimate change and economic development. The assessmentfound that climate change tended to have thegreatest negative effects precisely on those communitiesand households that lacked integration into marketsand hence were failing to participate in economicgrowth. One explanation for this could be that marketintegration allows households to be more flexible intheir livelihood strategies, and hence more adaptive toclimate change impacts. The policy implication fromthis study was that greater attention to economic integrationcould be an important means of reducing thevulnerability to climate change.The third assessment, on disaster risk management, wasled by the German Agency for Technical Cooperation(GTZ), in cooperation with the Mozambique RedCross (CVM) and the Catholic University ofMozambique (Ferguson 2005). This assessment focusedon the Búzi River basin in central Mozambique andlooked closely at conditions on the ground. Theresearchers engaged in two activities: First, they engagedin mapping to identify the need for specific infrastructurethat would mitigate the effects of droughts andfloods; second, they identified household- and community-levelcoping mechanisms and infrastructure. Amajor implication of their study is that policy interventionsneed to build upon local knowledge and localpractice, rather than interfere with it.In comparing the results of these three assessments,Eriksen et al. (2008) reached two main conclusions.First, it is essential to consider the multiple factors thatgive rise to vulnerability: not just the drought or floodthat is the triggering event for food insecurity or a lossof life, but the more fundamental patterns of developmentthat exacerbate or mediate the effects of these riskfactors on human suffering. Put into IntergovernmentalPanel on Climate Change (IPCC) language, this issaying that assessments need to consider not just exposure,but also sensitivity. The specific factors that makepeople more sensitive are isolation from markets, a lackof information and education, and a lack of basic infrastructure.Second, it is essential to consider vulnerabilityas it operates at the household and community levels,taking into account household- and community-levelknowledge and infrastructure. This is another way offocusing on the need to assess adaptive capacity, viewingthe household and community as the initial repositoriesof this capacity. Policy interventions need to build on,rather than act against or interfere with, this capacity.studies using the household food economyapproachIn Mozambique, the process of assessing householdfood economies has been undertaken by the FamineEarly Warning Systems Network (FEWS.NET), first in2002, and then again more recently since 2008. The2002 assessment resulted first in a breakdown of thecountry into 39 food economy zones. Essentially, thesewere regions with a similar economic context withinwhich people operated, and thus a homogenous set ofthreats to that economic context (FEWS-NET 2002b).For each zone, the report provides details on the administrativeboundaries and the populations of men andwomen within those boundaries. It then describes themain sources of food, the main sources of income, andthe main risks. It does not describe wealth levels orhousehold food economies within each zone.The 2002 assessment also resulted in two detailed foodeconomy baseline profiles for the Alto Limpopo zone(FEWS-NET 2002a), and for the coastal Nampula andSouthern Cabo Delgado zone (FEWS-NET 2002c).These provide more detailed information on actualhousehold food economies within each zone and weremeant as a proof of concept for putting the five-stepmethodology concept to use within an operationalagency. In 2008, FEWS.NET resumed work on this,based on an updated set of assessment guidelines (FEG

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n p A p E R s15Consulting and Save the Children 2008), and has so farprepared an additional set of five detailed livelihoodprofiles (FEWS-NET 2008a, 2008b, 2008c, 2008d,2008e). These profiles provide a richly detailed set ofinformation on the zones that have been covered so far.Connected with the work in 2008, FEWS-NET alsoprepared in cooperation with the United States Agencyfor International Development (USAID) office forMozambique a set of four briefing papers, in whichthey analyzed the impacts of the global food crisis onrural Mozambican households (USAID Mozambique2008), assessing the vulnerability of the poor, the linkbetween household food economies and vulnerability tonatural disasters, and the utility of the HEA approachfor development. These reports are rather brief, andprovide an overview of the factors that relate entitlement,food security, and vulnerability.Top-down approachesFinally, Patt and Nussbaumer (2009) conducted a studyof social vulnerability as part of the larger INGC studyon disaster risk (Asante et al. 2009). The authors examinedthe combined effects of both climatic drivers ofvulnerability and socioeconomic drivers. To pull the twotogether, they conducted a cross-sectional study oflosses resulting from climate-related disasters across theglobe and built a statistical model that relied on bothclimate variables and socioeconomic drivers, particularlythe Human Development Index. This part of the studywas similar to what had been done previously in thecontext of several global studies (Brooks et al. 2005;Brooks et al. 2004; UNDP 2004; and Yohe and Tol2002). They then assembled projections for both sets ofvariables—the climate ones from the team members inthe INGC study, and the socioeconomic ones fromIPCC and UNDP scenarios that had been assembled—in order to reach conclusions about likely changes overtime. Their results suggested that the socioeconomicdrivers could play an important role in driving vulnerability.Figure 7 shows their projections for vulnerability.As the figure caption indicates, the upper figure showsthe risk of being affected by a climate-related hazard,while the lower figure shows the risk of being killed.The three solid lines in each figure show estimatesobtained using a linear extrapolation of past disasterfrequency trends as the climatic driver in the future.fIguRe 7. esTIMaTes of ChangIng RIskleVels beTween 2000 anD 2060Persons affected, per thousand annually250200150100500AB2000 2010 2020 2030 2040 2050 2060Disaster frequency estimates based on observed trend, 1970–2007No Development A2 B1Disaster frequency estimated ranges based on GCM projectionsNo DevelopmentA2Source: Patt and nussbaumer (2009).note: figure 7a shows estimates for the change in the risk of being affectedby a climate-related hazard, while figure 7b shows the change in the risk ofbeing killed by a climate-related hazard. The figures contain three socioeconomicscenarios and two climate scenarios.The shaded regions, by contrast, derive climate projectionsfrom the INGC modeling results. For both, theblack lines and shaded regions correspond to a socioeconomicscenario with no changes from the present,such as in income or fertility; the blue lines and shadedregions correspond to the IPCC A2 scenario; and thered lines and regions correspond to the IPCC B1scenario. The A2 scenario generally suggests lesseconomic growth in Mozambique, and less change inB16543210Persons affected, per thousand annually

16 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U Eother development factors such as life expectancy andfertility. It does, however, project higher urbanizationrates. The B1 scenario projects greater improvementsin human development and lower rates ofurbanization.What can one learn from this study? First, there is a lotof uncertainty about future vulnerability that isaccounted for by uncertainty about future climatetrends. The linear extrapolation of observed disastertrends leads to much higher estimates of risk levels thando the climate models. Which is right? Given the difficultyof modeling extreme event risks, it is hard to say.Second, the effects of the faster improvement in humandevelopment associated with the B1 scenario are alsoquite large. Indeed, human development projectionsfrom the B1 scenario could lead to falling risk levels bythe middle of the century.Taking the bottom-up and top-down studies together, itis clear that social and socioeconomic factors do play alarge role in determining vulnerability to climate changenow, and will in the future. The bottom-up studiessuggest that factors such as people’s perceptions andattributions of the underlying risk, and the informationthat they have at their disposal to respond to changes inrisk levels, could influence their adaptive capacity. Thestudies also suggest that socially and economicallymarginalized groups will have a more difficult timecoping with and adapting to climate variability andchange and will therefore experience higher vulnerability.The top-down study shows that a rising economictide could raise the overall ship of Mozambique, makingit more resilient and less vulnerable to climate change.However, this study says nothing about how the socioeconomicchanges will reach the most vulnerablecommunities within the country.

173. ReseaRCh MeThoDologYReseaRCh sTRaTegY anD quesTIonsBoth the guiding questions and the research strategy forthis study were, to a large extent, set in advance in theWorld Bank’s original scoping materials for all countrystudies. The questions were to identify the specialconcerns faced by the most vulnerable populations inMozambique and to identify adaptation strategies andpathways that would best respond to those concerns. Ata project meeting held in Accra in June 2009, this wasfurther refined to identify the extent to which the prioritiesset forth in existing adaptation plans–such asNAPA–would differ if one were to give special considerationto the needs of the most vulnerable communities.To a large extent, this would require identifyinghow the concerns of the most vulnerable differ fromthose of the population in general.To address these questions, the research strategyincluded four key steps:• A review of primary and secondary literature, supplementedby stakeholder interviews, to identifyboth existing knowledge about social vulnerabilityin Mozambique and current and planned efforts toreduce that vulnerability.• A review of climate impact studies, including mapping,to identify key hotspots of social vulnerabilityand a set of sociogeographic zones.• A series of participatory scenario development(PSD) workshops, with the last one held in thenational capital and involving national-levelstakeholders, to identify the most vulnerable groupsand appropriate adaptation pathways.• Fieldwork in vulnerability hotspots to validate theresults from the PSD workshops. The fieldworkwas to consist of three methods: (1) a set of institutionalinterviews with representatives from keyorganizations; (2) a set of focus group discussionswith people from different socioeconomic strata;and (3) a set of household surveys, with a samplingmethod to capture variance in the anticipated driversof social vulnerability.The first of these steps was carried out quickly by thelead consultant, with the results reported in the inceptionreport. The inception report also proposed a set ofsociogeographic zones. Following this, the localconsultant suggested refinements to these zones, andengaged in GIS-based analysis to identify hotspotswithin these zones. Work then slowed, as bureaucraticcomplexities made the process of releasing funds forthe PSD workshops and fieldwork difficult. Eventually,the team held three PSD workshops in July andAugust, with technical assistance from IISD andESSA technologies. Fieldwork commenced in August,and was finished in mid-September. The initialprocessing of data yielded field notes by the end ofSeptember. These required translation into English,which took place by mid-October. Quantitative datawas available in a form suitable for analysis bymid-November.sITe seleCTIon anD saMPlIngIn the inception report, the lead consultant proposed aset of six zones. These are:

18 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U E• Coastal urban areas, most importantly Maputo andBeira. This zone is marked by highly differentialvulnerability across income groups, with large periurbanareas vulnerable to flooding from both riversand the ocean.• The non-urban coastal strip. This zone is markedby high vulnerability to coastal flooding and stormsurges from tropical cyclones, as well as threats oferosion. It is relatively food secure, with low rates ofpoverty.• The districts containing the Limpopo River valleyupstream of Xai Xai. This zone is unique in beingexposed to two very different threats at a high level:river flooding and drought. It has a relatively highpopulation density, and hence high numbers ofimpoverished people. There has been extensivestudying of this region, and so it is possible toaccess a great deal of baseline data.• The other river valleys, in particular in the Buziand Zambezi river valleys, which are highly susceptibleto flood risk and to flooding caused by tropicalcyclones, but are less susceptible to droughts. TheBuzi River region has also been extensively studied,as part of German-funded activities, and so there isno shortage of baseline data.• The drought-prone inland areas, in particular inthe south. These areas are highly susceptible todrought—indeed years of adequate rainfall to supportagriculture are the exception, rather than therule—while the people are often dependent onremittances for survival. Population densities inthese regions are low.• The inland areas of higher agricultural productivity,including the highly productive and populated areasin Zambézia. These areas are perhaps the least vulnerablein Mozambique, facing adequate rainfallmost years and no extreme risks from flooding ortropical cyclones. They are somewhat heterogeneousin terms of poverty rates and food security. Thehighly productive regions in Zambézia stand out fortheir high population density and relatively lowvulnerability.Further discussions among the entire project teamagreed on these zones. The local consultant thenengaged in two processes. First, he delineated themprecisely, in terms of particular districts covered withineach. Second, and more importantly, he identifieddistricts within each zone that constituted risk hotspots.To do so, he first engaged in mapping the differentlevels of risk, for each of the major classes of risk, andoverlaying areas of high population on these. Figure 8shows the result of this. It was agreed to combine thisanalysis with the maps of the zones, in order to identifyone or two districts within each zone that representedhotspots for that zone. Within each of these districts,the fieldwork would take place at multiple administrativeposts, in order to capture potential variation within thedistrict.Budgetary factors and time constraints, ultimately,limited the number of districts considered in the fieldstudy. Figure 9 shows the locations of these sites.In terms of the PSD workshops, it had been originallyplanned to hold one regional workshop, and onenational workshop, back to back. During discussions inJune in Accra, the project team decided instead to holdthree regional workshops, one each in the northern,central, and southern regions of the country, followed bya national workshop in Maputo. Time constraints led toa scaling back; ultimately the team held regional workshopsin Xai Xai and Beira, inviting stakeholders fromboth the central and northern regions of the country tothe Beira workshop. A national-level workshop inMaputo followed.

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s19fIguRe 8. soCIal VulneRabIlITY hoTsPoTsSource: authors.note: The district map identifies areas where risks from drought, flood, and cyclone are considered to be high or very high.

20 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U EfIguRe 9. fInal fIelDwoRk sITes

214. PaRTICIPaToRY sCenaRIoDeVeloPMenT woRkshoP ResulTsPsD woRkshoPs: DesIgn oVeRVIewThe team organized three PSD workshops. The originalintention was to conduct two one-day PSD workshopsin each country. One of these would focus oninvolving local-level stakeholders, while the otherwould focus on involving national-level stakeholders.To assist in the development of these workshops, theWorld Bank issued a contract to a consortium of twoCanadian partners—ESSA Technologies and theInternational Institute for Sustainable Development(IISD)—to develop a workshop program, prepare visualmaterials for the workshops, and in selected countries(so-called Tier 1 countries), attend and help to facilitatethe workshops themselves. In the first of theAfrican Tier 1 countries to hold a workshop—Ghana—the consultants from Mozambique andEthiopia also attended, in order to observe the workshopin operation, so as to be able to repeat its keyaspects. This Ghana workshop—involving nationallevelstakeholders—took place in June 2009.The consultants attending the Ghana workshop werenot entirely satisfied. Meeting the day after the workshop,they reached several new conclusions. First, it wasessential to place more emphasis within the workshopson different stakeholder groups in order to stimulatemore in-depth thinking about development and climateconcerns. At the Ghana workshop, participants hadbroken out according to their geographical expertise. Itwas decided that for future PSD workshops, theywould break out according to their sectoral or livelihoodgroup expertise or interest, and indeed the identificationof relevant sectors or livelihood groups couldbe an outcome in and of itself of the workshop. Second,the consultants decided that for Mozambique it wouldbe useful to have more than one local-stakeholder-levelworkshop, for two reasons. Holding a series of workshopsin different regions of the country could reducetravel distances for workshop participants. Also, differentregions of the country have different climate, development,and adaptation concerns, and it would beuseful to address these in separate workshops.Originally the plan was to have three such workshops,but due to organizational constraints the numberchanged to two. The third conclusion was that theworkshops would be over two days, rather than one.This would give more time for in-depth discussion.Fourth, ESSA and IISD would provide assistance tothe Mozambique team, as with the Tier 1 countries,attending at least the first workshop. Given languageabilities, the consultants Raul Varela and IsildaNhantumbo would jointly facilitate both workshops inPortuguese, with additional assistance from studentsrecruited from the University of Eduardo Mondlane inMaputo.There were, however, several important differences inthe designs of the three workshops. The first twoworkshops—in Xai Xai and Beira—each followed aneight-step procedure, while the national workshop inMaputo followed a seven-step approach. These appearin Figure 10. The two procedures differed in severalrespects. Most importantly, the eight-step approachwas built much more around the idea of imagining afuture vision for the country, and people’s lives in that

22 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U Efuture. After having done this, participants identifiedways that climate change could interfere with thatvision, and then thought of the appropriate adaptationsthat could lessen this interference. The seven-stepapproach, by contrast, was built much more aroundconsidering specific climate impacts, and identifyingappropriate adaptation pathways to lessen thoseimpacts. Roughly speaking, the eight-step approachreflects a development-oriented view of adaptation(making the development pathway more robust toclimate change), whereas the seven-step approachreflected a climate-oriented view of adaptation (reducethe negative consequences associated with a range ofpotential climate impacts).The second main difference was in the manner inwhich breakout groups were defined, and here the majordifference was between the Beira workshop and theother two. In all workshops, participants identified earlyon where they were from, and where their geographicalarea of greatest expertise lay (Figure 11). In Xai Xai andMaputo, participants then divided into breakout groupsalong the lines of different economic activities orgroups. These were agriculture and ranching, agroforestry,fishing, and commerce. In the Maputo workshop,the agriculture and ranching participants in turn dividedinto two groups, one concerned with subsistence activityand the other with commercial activity. In the Beiraworkshop, the breakout groups divided according togeographic location, with a separate group for each ofthe provinces represented at the workshop.These different approaches led to very different units ofanalysis. In the activity-oriented approach, discussionfocused on the constraints and opportunities people facewhile earning their livelihoods in particular ways, withthe recognition that climate change threatens someactivities more than others. The participants discussedgovernance issues at a more general level, with somewhatless attention to how different provinces wouldnecessarily implement different types of policies, or onthe combinations of policies needed to address theneeds of different economic groups. In the geographyorientedapproach, discussion focused more on the portfoliosof activities present within each province, and theneed for policies at the provincial and district levels todeal with the impacts of climate across multipleeconomic activities. The important justification for thiswas to stimulate a different type of thinking, onefocused on practical solutions that people workingtogether along existing jurisdictional lines could address.A critical question was whether changing the framingof the unit of analysis for adaptation—from activities inthe Xai Xai workshop to geographic regions in theBeira workshop—led to differences in substantiveoutcomes.oVeRVIew of ResulTs fRoM loCal anDnaTIonal woRkshoPsThe two local workshops were especially interesting,because unlike many other workshops they focusedparticipants’ attention first on development, and onlythen on the potential threats that climate would imposeon that development. This was especially the case in theXai Xai workshops, where each of the breakout groupsconsidered how development would proceed for aparticular sector of society. The result was somewhatless the case in the Beira workshop, where the separationinto provincial breakout groups did not allow fordiscussion to be as specific on different societal sectors,and instead the discussion was more general in terms ofimproved quality of life for all people.xai xai key InsightsFrom the Xai Xai workshop, the most interesting observationis that in very few cases did the developmentchallenges that people listed have much to do withclimate change. The main exception was for agriculture.In the fishing livelihood group in Xai Xai, for example,there were challenges related to low fish catches, to thedifficulty of selling what they had caught, and to thelack of a future for the children in these communitiesdue to poor education. All of the 2050 vision pointsrevolved around these issues. In the agroforestry group,the challenges for commercial actors related to poorinfrastructure and access to credit, while for informalactors they related to poor forest management practicesand ecosystem changes. In one case weather was listedas a challenge—the fact that charcoal producers needdry weather to work—but this is an issue more of insufficientability to invest in building infrastructure fortheir operations. They were not saying that inMozambique it rains too often, simply that rain presentsa challenge.

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s23fIguRe 10. sTePs In RegIonal anD naTIonal woRkshoPs7Reporting backfrom groups anddiscussions8Reflect on theday1Introductions;context setting;overviewof day2Review thecurrent situation6Revise visionof the future &timeline to includeadaptations5Review, identify& evaluateadaptation /Coping optionsEngagement andparticipation4Challenge thevision with boundaryconditions: Climateand socioeconomic3Develop a visionof the futureand timelineNote: The top figure shows the eight steps in each of the two regional workshops, while the figure on the bottom shows the seven steps taken in the nationalworkshop.

24 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U EfIguRe 11. MaPPIng of geogRaPhICaloRIgIn anD exPeRTIsereducing exposure or sensitivity to natural hazards. Onlyin the agriculture and ranching group did climatefactors play an important role as a barrier to development.This group listed drought and floods as two ofthe three most important barriers. For this group, abouthalf of the adaptations listed had something to do withreducing vulnerability to droughts and floods, while therest related to issues of general poverty alleviation andgreater access to markets.beira key InsightsNote: These are the locations identified by participants in the xai xai workshopand show the effort to obtain stakeholders from a wide geographicalrange. breakout groups, however, followed sectoral expertise.Again, the 2050 visions were consistent with reducingthese challenges. In the trade and commerce group,again the challenges were related to poor infrastructure(such as roads, water, and electric grid), poor institutionsand governance (such as lack of education, especially forgirls; poor spatial planning; poorly developed markets;and inadequate access to credit), and generalized effectsof poverty (such as a low standard of living, poor health,and high unemployment). This group did identify oneclimate-related challenge: the incidence of naturalhazards. The visions for 2050 were again consistentwith addressing these challenges, although it is noteworthythat the vision did not contain anything aboutTo test the effects of changing the workshop format,the project team divided participants into geographicallyidentified groups. Each of the provinces identifieda set of development targets, in terms of life expectancy,per capita income, fertility, population, education, andthe size of the economically active workforce. Theythen listed the factors that would allow these targets tobe realized, with similar ideas and suggestions acrossprovinces. They then identified the potential climateimpacts and the negative effects each could have in theprovince. There were some differences across provinces.First, Tete Province is in the driest part ofMozambique, where local communities have morelimited livelihood strategies. Second, Nampula is aregion normally hit by very strong hurricanes. Third,from Beira to Rovuma River, a long coastline ofcommunities depend on fishing, using poor fishingmethods, and thus are potentially more vulnerable tosea level rise and changes in storm intensity. Fourth, thecentral and northern regions have large forest areas,potentially susceptible to the combination of strongwinds and drought, leading to highly damaging firessuch as those that devastated Manica and Sofala in2008. Fifth, there is large-scale contract farming forcotton (for example, in Nampula) and tobacco (forexample, Tete); this may mean that those communitiesengaged in such activities may have more income, andhence greater adaptive capacity. It may also mean thatvolatile price markets, especially for cotton, can weakentheir capacity to cope.However, no group focused on the specific relationshipbetween the factors leading to developmenttargets and the climate impacts, and so it is hard tosee whether they identified linkages or lack of linkages.Likewise, each of the groups listed adaptations

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s25appropriate for different climate impacts, and theseoptions are consistent with those appearing in theNAPAs and the literature more generally, with nearcomplete agreement across provinces as to the types ofadaptations they see as important. What one can learnfrom these results, however, is that the adaptationneeds that people identified were quite homogenousacross the country. It was not the case that each provinceneeds its own unique ways of adaptation to eachclimate hazard.In terms of workshop design, one conclusion to draw isthat the geographic groupings may have been useful fordeveloping discussions among a group of colleagueswho could then later implement their collective ideastogether (since they worked in the same province).However, it was a less successful strategy for elicitingdetailed information about challenges associated withdifferent livelihood activities, but to a large extentcommon across provinces.Maputo key InsightsThe Maputo workshop devoted less time to identifyingways in which climate may affect particular sectors—indeed suggestions, collated from the regional workshops,were handed to participants on note cards—andmore attention to developing pathways of adaptation,something not done in the two regional workshops.Table 3 lists these pathways for each of the five breakoutgroups.These suggested pathways are similar to the resultsfrom Xai Xai, in that very few of the adaptation strategieshave much to do with climate change and muchmore to do with development in general. The majorexception is for agriculture, both subsistence andcommercial. This also finds its way into the longer termvision for the commercial sector, i.e., enhancing thesustainability of trade in agricultural products.synthesisAcross the workshops, the more interesting results werein Xai Xai and Maputo, where participants focused theirattention on particular economic and livelihood activities.Here, the common finding was that outside of agriculture,climate change does not present many clearlyidentifiable threats to the improvement in people’s livelihoods.Attention within these sectors has to concentrateon improving existing management practices, fromefforts to avoid overfishing and overharvesting of forests,to the continued development of a healthy commercialsector that can provide people with reliable sources ofincome. Within the agricultural sector, the threats fromincreased prevalence of both floods (in the fertile floodplains)and droughts (everywhere else) require a responsethat must take the form of both hard and soft measures.Table 3. aDaPTaTIon PaThwaYs IDenTIfIeD In The MaPuTo woRkshoPForestryTradeFisheriessubsistenceagricultureCommercialagriculture2009–2015 2015–2030 2030–2050wildfire management andchanges in standardsRehabilitation of ruralcommerce and trade networks;simplification of licensingprocedures for small andmedium enterprisesimproved fishing nets &improved fish management topromote sustainable use ofresourcesconservation agricultureincluding planting cropsresistant to droughtconstruction of dykes forwater retention (small scale)Alternative incoln! generatingactivitiespost-harvest processing;building silos; improved family/community storage facilitiesBuilding infrastructure forprocessing and conservation ofto fish (increase yield andincrease quality of product)water harvesting techniques (ie.water storage tanh forconsumption and irrigation)construction of flood gates anddams in areas vulnerableReforestation, forest management and(environmental) education particularly in forestconcession areas (commercial forestry)construction of infrastructure with can withstandclimate disasters especially related to trade inagricultureEducation for skills development influence inbehavior and attitude to enforce seasonalfishing restrictions; use improved fishing netsEducation so that people can learn to live withdroughtReforestation so the areas destroyed by forestharvesting to floods and wildfires are reduced

26 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U EaDaPTaTIon oPTIons: keY seCToRalTheMesIn all three workshops, participants identified the samefour basic activity sectors—agriculture, forestry, fishing,and commerce—as crucial for a consideration of adaptationpathways. In the Beira workshop these were allconsidered, even if discussion took place in geographicallydefined groups. Here are results from each of thefour sectors, across the three workshops.agriculture and RanchingTable 4 presents the specific results from each of thethree workshops.Across the workshops, the theme emerged that the agricultureand ranching sectors would be the hardest hit byclimate change. Potential increases in the likelihood andmagnitude of drought, flooding, cyclones, and coastalflooding and intrusion would lead to reduced yields andgreater insecurity. The adaptation options that peopleidentified were a combination of infrastructure projects,the improved use of drought-resistant crops, andoptions that would allow people to diversify theirincomes and self-insure.agroforestryAcross the workshops, deforestation emerged as arepeated problem. It was unclear in many people’sminds the extent to which this would be a direct consequenceof climate change, or simply of poor forestmanagement practices, or of some combination of thetwo. Without resolving this question, almost all groupssuggested approaches to improving forest managementand forest access for commercial uses. Many of theparticipants suggested greater empowerment of localcommunity groups, both in the design of managementpolicies and in their enforcement, such as throughcommunity-based natural resource management.fishingPeople were worried about the direction that fisherieswere heading, but there was very little consensus on theextent to which climate change was a driver of thisproblem, except for the fact that increases in cyclonestrength could threaten fishing fleets in the absence ofadequate early warning. In the Xai Xai workshop,participants suggested that the impact of climate changeon fisheries would be indirect, by threatening agriculturalproduction and in turn leading to a greater relianceon fish as a food source and source of income.Trade and CommerceThere was a concern that climate change could negativelyaffect trade and commerce through several differentpathways. One pathway was in terms of losses inagricultural productivity; since agricultural commoditiesare the main thing traded, this would have a directnegative effect, both on supply of commodities from therural areas into trade networks, and in terms of loss inpurchasing power by rural people. Another pathway wasin terms of a loss of infrastructure from extreme events,Table 4. woRkshoP ResulTsxai xaiBeiramaputoKey impacts fromclimate change Consequences for development Adaptation optionsdroughts, floods,cyclones, and sealevel risedrought, flood,cyclonedrought,desertification, lossof soil fertilityloss of production, loss of soilfertility and increased salinity,destruction of infrastructure,increased mortalitymalnutrition, loss of yields, loss ofpossessions in low-lying areasReduced crop yields, poverty,disease, malnutrition, foodinsecurity, loss of grazing land, lossof forest landconstruction of water retention and flood protectioninfrastructure; construction of barns for animals, silos for grainstorage, and food processing facilities to give greater marketaccess.drought-resistant crops, water conservation, incomediversification through improved market access, resettlementcrop diversification, rainwater conservations, expandedirrigation, improved social services and health care, livelihooddiversification, micro-finance for value added activities andmarket access

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s27Table 5. sPeCIfIC ResulTs anD suggesTIons fRoM The woRkshoPsxai xaiKey impacts from climate change Consequences for development Adaptation optionsdrought, flood, cyclone, and sealevel risechanges in forest ecosystems,destruction of forest accessways, loss of coastal vegetationReforestation, road construction, communityresource conservation, early warning systemsBeira drought, flood, cyclone wildfires, deforestation community forest management, increasedconsumption of wild fruits, reforestationmaputowildfires, loss of vegetation, lossof soil fertilityfood insecurity, reduced forestryyields, destruction of housingimproved forest management including betterenforcement of existing laws, community-basedreforestation, improved control of wildfires,establishment of fire breaks, infrastructure toprotect housing from fireTable 6. keY fInDIngs In The fIshIng seCToRxai xaiKey impacts from climate change Consequences for development Adaptation optionsdrought, flood, cyclones, and seallevel riseloss of agriculture leading to higherpressure on fisheries, reduction incatches, loss of fluvial fisheriesAdaptive agriculture, reforestation, earlywarning, introduction of new fish species,aquaculture.Beira drought, flood, cyclone none identified moving artisanal fishing areas to thosemore suitable, community riskmanagement, early warningmaputo changes in fish availability overexploitation and extinction of somespecies, reduced catches and attendantincome, food insecurityimprovements to fishing nets, storage,and processing facilities; education andskill training; livelihood diversificationaway from fishing; improved earlywarningTable 7. keY fInDIngs foR The TRaDe anD CoMMeRCe seCToRsxai xaiKey impacts from climatechange Consequences for development Adaptation optionsdrought, flood, cyclones, andsea level risewild fires, loss of agriculturalproductivity, destruction ofinfrastructure, displacement ofpeople, loss of land for economicdevelopmentReforestation, adaptive agricultural practices,more resilient infrastructure, improved coastalerosion controlBeira drought, flood, cyclone loss of infrastructure Building stronger infrastructure, improvingtransportation, food processing and storage,and market places, improved socialassistance, improved finance for smallbusinessesmaputoReduced and more variableagricultural production, loss ofinfrastructure, coastal floodingand erosion, cyclonesmarket price fluctuations andcommodity scarcity, degradation oftransportation networksconstruction of grain and fish storagefacilities, price monitoring, micro-insurance,improved roadways, flood protection barriers

28 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U Eboth buildings in town and roadways. In addition toaddressing these causes directly, participants suggestedthat better access to credit and information would stimulatea diversification and general development of thesector, thus making it more resilient.IDenTIfICaTIon of MosT VulneRablegRouPsOne of the important goals of the workshops was toidentify the people and places that could be mostvulnerable to the effects of climate change. The results,in general, were unsurprising as to both.In terms of vulnerable places, these seemed simply tomirror where climate-relevant activities took place andthe hazards were high. In the Xai Xai workshop, forexample, the participants identified areas within thesouthern region of “most vulnerable” and “second mostfIguRe 12. VulneRable RegIonsIDenTIfIeD In The xaI xaI woRkshoPvulnerable” status, for each of the four livelihood activities.Figure 12 shows these identified areas, and ingeneral they seemed to match up with the overlap ofhigh hazard level and the particular activity. For agriculture,it corresponded most closely to areas prone todrought but where people nevertheless farm. In the caseof fishing, it appeared to be the result of thinking aboutwhere inland fishing is taking place and resources areunder threat from overfishing. Coastal fishing also wasseen as a hazardous activity, especially as artisanal fishersincreasingly go out farther from shore to find stocks.In the case of trade and commerce, it appeared thatpeople were most likely to consider not the main cities,but rather smaller commercial centers that are marginal,and could be threatened in their existence by increasingrural poverty or decreasing crop yields.In the Beira workshop, participants split into workinggroups according to provinces, and so there was not theopportunity to identify those places that were mostvulnerable across the whole region. However, participantsidentified the vulnerable groups within their provinces.These included (a) fishermen vulnerable to hurricanesand possible sea level rise; (b) the farming communitiesusing fire for land clearing and the communities near theforests, both of whom were considered vulnerable to wildand manmade fires as propagation may become easierduring lengthy drought periods; and (c) pastoralists inTete, where the ranching of small ruminants such asgoats is a very important enterprise and the exacerbationof drought conditions has the potential to reduce pastureavailability and worsen erosion.In Maputo, participants again engaged in mapping—asin Xai Xai—with very similar results. Figure 13 showsthe results for two sectors—fishing and forestry—and inboth cases (as well as other sectors) the identifiedregions of high vulnerability were where risk levels arehigh and large numbers of employed people areemployed.Across sectors at the Maputo workshop, the participantsidentified the following areas as most vulnerable:• Coastal Nampula and southern Cabo Delgado,including the cities Nampula and Pemba. Thiswould appear to be driven by fears of sea level riseand increases in cyclone strength and frequency.

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s29fIguRe 13. RegIons IDenTIfIeD as MosT VulneRable In The MaPuTo woRkshoPNote: The left-hand picture shows results for the fisheries sector, while the right-hand figure shows results for the forestry sector.• Northern Sofala and southern Zambezia. Thiswould appear to be driven by fears about increasedrisk of flooding.• Both inland and coastal areas of Gaza. This wouldappear to be driven by fears of increased risk ofdrought.Participants also identified groups of people theyconsidered to be the most vulnerable. All workshopsconsidered this question and arrived at similar answers:Generally, the most vulnerable were those people withoutthe resources to withstand livelihood shocks, withoutthe capacity to adapt, and without social supportnetworks. Specifically, they included (a) children, inparticular orphans; (b) women, in particular womenwho head households; (c) the elderly; and (d) thosewith physical or mental handicaps.In all workshops, participants suggested that membersof each of the above groups were most vulnerable whenengaged in dry-land farming as their main livelihoodactivity. They suggested that existing social supportmechanisms, such as social security offered by thegovernment, are inadequate to ensure the members ofthese groups with adequate livelihood security.RelaTIVe PReValenCe of “haRD”VeRsus “sofT” aDaPTaTIon oPTIonsOne workshop result that stood out was that peoplesaw both hard and soft adaptation options as important.Within each category, people also saw a mix ofoptions that would be centrally planned and deliveredby the government or NGO community, and those thatwould be operationalized at the household or

30 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U Ecommunity level. Another axis where adaptationoptions differed was in terms of whether they directlyaddressed a particular climate risk, or would primarilyact by increasing adaptive capacity. In the followingtable, the latter appear in italics.One interesting result was that when participants had toidentify which options would directly benefit the mostvulnerable communities, they were much more likely toidentify soft centralized options, and in particular thosethat addressed underlying problems of adaptive capacity.These included better credit, better information, andimproved health care and social services and support.This is unsurprising, since what separates the mostvulnerable from the others is their lack of access tothese things. Specifically, the Maputo workshops listedthe following adaptation options as most directly benefitingthe most vulnerable:• Improved rural trade, with fewer barriers to entry,and technical assistance programs for vulnerablegroups• Improved agriculture practices, including more useof wild resources such as fruits, apiculture, and nontimberforest products, and better mapping of agroecologicalzones• Education for capacity improvement, combinedwith better micro-credit.Participants at the workshops covered a wide range ofexperience, both in terms of the level of activity inwhich they were engaged, and the sector in which theirexperience lay. The workshops did not record results atthe individual level, however, and all group workcontained stakeholders with a variety of backgrounds.Hence, it is not possible to identify whether there weresignificant differences in views as to adaptation optionsacross different classes or groups of stakeholders.CongRuenCe wITh naTIonal Plans,InCluDIng naPasOne intention of the PSD workshops was to find outwhether an approach to adaptation planning that explicitlyconsidered (a) the needs of most vulnerable communities,and (b) the potential for climate impactsinterfering with development visions, would differsubstantially from previous adaptation planning exercises,such as the NAPA. The main conclusion that one canreach is that there are very few differences in outcomesbetween the NAPA process and the PSD process.The NAPAs identified four priority areas for nationalplanning and policy development. These were:• Strengthening the early warning system for floods,droughts, and tropical cyclones. This would benefit thesafety of people engaged in all areas of economicactivity threatened by disasters. It would also havesignificant economic implications for the agriculturalsector—the activities of which are directlyaffected by flooding and droughts—and theTable 8. keY aDaPTaTIon oPTIons fRoM PsD woRkshoPshardsoftCentralized• flood control dikes and levies• coastal flood control gates• dams and irrigation channels• Improved roadways• Improved communication infrastructure• improved early warning of hazards• Better planning and management of forest, fish, and othernatural resources• Resettlement of populations to lower risk zones• More credit and financial services for small businesses andrural development• Better education and information for the rural areas• Improved health care, social services, and social supportfor the poorDistributed• farm-scale water storage facilities• more robust buildings• Grain storage facilities• Improved food processing equipment• Better utilization of drought-resistant crops• Use of highland areas for living and lowland areas forfarming• Better household and community management and use ofnatural resources• practice of soil conservation agriculture• Diversification of livelihoods away from agriculture• Better planning of how much grain to save for personalconsumption, and how much to sell for income generationNote: The items appearing in plain text directly respond to anticipated climate hazards, while those in italics respond to the need for improved adaptive capacity.

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s31fisheries sector, given the exposure of its infrastructureto tropical cyclones and the need to bring fishingfleets into safety.• Strengthening the capacities of agricultural producers tocope with climate change. This attention to capacityincludes both hard and soft adaptation measures,and also addressed the general state of development.Putting in place flood control and irrigation infrastructurewould clearly improve people’s capacities.So would improving their knowledge about how touse early warning systems to mitigate losses fromfloods and droughts. Improving capacities couldalso, however, mean developing other income-generatingactivities for people now engaged solely inagriculture, as these would better enable people towithstand climatic hazards.• Reduction of climate change impacts in coastal zones.This would likely involve a great deal of infrastructureto protect urban areas and transportationroutes, and thus would have the greatest impact,from a social perspective, on commerce and industry.It would also involve a wider range of measuresbeneficial to both fishing and coastal agriculturalcommunities.• Management of water resources under climate change.This would likely have the largest impact on theagricultural sector, which relies on a continuoussupply of rainfall or irrigation during the growingseason and is most vulnerable to flooding. It wouldalso have an impact on energy production, and thusindirectly on commerce, as well as on public healththrough improved sanitation.Other national plans have contained similar or identicalpriorities for adaptation and development. The congruencebetween these plans and the results of the PSDworkshops can be seen in two respects:• The PSD process independently identified all of theadaptation priorities previously noted in theNAPAs. First, the PSD workshops identified that,among sectors, agriculture is the most vulnerable,both because of its high climate exposure andbecause of the social vulnerability of the peoplepracticing it. Second, within the agricultural sector,it identified that a range of measures is required,from those that address climate impacts such asflooding and drought directly with newinfrastructure and knowledge, to those that allowfarmers to broaden their livelihoods by engaging incommercial activities. Third, the workshops dididentify the threats to infrastructure in the coastalzone, including roads, fishing boats, and buildings.Fourth, the PSD workshops identified early warningsystem improvements as a key cross-cuttingadaptation. Fifth, the workshops identified the bettermanagement of water, such as through the practiceof conservation agriculture, as an importantadaptation option.• The PSD workshops identified the options previouslynoted in the NAPAs as the most importantelements of adaptation pathways that directlyaddressed climate concerns. In the Beira workshop,participants focused most closely on the differentclimate impacts and how to adapt to each of them.All of the options they identified as most importantmatch one of the NAPA priority areas.In one respect, however, the PSD workshops—especiallythe one in Xai Xai—reached an implicit conclusiondifferent from that of the NAPA process. The conclusionwas that, except within agriculture, developmentgoals in general were not threatened very much byclimate change, and that far more important are thosedevelopment initiatives that make sense independent ofclimate change. This conclusion can be reached whenstarting by considering development visions withindifferent economic sectors, moving on to considering thethreats to those visions, and finally considering howclimate change may contribute to these threats. In theNAPA process, by contrast, the question was somewhatdifferent: what can be done to adapt to those climateimpacts that could pose a threat to development. This isthe same question that guided the Maputo PSD workshop.When considering this question, the Maputo PSDcame up with answers similar to the NAPA process.PolICY PReConDITIons anDInsTITuTIonal baseIn all workshops, some of the most important adaptationoptions represented improvements in existinggovernment programs and practices. For example, manyparticipants suggested that the forestry sector did notnecessarily need new laws to promote more sustainableforest management, but rather more effective and fair

32 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U Eenforcement of existing laws. Statements such as thesethus suggest a necessary improvement in the ability ofthe government to act effectively—what many mightconsider a precondition.The Maputo workshop was the only one to addressdirectly the issue of preconditions. Participants identifiedthe following list of options:• Improving the competence of government• Combating corruption• Improving vocational training• Creating a more attractive policy for investment inMozambique• Mainstreaming climate issues into sectoral policies.Does this list make sense? Certainly, having a competentgovernment free of corruption does make sense and isconsistent with the messages from the other two workshops.Vocational training and improving the investmentclimate are also, perhaps, the two most important waysof improving adaptive capacity. Whether one considersimprovement in adaptive capacity to be an adaptation inits own right, or a precondition for adaptation, is asubject of much lively academic debate. These participants,at least, though the latter. The final item on thelist, mainstreaming, simply represents a particularapproach to adaptation governance. Many would see thisas the best way to begin to address climate concernsacross a wide range of sectors, where climate is not themain problem on policy-makers’ list of concerns.ConClusIons fRoM The woRkshoPTRaCkThe workshop track was an extremely useful means ofgenerating intense discussion on the issues of climatechange vulnerability and adaptation in Mozambique,paying particular attention to the issues facing the mostvulnerable in society while taking a development-firstperspective. The main result from the workshops is tovalidate previous findings. These are:• The most vulnerable in society are people lackingthe capacity to cope with climate shocks or adapt tochanges, and without strong social and family supportnetworks. These include orphans, femaleheadedhouseholds, the elderly, and people withphysical or mental handicaps.• Adaptation options need to address both the specificrisks associated with climate change—droughts, floods, and cyclones—with a portfolio ofhard and soft options, as well as with the underlyingcapacity of people to adapt and their resiliencein the face of extreme events.• In the case of the most vulnerable people, theirmost urgent need is for assistance that will improvetheir resilience and adaptive capacity, such as education,social support, and programs that allow themto diversify their livelihood strategies.• Agriculture is the area of economic activity inwhich the future development is most threatened byclimate change impacts. It is essential both todevelop specific infrastructure and soft adaptationsto protect agricultural producers from extremeweather and climate events, and to enable them todiversify their activities into other income-generatingactivities.None of these findings contradicts the main findingseither of past adaptation assessment or planning exercises,such as the NAPA, or the findings of theMozambique sectoral findings in the EACC study.

336. fIelDwoRk ResulTsThe primary purpose of the fieldwork in Mozambiquewas to validate the results from the PSD workshops, intwo respects. First, we were interested in the question ofdifferential vulnerability: who were the most vulnerableand what were the effects of climate change on themrelative to everyone else. Second, we were interested inthe coping and adaptation options identified. Wouldthere be regional variation in the options identified?To address these tasks, the team gathered data usingthree methods. The first was a set of institutional interviews:speaking with particularly knowledgeable peoplein the community, or people with decision-makingresponsibility. The second was a set of focus-groupdiscussions: speaking with groups of people representingcommunities covering a range of vulnerability. The thirdwas a household survey: interviewing households ineach community, again representing a range of vulnerabilityprofiles, to find out their individual perspectives.The three activities took place in each of the districtswhere fieldwork was conducted, as shown in Figure 9.In some districts, however, there was not the opportunityto conduct institutional interviews.ResulTs fRoM InsTITuTIonalInTeRVIewsAll of the institutional interviews followed a commonformat. The steps in this format were:• Describe the participants in the interview session.There were usually several participants from thelocal government and NGOs operating in thedistrict.• Identify the main climate risks in the district.These included droughts, floods, cyclones, and insome cases wildfires and coastal erosion.Participants identified past events that had beenparticularly severe.• Describe the overall sensitivity of the community tothese risks, identifying the groups within the communitymost vulnerable to these risks.• Identify options to mitigate and adapt to theserisks. The participants went through the list of risksin their district, and suggested adaptation optionsfor each. They divided these into institutionaloptions (what their own organization could do) andoptions for the community (what the community asa whole could benefit from).• Identify the main constraints preventing the implementationof these options. In this case, every interviewidentified financial constraints as the singlemost important factor. Most also identified the lackof knowledge, information, and skills as an additionalconstraint.Overall, the results from each of these interviews werenot surprising. The hazards and risks that people identifiedwere in all cases consistent with the results fromrisk mapping, using national-level data. People werethus well-informed about the risks faced by theircommunity. In general, the results from the other questionswere also consistent across interviews, with peoplelisting exactly those adaptation options for each of therisks that had been identified in the PSD workshops.Beyond this, there were several other interesting findingsfrom the interviews. Table 9 lists each of them,

34 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U Enoting the people identified as most vulnerable, andinteresting lessons—with respect to risks, sensitivity, oradaptation options—that stood out from their otheranswers.It is clear that the people identified as most vulnerableare the same, by and large, as those identified in thePSD workshops. Unfortunately, the institutional interviewsdid not shed any particular light on the nature ofvulnerability among these groups.ResulTs fRoM foCus gRouPDIsCussIonsThe focus group discussions had the same intention: toidentify particularly vulnerable social groups, and tobrainstorm about potential adaptation options. Unlikethe institutional interviews, they were held not with keydecision makers in each community, but rather withpeople representing a particular social group withmedium to high vulnerability. The format that theyfollowed was:• Social and demographic characteristics of the group.Discussions started with those in attendance, andthen moved to the members of the same socialgroup in the wider community.• Social differentiation. Participants described the roleof the group in the overall structure of thecommunity.• Natural disasters. Discussion then moved on to themajor disasters in the region. People described thedisasters, when they had occurred with what consequences,and why they believed they happened.• Climate hazards and vulnerability. Discussion movedfrom natural disasters in general to focus on climateTable 9. InTeResTIng lessons fRoM The foCus gRouP InTeRVIewsLocation Participants Especially vulnerable Interesting lessonsAngoche1 ngo andgovernmentsubsistence farmers, singlemothers, children, elderly• hazards result from witchcraft• need to protect forest areas and trees to protect against drought andstorm surges• need to avoid living right at the coastchibabava 5 ngos none listed • need to build water storage damschokwe 3 ngos Elderly, children, widows,pregnantgondola government not listed • there is no disaster planninginhambane government orphans, elderly, physicallyhandicapped, chronicallysickmabote4 ngos,governmentElderly, widows, children,pregnant women• should live on high ground for floods, farm on low ground for droughts• stay away from the dams, where flooding is more common• stock surpluses to prepare for drought, sell them in times of drought• trying to resettle people on higher ground• need better infrastructure• need better access roads• need to conserve soil bettermanhiça six ngos Elderly and orphans • need more early warning, more diversification of income• people attribute disasters to ancestorsmassingir 4 ngos Elderly, children,handicapped, singlewomen, widows• need to stock surpluses and sell them• floods happen when dam sluice gates open, and can be forecast• need to diversify income• need to build irrigation systemsmatutuine government not listed • need dikes and small dams, and more informationmoatize government not listed • need better informationmopeia 2 ngos not listed • Early warning• Better organization of farmers, for risk management committees andto distribute surpluses more effectivelyxai xai 2 ngos none listed • need to store surpluses• Build irrigation systems• stronger houses to protect against cycloneszavala 5 ngos widows, single mothers,elderly, handicapped• farm low-lying areas, live on high ground• need to build more solid houses because of cyclones

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s35hazards and vulnerability. People described the factorsthat made them more or less vulnerable, andthe reasons why these factors mattered.• Adaptive and coping capacities. Participants describedthe steps that people have taken to adapt to andcope with climate hazards, and the factors that havehelped or hindered them from doing so. In this section,participants identified some of the institutionsthat have played a key role.• Long-term adaptation. From the discussion on adaptiveand coping capacity, participants then moved todiscuss possible long-term adaptations. What arethey? Who could implement them?• Barriers to implementation. Participants describedthe factors that could stand in the way of implementingthe long-term adaptations they haddescribed.Table 10 presents results from across all focus groups. Itconsiders six themes, omitting only the characteristics ofthe particular groups. It presents answers to three questions.First, what were the most common answers andissues identified across all focus group discussions?Second, to what extent did these answers vary accordingto geographical factors, such as exposure to particularhazards? Third, to what extent did these answers varyaccording to different stakeholder groups; that is, did thefarmers give the same answers as the fishers and thecharcoal collectors?These results are completely in line with those fromthe PSD workshops, which in turn were in line withpast studies and assessment. The important finding,however, is the general lack of variance in responses,either geographically or across social groups. Also, therespondents were quite clear that the most vulnerablewere that way because they lacked the capacity to work.To supplement the results in Table 10, Annex 1 presentsnoteworthy comments and observations made in eachof the groups. These begin to paint a picture of people’sstories of vulnerability.ResulTs fRoM householD suRVeYsThe final set of data collected during the fieldwork wasin the form of the household surveys. The threeenumerators interviewed 136 households across all ofthe districts where fieldwork was conducted. In eachdistrict, the intention was to interview householdsTable 10. aggRegaTe ResulTs fRoM foCus gRouP DIsCussIonsSocialdifferentiationNatural disastersClimate hazardsand vulnerabilityAdaptive andcoping capacitiesLong-termadaptationBarriers toimplementationMost common responses Geographic heterogeneity Social heterogeneityin most communities subsistence farmerswere listed as the major social group, withlivelihood diversification into forestry andfishingdroughts were the main concern, followedby floods and cycloneschildren, female-headed households, theelderly, and handicapped people are themost vulnerable, because of their limitedability to workin most cases people try to cope and adapt,but have very little assistance and supportpeople listed the same set of adaptationoptions identified in psd workshopsAll groups listed the inadequacy of financeand technical supportdiversification of strategiesdepended on location, andwhether suitable for fishing orforestryparticipants listed hazardsaccording to their exposure,consistent with risk mapsnonein some regions there are socialsupport networks, such asfarming or fishing cooperatives;these provide helpAreas heavily reliant on fishingand forestry were more likely tolist better natural resourcemanagementnonenonenonenonefishermen seemed most inneed of cyclone earlywarning; farmers in floodplains needed warning ofupstream dam releasessubsistence farmers weremost likely to mention incomediversificationnone

36 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U Erepresenting the range of anticipated vulnerability. Inthis section, we first present the important raw data ingraphical form. We then go on to analyze importantrelationships. Annex 2 contains a copy of the surveyinstrument.household size, Resource access, and selfsufficiencyThe first way of describing the data is in terms ofhousehold size and the access to resources. Figure 14shows the distribution of household sizes. The largesthouseholds had fifteen members, while the majority ofhouseholds were between four and seven people, withseven being the most frequent household size.The next data are the household’s access to farmland,and the sufficiency of the farmland to meet the household’sneeds. Figure 15 shows the distribution of fieldsizes in highland and lowland areas. Field sizes tend tobe somewhat smaller in lowland areas, where close to 50percent of respondents had less than 1 ha of land tofarm. In highland areas, the majority of people hadbetween 1 ha and 5 ha of land to farm. Figure 16 showsthe pattern of people bringing their harvest to market.Almost all respondents kept at least 40 percent of theirharvest for household consumption. A quarter ofrespondents took 20 percent or less of their harvest tomarket. Finally, Figure 17 shows how long the harvestskept for the household typically would last. Themajority of households reported the stocks lastingbetween three and six months. Only about 10 percentreported their stocks lasting throughout the entire year.The picture that emerges is of households that havemany mouths to feed, and in general do not growenough grain on their own to do so. Most householdskeep the majority of their harvest for their ownconsumption, and yet in only a few cases does thatharvest last the entire year.fIguRe 15. PRoPoRTIon of fIelD sIzes InhIgh anD lowlanD aReasProportion of Respondents0.350.300.250.200.150.100.050.00Note: n = 136.0 0.1–0.25 0.26–1.0 1.1–2.0 0.1–5.0 5.1–10.0 >10Highland fieldsLowland fieldsfIguRe 14. DIsTRIbuTIon of householDsIzesProportion of Respondents0.200.150.100.050.00Note: n = 136.1 2 3 4 5 6 7 8 9 10 11 12 13 14 15fIguRe 16. CuMulaTIVe DIsTRIbuTIon ofCRoPs kePT foR householD ConsuMPTIonProportion of Respondents1.000.750.500.250.000 25 50 75 100Note: n = 117.

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s37fIguRe 17. nuMbeR of MonThs ThaThouseholD gRaIn sToCks fRoM ownhaRVesT lasTfIguRe 18. PRoPoRTIons affeCTeD bYClIMaTIC hazaRDs anD ReCeIVIng eaRlYwaRnIng of TheseProportion of Respondents0.750.500.250.00upto 3 3–6 6–9 9–12 12 or moreNote: n = 117.NoYesNote: The left-hand circle represents the relative numbers of respondentssaying they have been affected by a climatic hazard. The right-hand circlerepresents the numbers who have received early warning of those hazards.n = 117.NoYesexperience with Climate hazardsThe next data pertain to the climate hazards thatpeople face. As Figure 18 shows, almost all respondentsreported being negatively affected by one or moreclimate hazards. About half of them had received priorwarnings about these, while half of them were caughtrelatively unaware. Figure 19 suggests which hazardsthey view as the most severe for them. Over threequartersof respondents listed droughts as the mostimportant climate-related hazard they face, withsmaller numbers listing floods and cyclones as the mostsevere. Half of the respondents list floods as either themost or second-most important hazard they face. Alarge number list fire and erosion—indirect climateimpacts—as the third most important threat they face.These results are consistent with what was learnedduring the PSD workshops, and indeed is commonwisdom about Mozambique’s vulnerability to climatechange.Respondents listed the ways they had been affected bythese hazards, and in an open-ended question listed theeffects they were most concerned about. These were:• Destruction and loss of property (56 respondents)• Hunger and lack of food (20 respondents)• Low agricultural production (15 respondents)• Degraded condition of the fields (10 respondents)fIguRe 19. RankIng of hazaRDs bYResIDenTs100%75%50%25%0%FirstSecondThirdDrought Flood Cyclone Fire ErosionNote: Proportion of respondents ranking particular hazards as the first, second,and third most important. n = 117• Lack of water in the wells (eight respondents)• Lack of rain (seven respondents)• Being left exposed to the elements (six respondents)• Panic and suffering (five respondents)• Death and physical damage (four respondents)• Forced relocation (one respondent)• Limited fishing activity (one respondent)The next question on the survey asked respondentswhether there was variation in the effects that different

38 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U Ehouseholds in the community had suffered as a result ofclimate hazards; 80 percent responded that there is variation.When asked to identify the factors influencingsuch differences, their answers were:• House construction, i.e., traditional versus conventionalmaterials (44 respondents)• Income level of household (28 respondents)• Being a member of an especially vulnerable group,e.g., elderly, children, female- headed households,handicapped (11 respondents)• Diversification into other economic activities (ninerespondents)• Geographic factors, such as location of farm onhighlands or lowlands (two respondents)Respondents then were asked if their own householdhad suffered differently from most other households inthe community from climate hazards. Again, 80 percentof the respondents said yes, they had suffered differentlyfrom other households in the community. In this case,the reasons most commonly listed were:• Their own income level, compared to that of thecommunity (34 respondents)• Their prevention efforts against climate hazards (23respondents)• Their income diversification (eight respondents)• The quality of their roof (eight respondents)• Being a member of a vulnerable group (fiverespondents)What can we learn from examining these two sets ofanswers? The first is that two sets of factors are themost important determinants of vulnerability: people’slevel and diversity of income, and the extent to whichthey have prepared themselves for extreme weather,such as through the quality of their houses. When viewingthe matter in the abstract, people were more likelyto list preparations as more important. When thinkingabout it in terms of their own personal experiences, theywere more likely to list income as the most importantfactor. This could reflect an unwillingness to takepersonal responsibility for having suffered more thanother households because of the failure to take precautionarymeasures. In terms of whom they have gone tofor assistance after climate hazards, 39 percent ofrespondents indicated that they turned to the localgovernment authorities, 32 percent to the traditionalleaders and authorities, and 29 percent to other organizations,including INGC and NGOs.Past adaptation and Coping PracticesThe heart of the survey investigated households’ adaptationcoping practices. Two open-ended questions askedrespondents to list their primary coping strategies for arange of climatic hazards. Figure 20 shows the results,in terms of the strategies identified most frequently toprepare for hazards and to cope with hazards duringand after their occurrence.To prepare for drought, about a quarter of the peopledid not identify any coping strategy they had implementedin advance. Since almost all respondents listeddrought as a major concern, this could simply indicatethat they did not see options available. Among strategies,the most common were planting crops in the wetter(and sometimes irrigated) lowlands, planting shorterseason(i.e., more drought tolerant) crop varieties, andimproving their buildings. The point about improvingtheir buildings does not make obvious sense, but couldinclude the construction of granaries in order to storemore surplus harvest. An additional ten different strategieswere mentioned, but in each case by only one ortwo respondents: these constitute “other.” These includedpreparing for fires, hunting rats, engaging in more weeding,and engaging in religious practices. During and afterdroughts, the three most common strategies were toplant any new crops in the wetter lowlands, to manageforest resources carefully in order to obtain income fromthose forests as a safety net, and to manage past surplusharvests and cash receipts carefully. The majority ofrespondents, however, suggested that they did nothing.The “other” strategies again included a wide variety ofmeasures, including spending more time in the shade.A larger fraction of respondents do not prepare forfloods, likely because many of them do not face a floodrisk in their district. Of those who do prepare, the mostcommon preparations were to plant in the highlands, tofortify their houses, and to plant short-season varieties.Why short-season varieties? On the one hand, thesewould be more appropriate for highland farming. On theother hand, in the floodplain these would be more likelyto be harvested before the flood hits. During and after a

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s39flood, most people listed nothing that they can do. Theonly common strategy listed was to plant in the highlands,while a number of other strategies enlisted thesupport of only one or two respondents. These includedbuilding canoes and keeping belongings in safe places.The pattern of preparation for cyclones was very similarto that for floods, albeit with fewer additional strategiescovered by the “other” category, and more people listingthe planting of shorter season crop varieties, in order toimprove the chances of gathering a rainy season harvestfIguRe 20. CoPIng MeChanIsMs befoRe,DuRIng, anD IMMeDIaTelY afTeR ClIMaTIChazaRDsDroughtbefore the cyclone hits. Over three quarters of respondentslisted nothing to do during and immediately aftera cyclone. The three most frequently listed strategieswere to plant short-season crops in the highlands, and togather wild fruits to make up for the lack of a harvest.One important question is whether the propensity, orability, to carry out these adaptation and coping strategiescorrelates with income or other resources. The finalsection of the survey asked participants to list whether amember of their household worked either for paymentin cash or in kind (61 percent of respondents answered“yes”) and whether the household had any monetarysavings to pay for special emergencies (55 percentanswered “yes”). We examined whether the answers toeither of these two questions significantly correlatedwith the six sets of results presented in Figure 20. Usinga X 2 test, we found no significant correlations (p > 0.10in all cases). In all cases, similar proportions of thosewith and without employment income, or with andwithout savings, engaged in adaptations, and the adaptationsin which they engaged were similar.OtherImprovehousingNothingConsumewIld fruitFloodCyxloneShort season cropsManage forestresources carefullyFarms in highlands/lowlandsManage surpluscarefullyWe repeated this analysis using not income or the presenceof cash reserves as an indicator of income orresources, but rather other factors. First, we looked atthe productivity of the farm. Figure 17 indicated thenumber of months that households’ harvests lasted; weassumed that those households where the harvest lastedless than four months had low productivity, those wherethe harvest lasted between five and eight monthsmedium productivity, and those lasting nine or moremonths high productivity. Again, using a X 2 test, wefound no significant correlations between this and theadaptation strategies that households reported making(p > 0.10 in all cases). Second we looked at the educationlevel of the household head, coding them as beingwithout schooling, having completed only some primaryschool, and having completed all of primary school ormore. Again, there were no significant correlations.Third, we repeated the test based on household size as ameasure of income: perhaps larger households wouldhave more hands available to engage in the adaptations.In this case, we did find one marginally significant relationship(p = 0.09), namely between household size andflood preparations. Households with more than sevenmembers were twice as likely as smaller ones to engagein some sort of adaptation; at the same time, there was a

40 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U Edifference in the favored adaptation strategies, amongthose who did adapt. Larger households were much lesslikely to move their farming operations to the highlands,and much more likely to engage in all of theother adaptation strategies listed. Presumably, it is hardto move a very big household. All other relationshipswere insignificant (p > 0.10 in all cases).Planned adaptation and strategiesThe survey asked people what, if anything, they woulddo if the climate hazards in their regions were tobecome significantly more severe. The most commonanswers were:• Nothing (70 respondents)• Move to a safer or more productive area (23respondents)• Seek help from others (nine respondents)• Raise and sell animals (seven respondents)• Improve the durability of the house (sixrespondents)• Practice drought-resistant cultivation (fiverespondents)Other ideas, given by three or fewer respondents each,included seeking employment, hunting, and eating less.These results suggest that most respondents do not havea clear picture of what they would do to adapt toclimate change, and of those who do, the mostfrequently listed option was migration.When asked whether they felt that their household wasmore or less able to engage in longer-term adaptations,compared to other households in the community, 48percent thought they were more able, 27 percent thoughtthey were equally able, and 25 percent thought they wereless able. This result—only 25 percent thinking they areless able than average to adapt—is broadly consistentwith findings of overconfidence in the psychological literature,such as the fact that 90 percent of people think thatthey are above-average drivers. There was no significantcorrelation between the belief of being more or less ableto adapt to climate change with any of the previouslyconsidered indicators of adaptive capacity: household size,number of months the harvest would last, education ofthe household head, paid employment, or cash savings foremergencies (p > 0.10 in all cases, X 2 test).who should help?Respondents gave their opinion on which organizationscurrently play the more important role in helpingto mitigate natural disasters. Forty-six percent ofrespondents suggested that it was agencies of thenational government, such as INGC or INAS, while31 percent thought it was local government or traditionalauthorities, and 23 percent thought that it wasan NGO, such as the Red Cross or World Vision.These organizations are providing assistance such asdirect support (food, clothes, money), early warning,and education. What do people need more of fromthese organizations? The most common response(from 39 percent of respondents) was more directsupport, followed by early warning (13 percent ofrespondents), awareness-raising (7 percent), andimproved provision of basic services, such as schools,roads, and hospitals (5 percent).synthesis of quantitative findings from householdsurveyThe household survey offered results in most waysbroadly consistent with the results from the PSDworkshops, and from the institutional interviews andfocus group discussions. People in general were worriedabout the same climate-related hazards, viewingdroughts, floods, and cyclones as the most severethreats to their livelihoods, in that order. They wereengaging in a variety of coping mechanisms, the samecoping mechanisms identified earlier: cultivatingshorter season varieties of crops, switching their cultivationbetween highlands and lowlands depending onthe relative threat of flooding and drought, reinforcingtheir buildings, and managing their other resourcesmore wisely. Very few, interestingly, listed workingmore for money as a strategy they were engaging in,perhaps because there are so few such jobs. In all cases,there seemed to be more options available to morepeople to prepare for a hazard ahead of time, ratherthan during or after. This would suggest the need forimproved early warning, something identified in thePSD workshop as important, and by the surveyrespondents as something they would like more of. Ofcourse what they really need and want is money, food,and clothing; this is consistent with a lack of resourcesbeing the greatest obstacle to coping and adaptation.

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s41What the survey did not show was a significant relationshipbetween the hypothesized—and in the literature,previously identified—determinants of adaptive capacity,and people’s propensity to have engaged in adaptation orcoping options, or their perceived ability to engage insuch options in the future. Only one relationship wasmarginally significant, between household size andcoping measures, with larger households showing adifferent pattern than smaller ones. But given 35 relationshipsinvestigated, chance alone could account forone of these appearing to be significant at the 10 percentconfidence level. It is certainly not a strong finding. Thelack of significance could be because the sample size wastoo small, because the survey was not ideally designed orpoorly administered, or simply because differences in thedeterminants of adaptive capacity within the communitiessurveyed simply do not matter very much.Finally, the survey paints a fairly bleak picture of futureadaptation at the household level. The majority ofhouseholds do not have any idea of how they wouldadapt to more severe climate hazards, and one of themost common strategies identified by those who didhave an idea—asking others in the community forhelp—would probably not work very well given thateveryone in the community would be suffering. Themost common strategy identified was migration. This isconsistent with the literature on adaptation, and yet it isa highly disruptive act for any household.sYnThesIs of fIelDwoRk ResulTsPut together, the three sets of data from the fieldworkvalidate the results from the PSD workshops, whichwas their primary intention. There were a few additionalnew findings:• The institutional interviews and focus group discussionsboth revealed very little regional heterogeneity,and the focus group discussions revealed verylittle heterogeneity according to social group.• The household surveys revealed no significant differentialvulnerability, on the basis of commonlyused indicators, such as household assets, education,or income.• The household surveys revealed a very low awarenesson the part of most people of how they wouldadapt to future climate change, should that bringwith it an increase in the frequency or severity ofnatural hazards.The last of these points is consistent with an importantconclusion from the PSD workshops, namely that thereis an urgent need for awareness raising and educationabout climate change, and options for adapting toclimate change, among subsistence farmers.

427. sYnThesIs anD DIsCussIonThis section consists of three steps. First, we discuss themain adaptation options gleaned from this research,comparing them with options that were suggestedbefore, such as in the NAPAs. In so doing, we addressthe first question: Does a development-first assessmentprocess result in different findings from a more traditionalimpacts approach? We also compare theseconclusions with the general conclusions from theeconomic track of the EACC study. Second, we look atthe issue of differential vulnerability. Who are the mostvulnerable, why, and what interventions can best assistthem? Third, we evaluate some of the weaknesses andlimitations in this study, and the extent to which ourresults on either of the first two questions may be sensitiveto these weaknesses. Finally, we offer policyrecommendations.aDaPTaTIon oPTIonsThe PSD workshops generated a list of adaptationoptions, described in Table 8. The fieldwork generatednew results that in most cases left this list of optionsintact, but in some cases changed them. Table 11 presentsthe revised list.As with Table 8, the options in plain text responddirectly to climate hazards, while those in italics representmeasures to increase the adaptive capacity of thepopulation, or to make them more resilient to shocks totheir livelihoods in general. Changes from Table 8 toTable 11 include the addition of wells for drinkingwater, the construction of hospitals and schools, bettersensitivity to how farmers can hedge their risks by farmingboth highland and lowland fields, and the importanceof wild fruits as a natural resource in times ofstress.All of the options in the left-hand column, beingcentralized, represent potential government interventions.The right-hand column, by contrast, representsmeasures that people can undertake on their own. Theliterature on adaptation often characterizes such actionsas “autonomous” (Patt et al. in press and Aaheim et al.2007), as if they took place on their own without anyparticular agency. This may be a misnomer in a countrylike Mozambique. First, there is a great deal of intentionalitybehind any of these actions. Second, and moreimportantly, many of these represent actions thatpeople do not have the resources or skills to undertakeon their own. This is particularly the case with the hardoptions: they take money. The people who have themoney have already undertaken them, in particularmaking their buildings more robust. But any of the softoptions are also difficult without the right economicenabling environment. Many, for example, would like todiversify their incomes, but there are simply no opportunitiesto do so. Direct government intervention, inthe form of rural economic development, may benecessary.None of these options are inconsistent with the NAPAsand previous planning efforts. What they do accomplish,however, is to suggest a number of specific waysof improving the capacities of farmers to cope withincreasingly adverse weather. They also highlight thefact that some of the most important adaptation optionsare in fact those that commonly would be labeled

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s43Table 11. keY aDaPTaTIon oPTIons fRoM fIelDwoRkhardsoftCentralized• flood control dikes and levies• coastal flood control gates• dams and irrigation channels• Improved roadways• Improved communication infrastructure• Improved hospitals and schools• improved early warning of climatic hazards, and of damreleases• Better planning and management of forest, fish, andother natural resources• Resettlement of populations to lower risk zones• More credit and financial services for small businessesand rural development• Better education and information for rural areas• Improved health care, social services, and social supportfor all peopleDistributed• more robust buildings• farm-scale water storage facilities• deep wells to provide drinking water for people and animals• Grain storage facilities• Improved food processing equipment• Better utilization of short-season and drought-resistant cropsto prepare for drought, floods, and cyclones• diversification of flood and drought risk by maintaining fields inboth highland and lowland areas• Better household and community management and use ofnatural resources, including wild fruits• practice of soil conservation agriculture• migration to lower risk areas• Diversification of livelihoods away from agriculture• Better planning of how much grain to save for personalconsumption, and how much to sell for income generationdevelopment, such as improved finance in rural areasand the provision of social services.They are also consistent with the economic track resultsof the EACC study. That part made several recommendationsfor adaptation strategies:• Development. The main result of the economicstudy was that the most important means of overcomingthe challenge of climate change wasthrough the development of human capital (forexample, education) and the growth of well-functioninginstitutions.• Regional river basin management. Mozambique is adownstream country, and it needs to cooperatemore closely with upstream countries to ensure thebest use of runoff.• Agriculture research and extension. Mozambique’scrop yields are due to decline given a continuationof current practices, but could increase significantlywith better utilization of existing technologies,based on an evaluation of suitability.• Rural roads. Roads are necessary for rural development,and unpaved roads will be washed out morefrequently by the expected increase in extreme precipitationevents.• Zoning and land-use planning. Given the economicgrowth that is projected for Mozambique, it isimportant to plan carefully so that the growthoccurs in low-risk areas.• Careful, and limited, use of hard adaptation options.Hard adaptation options are an enticing way of protectingagainst extreme events. The study foundthat often the costs of these options far exceededtheir benefits, in particular for coastal protection.The social component arrived at remarkably consistentfindings. There were only three inconsistencies. First, thesocial component did not identify regional river basinmanagement as a key priority. This was mentioned inone focus group discussion—farmers complained thatthey were uninformed about upstream dam releases thatcaused flooding for them—but was not highlighted inother areas. Second, the social component did notconsider the costs of hard adaptation options, and so didnot reach a conclusion that they are often inefficient.Third, the social component highlighted the importanceof improving social safety nets. This potentially fallsunder the category of development institutions, but wasnot explicitly mentioned in the economic study.DIffeRenTIal VulneRabIlITYIn all three PSD workshops, and in every institutionalinterview and focus group discussion, people identifiedthe elderly, children, single women and female-headedhouseholds, and the physically handicapped as being themost vulnerable. This is because they lack the ability towork and thus to earn a sufficient income, the skillsnecessary to engage in self-protective action, and the

44 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U Eassets to withstand extreme weather events. What thesepeople need more than anything is an improved socialsafety net, so that they do not use their productive assetssimply to survive. Several interviewees mentioned thatthe existing level of social support—approximately $4per month—is tragically low.It also emerged that the agriculture sector is far morevulnerable than alternative livelihoods. Farmers lose theircrops from droughts most of all, and then from floodsand cyclones. The effects of natural hazards on the agriculturalsector then trickle through into the other livelihoodstrategies. They place additional pressure on forestsand fisheries to feed people when the crops fail, and theyslow down the flow of money and goods in trade.Differential vulnerability failed to show up as expected inthe household survey. While many people in the surveyclaimed—as had the participants in the other activities—that those with fewer resources were unable to engage inself-protective behavior, there were no significant correlationsbetween the indicators of household assets andreported past self-protective behavior. It could be that thesample size was too small, or the survey design inappropriate,or the sampling of households too narrow, tocapture a differential vulnerability effect. Or, it maysimply be that the effects of differential vulnerability arein fact small. It is worth mentioning that other studieshave also found that income or assets are not goodpredictors of adaptive behavior, except when they lead tohouseholds acquiring additional information or skills(Phillips 2003; Pratt et al. 2005; Grothmann andReusswig 2006; and Grothmann and Patt 2005)Addressing differential vulnerability—if indeed it is amajor problem—needs to be part of the developmentagenda, simply because that agenda needs to focus oninterventions that benefit the poorest of the poor. Infact, the last few years have witnessed this focus takinghold among development agencies (Gupta et al. 2009).Hard adaptation options will not achieve this, but socialdevelopment, including strong safety nets, will.lIMITaTIons of The sTuDYThere are some important limitations to this study, aswell as lessons learned about how to—and not to—conduct a country case study such as this.First, there has been an insufficient review of thecurrent institutional context of adaptation research andplanning in Mozambique. While this report hasreviewed the NAPAs, it has omitted detailed descriptionsof some other important programs under way.SETSAN and FEWS.NET, for example, are engaged ina great deal of vulnerability and livelihood mapping,with the intention of using this information to reducefood insecurity in the face of increasing climate hazards.Originally it had been envisioned to interview nationallevelstakeholders, and to analyze these programs ingreater depth. Given unforeseen delays in the studyprocess, this did not take place. Since this review of theinstitutional context was not seen as a critical elementof this report—it describes features unique toMozambique and is difficult to generalize to othercountries—resources were redirected to the more participatoryareas of the project. As a result, this componentof the report is underdeveloped.Second, there were inconsistent approaches takenduring the PSD workshops. These had a valid justification,namely to observe whether variance in the PSDworkshop design led to a difference in substantiveoutcome. To a large extent these changes in format werethe result of post-workshop team discussions, in orderto “tweak” the workshop design for the next one.However, the Xai Xai format—a longer workshop, withdiscussions centered on activities and not geographicalregions—generated the greatest depth of discussion.Thus, some of the most interesting findings are from asingle workshop, not well-corroborated by the otherworkshops.Third, the quality of some of the data is questionable.In the case of the PSD workshops, arguably the mostimportant single workshop was that held in Maputo,with national-level stakeholders. In fact, challenges withtiming meant that very few high-level stakeholdersfrom the national government attended the Maputoworkshop. The majority of attendees were representativesof NGOs operating in Maputo, some of them withvery limited knowledge of adaptation policies and priorities.In the case of the institutional interviews andfocus group discussions, the field notes that served asthe basis of this report omitted a great deal of detail ofthe discussions. For example, the local consultant’s fieldreport (Annex 3) suggests that participants engaged in

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s45wealth-ranking exercises to identify the links betweensocial groups and adaptive capacity. The field notesstated that such wealth-ranking exercises took place, butdid not provide information about the results. In thecase of the household survey, the sample size was toosmall to observe significant statistical relationships.Adding to this, there were many apparent problems indata collection and entry. This consultant simplydropped the unexplained answers from the statisticalanalysis, but if the unexplained answers correlated withother variables, then this would have introduced biasinto the results.Fourth, the depth of analysis of the field data is not asgreat as it should be. Delays in the scheduling of theworkshops and the fieldwork led to very little time beingavailable for the analysis of the data generated there.How serious are these limitations? Not very; indeed,there is good reason to believe the core findings of thisstudy would be the same even if each of the problemshad been resolved. This is because they are so consistentwith past assessments, and with other results within theEACC study. This agreement provides reason to believethat improvements in data and analysis would not likelyhave led to qualitatively different findings.ConClusIonsFinally, it is useful to condense the many findings into acore set of key conclusions. These are:1. Rainfed agriculture takes the hardest direct hit fromclimate hazards. Across the focus group discussions,institutional interviews, and workshops, participantsmentioned climate impacts affecting avariety of livelihood activities, including agriculture,fishing, forestry, and charcoal production. Butin all cases, the most frequent and severe impactswere listed for rain-fed agriculture because ofdroughts. The most frequently mentioned way ofreducing these impacts was the construction of irrigationsystems. The most frequently listed barrierto this was the lack of finance. Likewise the leastvulnerable were those households that had managedto diversify their income away from just agriculture,such as through informal trade or thereceipt of remittances from family members livingin urban areas or in South Africa. But diversificationisn’t easy: many people suggested that therecipients of micro-finance ended up using thatmoney just to buy food, because the opportunitiesfor small businesses succeeding were not very good.This suggests that attention to integrating ruralareas into markets—including a great deal of attentionto improving transportation and communicationinfrastructure—will be an important activity.2. Subsistence farmers and the economically andsocially marginalized are the worst off. Across allfocus group discussions, institutional interviews,and workshops, participants identified subsistencefarmers—and among them the economically andsocially marginalized individuals—as the mostvulnerable and having the fewest resources availableto adapt. Economically and socially marginalizedindividuals include the elderly, orphans,widows and female heads of households, and thephysically handicapped. Most communities arelacking support networks for these people, eitherformally through the government or informallythrough well-functioning social networks. Thegovernment assistance that the particularly needyreceive is inadequate to help them make it throughadverse weather events. Social safety nets need tobe improved.3. Policies and institutions should enforce sustainableresource management and wise land-use planning.In many cases, participants in the discussions andworkshops suggested that the harvesting of forestresources—such as wood for charcoal production—aswell as fishing were important incomegeneratingactivities, which often helped to buffershocks to agricultural productivity. But theseactivities are suffering due to deforestation andoverfishing. Better management is needed. Thiscan include the enforcement of existing laws andgovernment policies, as well as improving thecapacity of community associations to managelocal resources effectively. Planning is also neededto make sure that activities gradually move fromthose areas of high risk to areas of lower risk.4. Knowledge is needed across the board. Lack ofknowledge is a problem in several respects. First,there appeared to be a great deal of confusionabout the causes of natural disasters and whetherthey would increase or decrease in the future.

46 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U EEducation is needed so that people can expectdisasters to be a constant feature in the future.Second, people need technical assistance concerningbetter land management, such as conservationagriculture. Third, people need access to real-timeweather forecasts—effective early warning—tomitigate their losses to floods and cyclones. Insome cases, they also need information about whendam operators upstream will be releasing water, sothat they can prepare for the local flooding thiscauses. Fourth, and most importantly, people needmore education in general, if they are to diversifytheir livelihoods away from agriculture and towardhigh value-added activities that can enhance theirwell-being and provide them the resources towithstand extreme weather.

47RefeRenCesAaheim, A., et al. 2007. “Why we will need adaptationand how can it be implemented.” In ADAM / CEPSPolicy Brief. Norwich, UK, and Brussels: ADAM Projectand Center for European Policy Studies.Adger, W.N. 1999. “Social vulnerability to climatechange and extremes in coastal Vietnam.” WorldDevelopment. 27 (2): 249–269.Asante, K., et al. 2009. Study on the impact of climatechange on disaster risk in Mozambique: synthesis report.Maputo: National Institute for Disaster Management.Bengtsson, L., et al. 2007. How may tropical cycloneschange in a warmer climate? Tellus, A. 59 (4): 539–561.Boko, M., et al. 2007. “Africa.” in M.L. Parry et al., eds.Climate change 2007: Impacts, adaptation and vulnerability.Contribution of Working Group II to the FourthAssessment Report of the Intergovernmental Panel onClimate Change. Cambridge UK: CambridgeUniversity Press.Brooks, N., and W. N. Adger. 2004. “Assessing andenhancing adaptive capacity.” Adaptation PolicyFramework. Geneva: United Nations DevelopmentProgram.Brooks, N., W. N. Adger, and P.M. Kelly. 2005. “Thedeterminants of vulnerability and adaptive capacity atthe national level and the implications for adaptation.”Global Environmental Change 15 (2): 151–163.Carmo Vaz, A. 2000. “Coping with floods: the experienceof Mozambique.” First WARFSA / WaterNetSymposium: Sustainable use of water resources. Maputo:WARFSA/ WaterNet.Center for Research on the Epidemiology of Disasters.2008. The OFDA/CRED international disaster database.Université Catholique de Louvain. Available at: .Comité de Conselheiros. 2003. Agenda 2025: Visão eEstratégias da Nação. Comité de Conselheiros. Maputo:República de Moçambique.Cutter, S. 2001. A Research Agenda for VulnerabilityScience and Environmental Hazards. InternationalHuman Dimensions Program Update 01 (2): 8–9.Emanuel, K., R. Sundararajan, and J. Williams. 2008.“Hurricanes and global warming: results from downscalingIPCC AR4 simulations.” Bulletin of the AmericanMeteorological Society 89: 347–367.Ehrhart, C., and M. Twena. 2006. “Climate Change andPoverty in Mozambique: Realities and response optionsfor CARE.” Background report. Geneva: CAREInternational.Eriksen, S., et al. 2008. “Vulnerability assessments in thedeveloping world: Mozambique and South Africa.” InA.G. Patt et al., eds. Assessing vulnerability to globalenvironmental change. London: Earthscan.FEG Consulting and Save the Children. 2008. Thepractitioner’s guide to the household economy approach.

48 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U EJohannesburg: Regional Hunger and VulnerabilityProgram.Ferguson, J. 2005. Disaster risk management along the RioBúzi. Case study on the background, concept, and implementationof disaster risk management in the context of theGTZ program for rural development (PRODER).Eschborn, Germany: German Agency for InternationalCooperation (GTZ).FEWS-NET. 2002a. Food economy baseline profile: AltoLimpopo. Maputo: Famine Early Warning SystemNetwork.FEWS-NET. 2002b. Food economy zones of Mozambique.Maputo: Famine Early Warning System Network,Ministry of Agriculture and Rural Development, andthe National Directorate of Agriculture.FEWS-NET. 2002c. Food economy baseline profile:Coastal Nampula and Southern Cabo Delgado. Maputo:Famine Early Warning System Network.FEWS-NET. 2008a. Livelihood profile: Chinde Delta,Zambézia, Mozambique (Luabo Sub-zone). FEWS-NET:Washington, DC.FEWS-NET. 2008b. Livelihood profile: NorthernNampula coastal (agricultural) zone. FEWS-NET:Washington DC.FEWS-NET. 2008c. Livelihood profile: SouthernNampula coastal (agricultural) livelihood zone. FEWS-NET: Washington DC.FEWS-NET. 2008d. Livelihood profile: Zambézia coastallivelihood zone. FEWS-NET: Washington DC.FEWS-NET. 2008e. Zambezi Riverine Livelihood ZonesProfile. FEWS-NET: Washington DC.Grothmann, T., and A. G. Patt. 2005. “Adaptive capacityand human cognition: the process of individual adaptationto climate change.” Global Environmental Change15 (3): 199–213.Grothmann, T., and F. Reusswig. 2006. “People at risk offlooding: why some residents take precautionary actionwhile others do not.” Natural Hazards 38 (1–2): 101–120.Gupta, J., et al. 2009. “Mainstreaming climate change indevelopment cooperation policy: conditions for success.”In M. Hulme and H. Neufeldt, eds. Making climatechange work for us. Cambridge, UK: CambridgeUniversity Press.International Monetary Fund and InternationalDevelopment Association. 2001. Joint staff assessment ofthe poverty reduction strategy paper. Maputo:International Monetary Fund and InternationalDevelopment Association.International Monetary Fund. 2007. Republic ofMozambique: Poverty Reduction Strategy Paper—JointStaff Advisory Note. Washington DC: InternationalMonetary Fund.Lucio, F., A. Muianga, and M. Muller. 2007. “Floodmanagement in Mozambique.” In M. Hellmuth, etal., eds. Climate risk management in Africa: learningfrom practice. New York: International ResearchInstitute for Climate and Society (IRI), ColumbiaUniversity.Mavie, A. 2003. Battling nature in Mozambique.USAID/FEWS NET MIND: 2.McDonald, R., et al. 2005. “Tropical storms: representationand diagnosis in climate models and theimpacts of climate change.” Climate Dynamics 25 (1):19.MICOA. 2003. Mozambique initial national communicationto the UNFCCC. Maputo: Ministry forCo-Ordination of Environmental Affairs.Mirza, M. M. Q. 2003. “Climate change and extremeweather events: can developing countries adapt?”Climate Policy 3: 233–248.Muianga, M. 2007. Sustainable land use planning forintegrated land and water management for disasterpreparedness and vulnerability reduction in the LimpopoBasin. Maputo: United Nations Human SettlementsProgram (UN-HABITAT).Nakicenovic, N., and R. Swart, eds. 2000. Special reporton emission scenarios. Geneva: Intergovernmental Panelon Climate Change.

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s49Nkomo, J.C., A. O. Nyong, and K. Kulindwa. 2006.“The Impacts of Climate Change in Africa.” Papersubmitted to The Stern Review on the Economics ofClimate Change.Osbahr, H., et al. 2008. “Effective livelihood adaptationto climate change disturbance: scale dimensions ofpractice in Mozambique.” Geoforum 39: 1951–64.Patt, A.G., et al. In press. “Adaptation in integratedassessment modeling: where do we stand?” ClimaticChange.Patt, A. G., et al., eds. 2008. Assessing Vulnerability toGlobal Environmental Change: Making information usefulfor adaptation policy and decision-making. London:Earthscan.Patt, A. G., and P. Nussbaumer. 2009. Vulnerability toclimate change up to 2060 in Mozambique. Laxenburg,Austria: International Institute for Applied SystemsAnalysis.Patt, A.G., and D. Schröter. 2008. “Perceptions ofclimate risk in Mozambique: implications for thesuccess of adaptation and coping strategies.” GlobalEnvironmental Change 18: 458–467.Patt, A.G., P. Suarez, and C. Gwata. 2005. “Effects ofseasonal climate forecasts and participatory workshopsamong subsistence farmers in Zimbabwe.” Proceedings ofthe National Academy of Sciences of the United States ofAmerica 102: 12623–12628.Phillips, J. 2003. “Determinants of forecast use amongcommunal farmers in Zimbabwe.” In K. O’Brien and C.Vogel, eds. Coping with climate variability: the use ofseasonal climate forecasts in southern Africa. Aldershot,UK: Ashgate.República de Moçambique. 2001. “Action Plan for theReduction of Absolute Poverty (PARPA) 2001–2005.”In Poverty Reduction Strategy Paper (PRSP). Approvedby the Council of Ministers. Maputo, Moçambique.República de Moçambique. 2006. “Action Plan for theReduction of Absolute Poverty (PARPA II) 2006–2009.”In Poverty Reduction Strategy Paper (PRSP). Approved bythe Council of Ministers: Maputo, Moçambique.República de Moçambique. 2007. National AdaptationProgram of Action (NAPA). Maputo, Mozambique:Ministry for the Co-ordination of Environmental Affairs.United Nations Development Program. 2004. Reducingdisaster risk: a challenge for development. New York:United Nations Development Program, Bureau forCrisis Prevention and Recovery.USAID Mozambique. 2008. Thematic briefing paper 1:the global food crisis: a blessing or a curse for ruralMozambican households? Maputo: United States Agencyfor International Development.USAID Mozambique. 2008. Thematic briefing paper 2:just how poor are “the poor.” Maputo: United StatesAgency for International Development.USAID Mozambique. 2008. Thematic briefing paper 3:“Vulnerable” or “At Risk”? What does household food economytell us about disaster risk reduction? Maputo: United StatesAgency for International Development.USAID Mozambique. 2008. Thematic briefing paper 4:What can HEA livelihood baselines tell us about opportunitiesfor development? Maputo: United States Agency forInternational Development.Vincent, K. 2004. Creating an index of social vulnerabilityto climate change for Africa. Norwich, UK: TyndallCenter for Climate Research.Vincent, K. 2007. “Uncertainty in adaptive capacity andthe importance of scale.” Global Environmental Change17 (1): 12.World Bank. 2000. A Preliminary Assessment of Damagefrom the Flood and Cyclone Emergency of February–March2000. Washington, DC: World Bank.Yohe, G., and R. S. J. Tol. 2002. “Indicators for socialand economic coping capacity—moving toward a workingdefinition of adaptive capacity.” GlobalEnvironmental Change. 2: 25–40.

50aPPenDIxesaPPenDIx 1. InTeResTIng ResulTs fRoM InDIVIDual foCus gRouP DIsCussIonsLocationAngocheAngocheAngocheParticipantsElderly andhandicappedfishermenwith boatssubsistencefarmersInteresting lessons on social differentiationand incomethe poorest group comprise 11 percent ofpopulationgovernment support to them of 100metecaix ($4) a month is totally inadequatesocial protection network existsfish are no longer available in largequantitiespeople use mosquito nets for fishing, whichdeplete fish stockspeople help each other to reconstruct after astormfishing and agriculture are main sources ofincomeElderly, single women, widows suffer mostAngoche fishermen fishing and selling fish are the main incomeactivitiesBuziBuzichibabavasubsistencefarmerscharcoalproducerssubsistencefarmersAssociations exist in the communitywidows are the most vulnerable groupincreasing hardship over last 10 yearsAccess to trees to fell is determined byrelevant permitsno cooperativesAgriculture and labor are main incomesourcesneed better inputs for agriculture (tractors,etc.)Interesting lessons on vulnerability and adaptationpeople with more diversified incomes are in a betterpositionthere is one guy with a really strong house, and peoplego to him when cyclones come; more such houses areneededneed to establish a fishing association to manage theresource bettercyclones and strong winds are the big threat, becauseof the destruction to houses, boats, and coconut treesno forecasts of adverse weathermen and young people are better at adaptingneed to stop cutting the mangrovescyclones are biggest hazard, destroying houses andboatswealthier people have built stronger houses that canwithstand stormsneed cash to build better homescyclones lead to hunger and diseasemaybe need to move to a safer areadrought happens annually; locust plague was a recentproblemnobody can adapt to cyclonesnatural disasters would become less common if theycould hold more traditional ceremoniesneed to build dams and irrigation, better homes, andhomes on higher groundContinued on next page

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s51aPPenDIx 1. InTeResTIng ResulTs fRoM InDIVIDual foCus gRouP DIsCussIons (continued)LocationchibabavachibabavachibabavachibutochibutoParticipantssmallbusinesspeoplesubsistencefarmerssubsistencefarmerswidows andsinglemotherscommunityleadersInteresting lessons on social differentiationand incomeno social protection network or associationexists in the communitythis group works in beekeepingsubsistence farmers, elderly, orphans arethe most vulnerableno social protection network or associationsexist in the communityAgriculture and charcoal production are themain sources of incomeno social protection network existsthere are activist associationscommunity controls the firewood resourceno cooperatives in communityno paid workAgriculture and lake fishing are main incomesources, also charcoal productionAgriculture and livestock are main activities.Also fishing and selling rush mattingchókwe widows An association exists for wood harvestingand bakingcharcoal is the main source of incomeorphans and elderly are the most vulnerablewidows represent 30 percent of populationchókwe Elderly men Young men go to south Africa for work,leaving women behind to run the farmRemittances from south Africa are importantwidows represent 30 percent of populationmabote farmers Agricultural production, includingbeekeeping, is the biggest activitymabote Elderly Agriculture, ranching, and forestry are themain income activitiesmaboteold womenand widowsthe elderly are the only ones who cannotadaptAgriculture and charcoal production are themain income-generating activitiesInteresting lessons on vulnerability and adaptationneed to get access to 7 billionpeople can stock up on food to prepare for drought,using money obtained from selling charcoalworking age people have the greatest adaptive capacityclimate hazards occurring because of rapiddeforestationneed to build better houses to prepare for floods andcyclonesRepairing broken water pumps would be a goodadaptationtree planting is neededfuture well-being will either get worse or better, nobodycan predictdesperately need alternative sources of incomeproblems are lack of health post, electricity, and pavedroadsBelieve that deforestation is causing the naturaldisastersBelieve that climate conditions will improve over next 10yearsneed early warningmicrocredit to help people start small businesses anddiversify income, but small businesses are also tough towork, and loaned money is often used to buy foodtechnical assistance is neededforecast of drought and flood has not been forthcomingUnassisted elderly people have the hardest timeadaptingdrought is biggest threatpeople have pulled together after disastersdrought is caused by climate change, which is causedby deforestationthere isn’t even enough drinking waterthey suffer from a lack of mobile telephone coveragewater shortages are so severe that there is no water forpeople and cattle to drink.Better storage of grain would help smooth out the hardtimesContinued on next page

52 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U EaPPenDIx 1. InTeResTIng ResulTs fRoM InDIVIDual foCus gRouP DIsCussIons (continued)LocationParticipantsInteresting lessons on social differentiationand incomemabote leaders A savings association exists that supportselderly and orphansmanhiçaElderlywomen andwidowsthe young people often leave to seek workelsewhereElderly subsistence farmers are the mostvulnerablelittle solidaritylots of people go to south Africamanhiça leaders no social protectionmassingirmassingircharcoalproducersneedywomenfarming, charcoal, and cattle are biggestincome sourcespeople travel to south Africafarming, charcoal, and hunting are mainincome sources; also beekeepingAssociation for tree felling and charcoalproduction existslack of social services is a key problem,especially a health postno social protection networkmatutuine leaders forestry and farming are main activitiesmatutuinemoatizemoatizemoatizemopeiawidows andsinglemotherssubsistencefarmerssubsistencefarmerssubsistencefarmerssubsistencefarmersElderly and orphans are most vulnerablethe worst affected is the entire populationno social protection for the elderlynew association formed to share in farmingworkmain income is working in south Africa andswaziland, farming, and forestryAgriculture and fishing are main incomesourcescooperatives of small businesses andfishermen existAgriculture and fishing are main sources ofincomesubsistence farmers are the mostvulnerable, because they go hungry withdroughtgoing to cities to work has increasedstandard of living – the young can do thisno associations in the communitysmall business owners have highestincomes, subsistence farmers the lowestInteresting lessons on vulnerability and adaptationmain problems are a lack of wells, lack of health care,and lack of roadstree planting for charcoal production is neededdroughts dry up the wells, leaving even no water fordrinkingYou can’t adapt if there isn’t any waterneed to diversify incomeplanting trees to increase charcoal productiondeforestation is a big problem, and seen as causingdroughts and floodinglack of water even for human consumption, as well asfor cattle, because the wells have dried upwithout enough water, no adaptation is possiblepeople travel to south Africa to escape the lack of waterat homeconstructing a dam would be goodnot entirely convinced that hazards result from failure toobserve traditional ceremonieslack of electricity is a problemflooding is annual when dams upstream opens its sluicegates. the dam in south Africa sometimes informs themin advance, the one in swaziland does notneed more traditional ceremonies to ward off droughtpoor access roads and lack of facilities to buy basicproductsno differentiation in terms of adaptive capacityneed more dams, irrigation, and technical assistanceBig problem is being eaten by crocodiles when farminglow-lying landneed to build irrigation canals, and avoid cutting downtrees for charcoalneed better road access, a health center, and policepeople do receive weather forecasts over the radio, andsuffer less than in the pastneed tools, and motor-driven pumps for irrigationno differences between groups in terms of adaptivecapacitypeople try traditional ceremonies to avoid droughts, butthese don’t seem to workBurnings are a big problemflooding is the main hazardcanoes are being builtfarming should be in the lowland during drought periodsContinued on next page

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s53aPPenDIx 1. InTeResTIng ResulTs fRoM InDIVIDual foCus gRouP DIsCussIons (continued)LocationmopeiamorrumbalamorrumbalamorrumbalaParticipantssubsistencefarmerssubsistencefarmerssubsistencefarmerssubsistencefarmersInteresting lessons on social differentiationand incomeno social protection networksAssociations to market food have beensuccessful at getting people income andraising standards of livingAssociation existsAgriculture and casual labor are primarysources of income, with fishing as wellsubsistence farmers the most vulnerableno social protection network in thecommunitymost people are subsistence farmers, whoare the most vulnerablepeople living in stronger built houses do notsuffer as muchVilanculo farmers no associations or cooperativesVilanculozavalazavalazavalacommunityleaderscommunityleaderssubsistencefarmerswidows andelderlyfarmersmost vulnerable are elderly, single mothers,and childrenfraud has led to unequal distribution of foodaidcommunity has suffered a loss of social andmoral valuesno associations in communitychildren suffer mostno cooperativesElderly and children are most vulnerablepeople here don’t travellots of conflicts between familiesAgriculture and ranching the main incomesourcespeople here don’t travelno social solidarity or cooperativesAgriculture is biggest income sourceInteresting lessons on vulnerability and adaptationlack of market and access roads are major concern;there are no people to buy their productshazards occur because of witchcraftproblems from natural hazards will improve in next 10yearshazards occur because of failures of ancestorsfrequent floods and droughtsto protect against drought, farm in the lowlands, toprotect against flood, farm in the highlandcVm has provided the most assistancewant to construct dikes to protect against floods.flooding is the most frequent event, followed by droughtneed to have two machambas, one high and one lowhippos have reduced agricultural production, along withfloodsfloods destroy all their propertysituation would improve if the climate would change tohave fewer natural disasterscyclones and drought also a problemmain problems are lack of information and transportationdrought is main hazardhazards have increased because of failure to followlocal customsdo not receive radio weather forecastsillegal tree felling is a problemlack of organization is the main barrier to communityadaptationfood for work program is mentioned many times asimportantplanting fruit trees helpsno formal forecasts available, but traditional indicatorswomen most able to adaptneed to focus more on raising and selling animalsneed to increase the area under cultivationneed more organization and knowledgecyclones are worst hazard, along with droughtneed more water pumpslack of rain because have failed to do the traditionalceremoniesneed early warningfamilies that receive remittances are most able to adaptBetter employment opportunities are needed

54 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U EaPPenDIx 2. suRVeY InsTRuMenT

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s55

56 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U E

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s57

58 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U E

59aPPenDIx 3. fIelD RePoRTsubMITTeD oCTobeR 5,2009, bYRaul VaRelabaCkgRounDThe purpose of the study was to create an environmentfor understanding economic adaptation to climaticchanges and to generate recommendations for subsequentactions leading to better interventions to mitigateand adapt to the effects of this phenomenon. Particularconcerns are with fragile livelihood groups, mainly thosehouseholds and communities with high exposure tonatural hazards.In order to address such concerns, it becomes importantthat a consistent criteria be adopted in the selectionprocess to visualize geographical zones that allow us tograsp, confidently, representatives of all livelihoodgroups followed by an in-depth analysis of their level ofadaptation to the effects of climatic changes, formal andtraditional interventions, level of resilience, interventiongaps and cost benefit analysis of additional actions toovercome the problem in a short, mid and long termperspectives.The WB recommendation is that the geographicalzones-hotspots be defined based on the combination ofthree factors: agro-ecologic zones (geographical patternsdefined based on combinations of soil characteristic,rainfall pattern, vegetations, agriculture potential andenvironmental conditions) 1 ; physical aspects (key naturalevents that may expose households and communitiesat risk of natural hazard) 2 ; social or socioeconomicaspects (key elements at disposal and affecting household’swellbeing that can be translated as combinedfactors influencing household levels of wealth, foodsecurity and nutrition, health and sanitation) 3 .In order to address the above concerns and instructionprovided by the World Bank and by taking intoaccount that adaptation to climatic changes is intrinsicallylinked to anthropogenic multi-factors, a frameworkwas laid out to capture the cause-effectrelationship among anthropogenic factors integratingagro-ecologic factors with physical and social aspects.As a result of this, it was possible to identify six zonesas well as physical hotspot with respect to climaticchanges. In addition, elements related to food securityand nutrition, poverty indicators as well issues relatedto health were integrated for a final definition ofgeographical hotspot.With this background, the study was conducted in threestages:• Definition of hotspots• Three workshops(two regional and one national)• Fieldwork at representative geographical zones• Analyses and presentation of overall trends1 IIAM had set up 10 agro-ecologic zones2 For Mozambican context, drought, cyclone, flood and sea level rise arethe most common hazards3 For the purpose of this study, food security, health and nutrition,HIV-AIDS and poverty should be the most relevant socioeconomicfactors.

60 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U EThe present report specifically addresses only issuesrelated to the fieldwork and it will deal with instruments,geographical coverage, procedures and activitiescarried out after fieldwork.fIelDwoRkFieldwork was initially planned to be conducted by localenumerators under the supervision of three junioruniversity graduates. The proposed approach was toassure high quality work because that way we could uselocal knowledge, control language constraints and carryout a more in depth survey, particularly with respect toidentification and characterization of livelihood groups.However, after various attempts to go to the field,which was postponed for three or four times due toadministrative reasons, the decision was that fieldworkwas going to be conducted by two junior students underthe supervision of the local technical coordinator.Thus, the approach was as follows:• In each district, the team recruited one local staff,preferably, one of the participants who attended theregional workshop to be part of the exercise.• In each administrative post, assistance was also providedby a technical staff who played the role offacilitator as well as translator.InstrumentsThe purpose of fieldwork was to cross check informationfrom the desk review as well as to gather insightson climatic changes at local level, particularly withrespect to livelihood groups and individual household.Furthermore, it was important to gather perceptionfrom different institutions and civil society about exposureto climatic changes, sensitivity, and adaptation,including extreme coping mechanisms, which wouldcomplement information already compiled from thethree workshops. Therefore, three instruments weredeveloped:1. Household questionnaire – was divided in 4sections:• Sociodemographic information• Agriculture performance and income source• Shocks and coping strategies• Climatic Changes and Vulnerability, whichinclude exposure, sensitivity, resilience andadaptation as well as institutional andcommunity´s performance and perception aboutclimatic changes.2. Institutional questionnaire – organized in four parts:• Institution profile, which included the identificationof area of intervention, group entitlementand partnerships.• Institutional perception about risk includingidentification/recall of past events, the effects ondifferent social groups, likely chance of repetitionand lesson learnt.• Sensitivity, which deals with perception aboutevents, preventive measures and action formitigation.• Inventory of adaptation (prevention, adaptationand mitigation done by the community andinstitutions) and constraints.3. Focus Groups – This instrument was designed tocapture information about various social groups andto understand common perception of groups aboutclimatic changes, variations on adaptation and senseabout future events. This included wealth rankingexercise and specific interview with the representativeof part of the identified social groups.Once designed, the instrument was pre-tested and,thereafter, it was adjusted for the final version. It isimportant to mention that the results of the variousworkshops were also keen for the final adjustment ofthe instruments because they were carried out beforefieldwork,spatial RepresentationDuring the preliminary exercise, six zones sensitive tothe climatic changes were identified. Later on, theywere combined with the definition of hotspots, whichwere crucial for the identification of zones for fieldwork.However, such definition had to be done by taking intoaccount administrative boundaries. All zones werecovered and regional variations (inland and coastal) aswell as population density were taken into accountAs per the map, a total of 7 out of 10 provinces werevisited of which 2 in the North, 2 in the Center and 3in the South.

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s61ProceduresThe following steps were followed:• At District Level• Upon arrival in each district, the team met withlocal administrator and his/her staff to explainthe purpose of the mission and to gather preliminarybackground about agenda on climaticchanges, past events, and evidences of risks andadaptation.• Additionally, the team met with various institutionsto fill out the questionnaire about climaticchanges. When it was possible,separated interviews were conducted with officialinstitutions and then with NGOs or, otherwisetogether. Such procedure was intendedto control for biases and for crosscheckinginformation about the same issue. Overall,this was a good opportunity to discuss variousinsights about institutional perception on climaticchanges in terms of: exposures;sensitivity and adaptation; individual actiontaken to mitigate the problem; perceptionabout outlooks.• One of the objectives of the above mentionedmeetings was to discuss spatial variationsabout types of climatic change exposures,population density, examples of adaptationand variations on exposure to risk. For a practicalpurpose, participants usually describedthe characteristics of various administrativeposts 4 .• After the discussions, the team, consensually,selected two administrative posts to be part ofthe study. In general, the selection of thesesposts was done after an in-depth discussionabout past events, exposure and sensitivity ofeach post and, thereafter, the selection wasmade.• At Administrative Post• Meetings were held with the population of differentsocial strata as well as with traditionalauthorities.• The discussion started with an introductionabout the purpose of the mission, its neutralityin addressing the issue and clarification thatthere were no political, economic gains or evencommitments neither from the organization norfor follow up of projects or programs.• Following that, discussion about socioeconomyof administrative post linked to climaticchanges took place. The combination of thesetwo elements guided the direction of the discussionthat was often supported by varioustechniques of participatory approach and communitydevelopment.• Once again, a clear picture about socioeconomicvariation and climatic change exposures at locallevel were addressed. This was a good opportunityto select the 2 “localidades 5 ” should bevisited.4 Geographical/administrative units below districts. In average eachdistrict has two to four administrative post and each administrativepost has about three “localidades”5 There is wider variation among localidades in terms of livelihood andsocial organization as well as exposure to the climatic changes. Somelocalidades were in the very coastal area and typically linked to fishery,while others were in the interior and mainly subject to subsistenceagriculture.

62 t h E s o c i A l d i m E n s i o n s o f A d A p t A t i o n t o c l i m A t E c h A n g E i n m o z A m B i q U E• Wealth Ranking• The meeting at administrative posts was completedwith the exercise of wealth ranking toidentify and characterize livelihood groups aswell as their proportion and distribution withinthe “localidade”. Typical items (i.e. maize, varioustypes of beans, stones and other local items)were used based on their value/role to characterizelivelihood groups. Such exercise was welltaken by the populations who embraced it andparticipated actively; therefore, the teambecame mere facilitators.• As a result, there was a general perception thatwe could find concentration of various livelihoodgroups at one “localidade” and, based onthat, the team selected two localidades for hhand focus group interviews.• At localidadeAs per table 1, a total of 24 Administrative Posts and 28localidades were visited. In each localidade, the followingtook place: 15 HH interviews with representativesof various identified livelihood groups; 3 focus interviews;1 institutional.In some cases, purposive interviews were conducted tocapture and enhance social variation, despite the factthat they were not identified during the wealth rankingexercise (i.e. gender issue, widows, orphan child as headof HH).Post-field Visit• Data Cleaning and Post-CodingA total of 137 HH questionnaires were completed aswell as 18 institutional and 45 focus interviews. Due tothe fact that there were close and open questions in thequestionnaires, there was a need for the revision andpost-coding. This exercise took longer than expectedbecause only two junior collaborators under my supervisionwere carrying out the task. They spent about sixdays on that task.• Data EntryA database had to be designed in order to enter thedata. For that, the supervisor worked closely withtechnicians to go over each part of the questionnaire tomake sure that everything was accordingly. CSPro wasthe selected software because it will allow data to beentered friendly and can be easily transferred either toSPSS or Excell.focus Interviews and Institutional questionnairesThe post-coding was difficult due to a wide variation ofresponses, even though, a template was designed forthat purpose. To counteract the problem, the option wasto write a synthesis, which, at this stage, is all inPortuguese, but if needed, it can be translated.lesson learntOverall, the results for all the components of this studywere very interesting. However, they could have beenmuch better. Several issues prevented us from gettinggreater results and they are highlighted below:• Contrary to what happened in other countries, thestudy in Mozambique did not have any institutionalintegration at all. Not only that, but Mozambiquelacked a multi-disciplinary core team.• Theoretically, the study had a national coordinator.Even though we recognize all her effort, shewas overloaded with other work and often awayfrom Mozambique and, thus, she could not followall the day-to-day activities. Furthermore, herrelationship with WB was not all clear at all,which often created some difficulties when sheneeded to drawn upon support from theorganization.• The team had also a technical coordinator who wasaway from the country at both initial phase andfieldwork. If he were here, he could have made adifference and many issues could have been dealtwith more profoundly. Thus, this led to some limitationson addressing technical issues.• There was limited consensus in terms of procedures,methodologies and overall technical aspects. Often,what was produced in the ground was not totallyintegrated or taken into account. For example,issues on socioeconomic aspects, including povertyand food security and nutrition were not well integratedwith climatic changes, even though, somevery good materials were produced.

d E V E l o p m E n t A n d c l i m A t E c h A n g E d i s c U s s i o n pA p E R s63• Resource allocation and very limited support fromWB office in Mozambique made it difficult toimplement the activities.• Often, the team was asked to prepare agenda withchronograms, budget and procedures, which wereoften changed for more than once. Such situationinterfered with the consultant’s scheduled andmade planning difficult. This was particularlyaggravated by great delays (i.e. fieldwork) whichmade it impossible to carry out the work smoothlyand; in some cases, it generated conflicts withrespect to the record of number of days on consultancywork.Overall, this activity was quite an experience and usefulto exchange ideas and to shape up methodologies tobring about outstanding results on economic costs ofadaptation to climatic changes.

The World Bank Group1818 H Street, NWWashington, D.C. 20433 USATel: 202-473-1000Fax: 202-477-6391Internet: www.worldbank.org/climatechange