When Disasters and Conflicts Collide

7/28/2019 When Disasters and Conflicts Collide

1/68

When disasters

and confictscollideImproving links between

disaster resilience andconfict prevention

Katie Harris, David Keen

and Tom Mitchell

February 2013


2/68

Overseas Development Institute

203 Blackfriars Road| London SE1 8NJ| UK

Tel: +44 (0)20 7922 0300Fax: +44 (0)20 7922 0399

www.odi.org.uk

The views presented in this paper are those of the author(s) and do not necessarily representthe views of ODI or our partners.

This material has beenfunded by UK aid fromthe UK Government,however the viewsexpressed do notnecessarily reflectthe UK Governmentsofficial policies.


3/68

When disasters

and confictscollideImproving links between

disaster resilience andconfict prevention

Katie Harris, David Keen

and Tom Mitchell

February 2013


4/68


5/68

Contents

Executive summary vi The scale of the challenge vii

The impact of natural disasters on con ict viiThe impact of con ict and fragility on natural disasters viiiLessons from current efforts to strengthen resilience viiiTowards a conceptual framework ixRecommendations x

1 Introduction 12 The challenge 33 The impact of natural disasters on con ict 103.1 Grievances 12

Natural resource and environmental scarcity 12Resource allocation pre- and post-disaster 13

3.2 Opportunities 14Economic opportunities 14Political opportunities 14

3.3 Feasibility 15

4 The impact of con ict on natural disasters 164.1 Impact of con ict on the causes of disasters 184.2 Impact of con ict on disaster response 234.3 Impact of disasters on individual coping strategies 23

5 Integrated approaches to managing con ict and natural disaster risk 255.1 Managing natural disasters and con ict in silos 275.2 Holistic approaches to resilience: integrated risk management 28

Building disaster resilience in fragile and con ict-affcted states 28Resilience to the effects of con ict 31The role of the state 32

5.3 Forging stronger links between disaster and con ict specialists 32

6 Conceptualising the linkages 366.1 Unintended consequences 376.2 From collision to collaboration 37

Entry point one: con ict prevention and statebuilding 39Entry point two: natural disasters and risk management 40

7 Recommendations 42 International policy 43

Programming and nance 43Research and evidence 44

References 45Annex 1: Quality of the evidence 48Annex 2: Background explanation: grievances, opportunities and feasibility of conflict 51Annex 3: Climate change and conditions of conflict 52

iii


6/68

Figures, Boxes and Tables

Figures1: Cases supporting the relationship between natural disasters and con ict ix2: Fragile and con ict-affected states, an aggregated list 73: Global Distribution of Drought Risk - Mortality 74: Climate Change Vulnerability Index 2013 85: Global distribution of multiple hazards mortality risk 86: Continuum of intent: disaster risk reduction and con ict prevention 28

7: Characterisation of the relationship between actions to address natural disasters and con ict 388: Schematic to show how conceptual frameworks for natural disasters and con ict could evolve 389: Building peace states and societies framework 3910: Adapted version of DFIDs Building peaceful states and societies framework illustrating the natural disaster 40

components of the peacebuilding and statebuilding agenda11: Operationalising con ict sensitivity 4012: Situating con ict within an adapted Pressure and Release Model 41

Boxes1: Selected recent natural disasters in fragile and con ict-affected states 52: Future trends shaping the con ict-natural disaster interface 5

3: Building resilience in Chad 184: Policy architecture for dealing with disaster risk 275: Building resilience in the Sahel region 296: The Nepal Risk Reduction Consortium practical funding for resilience 297: Tearfunds DRR work in Kandahar, Afghanistan 308: Disaster-con ict linkages and community based DRR in Karamoja 319: Reducing disaster risk in a fragile state Nepal 3310: Misconceptions among actors working to reduce risk to con ict and natural disasters 3411: Assessing protection and livelihoods 34

Tables1: Ranked list of countries demonstrating high levels of fragility, disaster risk, poverty and climate 9

change vulnerability2: The practical implications of undertaking disaster risk management in fragile and con ict-affected states 20

iv when disasters and conflict collide : improving links between disaster resilience and conflict prevention


7/68

Acronyms

DFID Department for International DevelopmentDRR Disaster risk reductionGFDRR Global Facility for Disaster Reduction and RecoveryHFA Hyogo Framework for ActionIFRC International Federation of Red Cross and Red Crescent SocietiesLTTE Liberation Tigers of Tamil EelamOCHA UN Office for the Coordination of Humanitarian Affairs

UNDP United Nations Development ProgrammeUNISDR UN International Strategy for Disaster Reduction

v


8/68

Executive summary

teun voeten / panos


9/68

In 2011, drought, and food and political insecurityin East Africa contributed to a full-scalehumanitarian crisis. A combination of naturalhazards, conflict and fragility provided a recipefor human suffering. From 2005-2009, more than50% of people affected by natural disasters livedin fragile and conflict-affected states (Kellett andSparks, 2012: 31).1

This convergence poses particular challengesfor governments and agencies working to securedevelopment progress and puts great strain onthe international humanitarian system.2 Withpoverty expected to be highly concentrated infragile and conflict-affected states by 2025 (Kharasand Rogerson, 2012) and with many fragile and

conflict-affected states still some way off meetingany of the Millennium Development Goals, theimpetus for tackling conflict and disaster riskcoherently in such settings is becoming a priorityfor both the development and humanitariancommunities. Without sufficient attention, hard-wondevelopment and security gains will be undone.

This study assesses the evidence base for hownatural disasters3 affect conflict, how conflictaffects natural disasters, and how people livingin complex environments are affected by multiplerisks. We also consider what can be learned fromcurrent practices to improve conflict prevention,statebuilding and disaster risk management inways that help build resilience. The evidencebase challenging: it is fragmented and contested,with a number of studies highlighting directlyopposing lines of arguments. This suggests thatthe complexity of conflict and disaster dynamicscan only be understood when grounded inspecific contexts.

This report focuses on the links betweenconditions of vulnerability and risks associatedwith the nexus of natural disasters, conflict andfragility. However, it is recognised that any givencontext will be mired in an even more complexarray of intersecting risks. Thus while there isa tendency to consider how one risk impactsanother, our overarching focus is on advancingan understanding of how multiple vulnerabilitiesstack up for populations. How disasters andconflict collide is an entry point to understandinghow vulnerability is dynamic and shaped byinterconnected shocks and stresses, and how itmust be addressed as such.

The scale o the challenge A number of high pro le disasters in fragile andcon ict-affected states have increased attention onthe concurrence of disasters and con ict, and there isan expectation that disasters and con ict will coincidemore in the future. Climate change, continuedurbanisation, food price uctuations, nancial shocksand other stresses may all shape and complicate future trends in the disastercon ict interface.

Based on a rudimentary analysis, there appears tobe a close association between the risk of mortalityfrom drought, state fragility and climate changevulnerability. However, the intersection betweenmortality risk from other natural hazards (such as

cyclones and earthquakes) and state fragility appearsto be much less pronounced, though still signi cant.

Somalia, Afghanistan and Niger are ranked high in acomposite list of countries considered to be fragile and/or con ict affected, with high disaster risk, high levelsof poverty and high vulnerability to climate change.

The impact o natural disasters on confict Though the picture is far from clear, the balance of evidence suggests that natural disasters exacerbatepre-existing con icts. There are only a limited number of cases where natural disasters have supportedpeacebuilding and led to the resolution of con icts,such as Aceh.4 In every complex situation, numerousinteractions exist, where natural disasters reducesome con ict drivers while exacerbating others.

1. These figures refer specifically to definitions and sources fromDevelopment Initiatives work on confli ct and the EM-DAT Centre

for Research on the Epidemiology of Disasters work on disaster-affected persons.2. Between 2007 and 2010 the UN consolidated appeals more than

doubled, reaching $11.3 billion. In the context of growing protractedcrises, global food crisis, economic crisis and increasing humanitarianneeds, there was a reduction in the volume of needs met (Poole andWalmsley, 2012: 59). For example, a 58.3% increase in needs between2007 and 2009 in Sudan, 88.7% increase in Occupied PalestinianTerritories in 2009 due to the Gaza crisis, and 121.9% increase inSomalia in 2009 owing to drought, ooding and increase insecurity(Poole and Walmsley, 2012: 59).

3. It is widely recognised that disasters are not natural but a product of aset of interactions between natural hazards, conditions of vulnerability,and socio-political-economic conditions. Although uncomfortable for theauthors, the term is used here to differentiate natural-hazard related

disasters from the con ict and fragility components of a disaster. For more on the paradigm shift towards recognising disasters not as atechnical issue but as inherently unnatural, with vulnerability at thecentre of the conceptual frame, see Wisner et al., 2004.

4. A celebrated though much debated case.

vii


10/68

Grievancescan be deepened by natural disastersthat increase resource scarcity or cause more acuteimbalances between areas of scarcity and abundance.Grievances can also increase with the unequaldistribution of ex-post humanitarian aid or ex-antepreventative/protective measures by governmentsor other agencies. Weak government responses tonatural disasters can also contribute to con ict.

The disruption caused by natural disasters canpresenteconomic opportunitiesfor criminalactivity, while their impact on livelihoods can leadindividuals to join armed groups. In some cases,though, good access to reconstruction aid canincrease the opportunity cost of conflict.

This study also finds that political opportunitiesfor engaging in conflict can arise when disasters createa smokescreen for advancing political or militaryobjectives (such as increasing military spending,deploying troops to sensitive areas, or manipulatingaid to some groups over others).

The feasibility of conflict can also be changedby natural disasters, either by strengthening or weakening one side in a conflict directly or throughthe appropriation of aid.

The impact o confict and ragility onnatural disastersThere is strong evidence that conflict and fragilityincrease the impact of natural disasters, notably byincreasing vulnerability to natural hazards. Conflictincreases disaster risk by displacing people intoareas more exposed to hazards and through theimpacts it has on physical and psychological health,basic service provision and the security of liveli-hoods. Conflict can drive individuals to sell assets,

which increases disaster risk. In a limited number of cases, individuals and groups can gain from con-flicts (through the so called war economy) in waysthat increase their resilience to disasters.

This study finds that conflict can undermine thecapacity of governmental and non-governmentalactors to plan for and protect people againsthazards for example, by inhibiting the ability toprovide basic early warning systems and to deviseand enforce building codes.

Governments can also exacerbate post-disaster suf-fering by inhibiting aid on security grounds or appro-priating humanitarian aid to support con ict objectives.

Some countries refuse, delay or complicate interna-tional help, fearing it will undermine sovereignty.

It is valuable to draw a distinction between fragileand conflict-affected states that arewilling but unable, and those that areunwilling and unabletoreduce the vulnerability of populations to disaster risks and impacts. Disaster risk management tendsto assume a positive state-society social contract5 exists where the state adopts the management of risk as a public good. In some states this may bethe case, but in others it is not. Figure 1 highlightsthe relationship between conflicts and naturaldisasters. The weight of evidence suggests astronger leaning to the left quadrants of the circle.

Lessons rom current e orts tostrengthen resilienceIn both policy and practice, conflict prevention anddisaster risk management are treated as discreteissues, with limited crossover of expertise or jointworking. Misconceptions, different languages, andlow levels of coordinated analysis and programminginhibit the potential for stronger collaboration.

The resilience agenda is helping to improve links

between humanitarian and development action andfostering stronger integration of risk management.While there are few well-documented examples of interventions that actively seek to integrate naturaldisaster risk reduction and conflict prevention,there does appear to be a growing realisation thatmanaging disaster risk in fragile and conflict-affectedstates cannot be a matter of business-as-usual.Hard won peace dividends may be underminedunless natural disaster risk is taken seriously.

The 2011 World Development Report warns not to

expect too much too soon of national institutions:action to transform governance needs long-terminvestment and sustained support. Yet less than 4%of humanitarian aid and less than 1% of developmentassistance is spent on ex-ante disaster prevention,preparedness and risk reduction (Poole and Walmsley,2012: 2). This severely constrains the potential for integrated risk management in fragile and con ict-affected states. While there are good reasons for notinvesting in ex-ante risk reduction in fragile situations

5. A dynamic agreement between the state and society on their mutual roles and responsibilities (Chandran and Jones, 2008: 17,in Harvey, 2009).

viii when disasters and conflict collide : improving links between disaster resilience and conflict prevention


11/68

(e.g. corruption, lack of capacity and political will), notdoing so makes little sense in the long term.

Towards a conceptual rameworkInterventions aimed at reducing natural disaster risk can have positive or negative effects on thedynamics of con ict; conversely, interventions aimed atpreventing con ict can have positive or negative effectson the likelihood and impact of natural disasters. Theideal scenario is to have interventions that reduce the

likelihood of natural disastersand con ict. Disaster riskmanagement should be integrated more systematicallyinto peacebuilding and statebuilding frameworks.The reverse should also happen. Integrating con ict

and fragility into natural disaster frameworks canhelp elucidate the links between natural disastersand con ict. Greater cross-integration of frameworkswill help encourage a transition from collision tocollaboration between the two communities.

Over time it will be necessary to devise integratedapproaches to natural disaster and con ict riskthrough a conceptual framework. The joint frameworkshould encourage accountability, learning, evidence-gathering, cross-organisational exchanges andshould draw on con ict sensitivity and politicaleconomy analysis. It should aspire to understandthe factors that produce vulnerability to disastersand con ict, as well as trade-offs and points of convergence where there is the most to gain.

Figure 1: Cases supporting the relationship between naturaldisasters and confict

Prolonged drought in Balkh province, Afghanistan in 2006-2007 resulted in youth being forced to join armedgroups (Heijman et al., 2009). Recurrent drought and food insecurity in post-war Burundi through the mid-2000s contributed to increased

levels of migration and tensions with host communities (Heijman et al., 2009). The 2004 tsunami in Sri Lanka has seen military expansion justi ed through its proposed role in reconstruction and

growing rebel capacity through increased nancial independence (Keen, 2009; Mampilly, 2009). Flooding and overflo w from the Koshi River, Terai region of Nepal in 2008, resulted in large-s cale

resettlement resulting in increased tension with host communities. These tensions were exacerbated bypolitical groups who used flood victims dissatisfaction over lack of clean water and shelter to feed anti-government sentiments (Vivikananda, 2001).

Migrants eeing from 1997 El Nino oods (together with other migrants) into the southern Somali region of Lower Juba has attracted clan raiding and land-grabs from relatively well-armed and powerful clans and sub-clans as wellas ghting between these more dominant groups (Keen, 1994; Narbeth and McLean, 2003).

The 1970 cyclone in former East Pakistan, and inadequate relief response from West Pakistan,gave major impetus to civil war and eventually to the establishment of Bangladesh (Olson andGawronski, 2003).

The 1972 earthquake in Managua, Nicaragua, led to massive government corruption inrelief and reconstruction, allowing Sandinista rebels to capitalise politically and open amilitary campaign in 1975 (Olson and Gawronski, 2003).

Slow onset disasters, changing natural resource availability and environmentalscarcity have contributed to an ongoing complex crisis in Darfur, Sudan (Flintand de Waal, 2005).

2004 tsunami in Ache,Indonesia is regarded ascontributing to the peaceprocess (Waizenegger, 2010).

The drought in Mozambiquein the early 1990s weakenedthe already struggling Renamorebels and helped end the war (Keen and Wilson, 1994).

During war between Pakistan andIndia in 1965, a cyclone struck,a cease re was negotiated toallow a response in Kashmir, andboth Pakistani and Indian troops

pulled out; but the con ictresumed a few

months later

(Nelson,2010).

For a small minority of individuals, gains from awar economy can be usedto reduce the risk of naturaldisasters and their negativeimpacts.

In Chad, state fragility and corruption have undermined equitable gover-nance and the ability to facilitate international aid and development in sup-port of livelihood security, disaster risk reduction and risk management. For instance, during the 2010 food crisis the National Office for Food Securitylacked the logistical resources to transport food from major towns to the mostseverely affected rural areas (Grubbels, 2011).

In Sudan, Darfurs the ongoing crisis have inhibited mechanisms for natural resourcemanagement, exacerbating slow onset disasters and environmental scarcity (Flint and

de Waal, 2005). Strategies designed to deal with drought and livelihoods more broadly have been hampered in

Kenyas Turkana and Kitui districts, by (often violent) livestock raiding (Eriksen and Lind, 2009). In 2008, Cyclone Nargis in Myanmar saw substantial delays in disaster response as the government limited

humanitarian access. Of the agencies that were permitted entry many were subject to restricted movement,limiting their ability to provide adequate relief (Asia-Pacific Centre for the Responsibility to Protect, 2008).

In flood-prone La Mojana, Colombia, the lack of government control, the presence of an active rebel group andcommunities weak political voice has seen a lack of investment in risk redu ction measures (Williams, 2011).

Moreover, whether the affected state labels a crisis as such is not merely a technical matter; the labelling of a situation has strong political dimensions. Examples include failing to declare a disaster to avoid appearingweak or undermining national pride (Cuba 1998 drought) and refusing international help, even when nationalcapabilities have been outstripped (Mozambique 2007 flooding) (Harvey, 2009).

Naturaldisasters canincrease the

risk of con ict

Naturaldisasters candecrease therisk of con ict

Conditions ofcon ict canincrease the

risk of naturaldisasters

Conditions ofcon ict candecrease therisk of naturaldisasters

ix


12/68

Recommendations

International policy

In order to raise the profile of the conflictdisaster nexus in fragile and conflict-affectedstates, UN member states andinternationalagencies should:

Ensure that managing risk in fragile andconflict-affected states is a key feature of thepost-2015 agreement on disaster risk reduction(Hyogo Framework 2) and that there are clear institutional mandates set to tackle this. TheWorld Bank2014 World Development Reporton risk, uncertainty and crisis should link backto the 2011 World Development Report onConflict, Security and Development and set anew agenda for managing risks in fragile andconflict-affected states.

Resilience, vulnerability, disaster and conflictshould be featured themes of post-2015development goals .

The Political Champions for Disaster Resilience Group should promote inter-agencyco-ordination to build resilience in fragile andconflict-affected states, developing regional

and national approaches to ex-ante riskmanagement in such settings.

Programming and inance

The Global Facility for Disaster Reductionand Recovery (GFDRR)should scale-upprogramming in fragile and con ict-affected statesand forge closer links with the con ict preventionwork of the World Bank, such as the GlobalCentre on Con ict, Justice and Development.

Bilateral donors and UN agencies should: Constitute joint risk taskforces in key fragile

and conflict-affected states to integrateconflict, natural disaster and climate changepractitioners, plans and programmes.

Explore new partnerships and new waysof working and build the evidence baseabout how to better invest in ex-ante riskmanagement measures in fragile andconflict-affected states.

Donors must be preparedto risk greater levels of up-stream investment in fragile andconflict-affected states .

Donors and other financing bodies should workto ensure that short-term funding restrictionsdo not inhibit opportunities to build resilience.Where possible,multi-year funding should bethe norm and the UN should look to expand theuse of multi-year consolidated appeals.

Civil society organisations and donors alikeshould invest in thecapacities of programmestaff in fragile and conflict-affected states tobetter link approaches to conflict, disastersand climate change. This may require trainingand new ways of formulating strategies anddesigning programmes. Donors, NGOs andother implementing agencies should developintegrated monitoring and evaluationframeworks for assessing needs, results, valuefor money and outcomes in fragile and conflict-affected states.

Research and evidence

We do not currently know how to measurethe scale and nature of risk facing fragile andconflict-affected countries, nor do we know whichinterventions are likely to be most effective inmanaging risk and building resilience in theseenvironments. Priorities for investment include to:

Develop amultidimensional risk index whichintegrates existing data on conflict and fragility,natural hazards, vulnerability, poverty andclimate change. Ideally this should be detailedenough to consider sub-national areas andshould include a process for weighting riskfactors depending on the focus of differentagencies. Monitoring changes to this index over time will help to highlight the co-dependencybetween different aspects of risk andvulnerability and allow progress to be trackedand analysed.

Develop and testconceptual frameworksand analytical tools . This should includemodifying existing analytical tools (suchas conflict sensitivity frameworks andstatebuilding and peacebuilding frameworks)to reflect disaster risk and vice versa. Thisprocess could provide the model for moreintegrated risk modelling.

Build the evidence base about what works in

increasing resilience to multiple shocks andstresses in fragile and conflict-affected states.

x when disasters and conflict collide : improving links between disaster resilience and conflict prevention


13/68

Introduction

1

espen rasmussen / panos


14/68

In 2011, drought, food and political insecurityin East Africa contributed to a full-scalehumanitarian crisis. Like many other recenthigh profile disasters, a combination of naturalhazards, insecurity, conflict and fragility provideda recipe for human suffering. From 2005-2009,more than 50% of people impacted by naturaldisasters lived in fragile and conflict-affectedstates (Kellett and Sparks, 2012: 31).6 Thisconvergence poses particular challenges for agencies working to secure development progressin such environments and puts great strain onthe international humanitarian system.7 While it iswidely acknowledged that more needs to be doneto tackle the risk posed by disasters and conflict(e.g. World Bank, 2010), in practice as littleas 1% of official development assistance wasinvested in reducing disaster risks between 2000and 2010 (Kellett and Sparks, 2012: 10).

Fortunately, the link between natural disasters8 and conflict is receiving growing attention fromresearchers and development agencies (e.g.UNDP, 2011; Mitchell and Smith, 2011; Walch,2010; DFID 2011b).9 These studies acknowledgethat the convergence of disasters and conflictsignificantly compounds their impacts, impairs

recovery, and increases the risk of future crises.They call for action that addresses the impacts of disasters in a way that supports a countrys socialand institutional fabric and its transition out of violence (Kostner and Meutia, 2011).

The problem is that the evidence for how conflictand disasters interrelate is fractured, highlycontext-specific and poorly understood (see Annex 1 for an assessment of the evidencebase). Do conflicts necessarily predisposecountries to vulnerability to natural disasters?

And if so , how? Do natural disasters exacerba teor ameliorate conflicts? What evidence existsabout efforts to tackle the causes of naturaldisasters and conflicts jointly? What does theevidence mean for the way stakeholders seek toachieve progress towards peace, stability anddevelopment in such contexts? This study seeksto answer these questions through a detailedreview of the available evidence. In doing so,it summarises lessons, highlights complexitywhere it emerges, and details recommendations

for strengthening resilience.

The study is structured as follows: Section 2reviews the scale of the challenge by examiningthe extent to which conflict and natural disastersoverlap and which countries are of most concernnow and in the future. Section 3 looks at theevidence for how natural disasters affect conflictand fragility, before Section 4 examines thecomplementary question of how conflict andfragility affect resilience to natural disasters.Section 5 explores how the disasterconflictinterface might be understood conceptuallyand how the relationship between disaster riskmanagement and conflict prevention mightbe strengthened. Section 6 reviews currentpractices and includes a set of case studies,and section 7 outlines recommendations for strengthening international policy, programmingand finance and the current evidence base.

6. These figures refer specifically to definitions and sources fromDevelopment Initiatives work on conflict and the EM-DAT Centrefor Research on the Epidemiology of Disasters work on disaster-

affected persons.7. Between 2007 and 2010 the UN consolidated appeals more

than doubled, reaching $11.3 billion. In the context of growingprotracted crises, global food crisis, economic crisis and increasinghumanitarian needs, there was a reduction in the volume of needsmet (Poole and Walmsley, 2012: 59). For example, a 58.3% increasein needs between 2007 and 2009 in Sudan, 88.7% increase in theOccupied Palestinian Territories in 2009 due to the Gaza crisi s, and121.9% increase in Somalia in 2009 owing to drought, flooding andincrease insecurity (Poole and Walmsley, 2012: 59).

8. It is worth noting that the natural part of natural disasters is amisnomer. Experts link the cause of disasters to vulnerable peopleliving in locations exposed to natural hazards. Yet the natural labelremains in disasters discourse hence we have chosen to include thenatural element here so as to clarify the distinction between disasters

associated with natural hazards and disasters associated with con ict.9. Even the UKs Building Stability Overseas Strategy (DFID, FCO

& MoD, 2011) includes a reference to the relationship betweenconflicts and disasters stating that conflict-affected environmentsare more affected by natural disasters.

2 when disasters and conflict collide : improving links between disaster resilience and conflict prevention


15/68

The challengeThis section examines the scale o the challenge by

reviewing the evidence about the extent to whichnatural disasters, ragility and con lict overlap.


2


16/68

A number of high-profile disasters from the lastfive years are widely understood to have beenshaped by conflict, fragility and insecurity (Box 1).Between 1999 and 2004, 140 disasters associatedwith natural hazards were in contexts affected bycomplex political emergencies (Buchanan-Smithand Christoplos, 2004). And as noted above,the number of people affected by disasters infragile and conflict-affected states appears to bedisproportionately high.10

This phenomenon of intersecting disaster-conflictshas already received increased policy attention.The 2011 Humanitarian Emergency ResponseReview highlighted how multiple risks are drivingever-increasing demands on humanitariancapacity, arguing that current approaches tomanaging emergencies are unsustainable (DFID,2011a). The 2011 World Development Reportincluded natural disaster deaths as a key indicator of overall security concerns. Analyses like these

are driving the demand for a better understandingof future trends in the conflictnatural disaster interface (see Box 2).

The hard data currently available on the co-location of disasters and conflict is limited inquantity and somewhat contested. This is partlyto do with the quality of the available datasets,the way environmental stresses are analysed,and the challenges of reporting and accuratelyrecording situations where natural disasters andconflict intersect. For example, disentangling thehydro-meteorological, conflict, vulnerability andcapacity-related causes of famine or floodingis notoriously difficult and there is certainly noconsistency in reporting.

A few studies have reviewed the correlationbetween conflict and disasters. With a smallsample over a short time period, Drury andOlson (1997, 1998) find a significant relationshipbetween natural disasters and political unrest,identifying the key variables as insufficient andinequitable government responses and the

SummaryA number of high profile disasters in fragile and conflict-affected states haveincreased the attention being paid to how disasters and conflict collide, thoughsystematic analysis is limited and sometimes contested. The evidence that does existpoints to disasters and conflict coinciding more in the future. Climate change, urbanisation,food price fluctuations, financial shocks and other stresses may all shape and complicate future trends in the disasterconflict interface.

A rudimentary analysis of indicies suggests a concurrence between droughtmortality risk, state fragility and climate change vulnerability. However, theintersection between mortality risk from other natural hazards (such as earthquakes andcyclones) and state fragility appears to be much less pronounced, though still significant incertain locations.

Somalia, Afghanistan and Niger are ranked high in a composite list developedfor this report of countries considered to be fragile and/or conflict-affected, withhigh disaster risk, high levels of poverty and high levels of vulnerability to climatechange . Such rankings, it must be noted, can only provide indicative guidance owing to thedifficulty of combining data sets and the shortcomings those data sets contain.

10. See the figures cited earlier for 2005-2009 (Kellett and Sparks,2012: 31).



17/68

Box 1: Selected recent natural disasters in ragileand con lict-a ected states The relief effort following the 2004tsunami

that hit Aceh, Indonesia is widelydocumented as an example of disaster diplomacy, where the disaster createdspace for negotiation and assisted thepeace process.

Cyclone Nargis which hit Burma in 2008,illustrated the challenges of humanitarianresponse being delayed by restrictedaccess, which was complicated by theconcerns of international actors aboutcooperating with the Burmese leadership onaccount of their human rights violations.

The 2010Haiti earthquake raisedquestions about the lack of disaster preparedness in contexts mired by sexualviolence, poverty, weak institutions andpoor governance.

Widespreadflooding across Pakistan in2010-2011 raised international concernsabout the role of non-state actors,insurgents and the (political) securityimplications of disaster relief.

In 2011, predicted drought andfoodinsecurity in East Africa developed into afull-scale humanitarian crisis accentuatedby political insecurity.

Box 2: Future trendsshaping the con lict-naturaldisaster inter aceWhat will drive changes in the relationshipbetween conflict and natural disasters in thefuture? While there is a growing interest inthe way climate change will shape conflict(see Annex C) and have an impact on naturaldisasters, the combination of all three factorsis rarely discussed. Similarly, differinganalyses of where the poor will be located inthe coming decades (Sumner, 2012; Kharasand Rogerson, 2012) have not fully consideredthe co-location of conflict and naturaldisasters. We know that the conflictnatural

disasters interface will not remain static: wecan be sure that changing settlement patterns,migration, urbanisation and changes to socio-economic conditions will all affect underlyingexposure and vulnerability (IPCC, 2012: 5).Moreover, the World Development Report 2011(World Bank, 2011: 17) finds that multiplestresses raise the risks of violence and acombination of economic, political and securityactors both internal and external to the givencontext have a role to play. These stresses,and the future trends associated with them,often paint a bleak picture: climate change,financial shocks (particularly in relation tofood price fluctuations), youth unemployment,inequity and injustice, criminal networks, rapidurbanisation and the growth of megacities,informal settlements and inadequate landmanagement (World Bank, 2011; IPCC, 2012).

Migration is likely to be an important factor shaping future trends. In 2010, 42 millionpeople were displaced by natural hazards, upfrom 17 million in 2009 (Foresight, 2011: 6).One particular area of concern in this regard is

the existence of trapped populations unableto move from vulnerable locations the IPCC(2012) deems potentially unfit to live and workin in the future. Somalia is one such example:pastoralists are unable to follow traditional or alternative migratory routes to escape droughtdue to armed conflict, and humanitarianaccess is severely restricted.

5


18/68

severity of the disaster. Miguel, Stayanath, andSergentis statistical analysis (2004) concludesthat droughts increase civil war in Africa. Berebiand Ostwalds study of 167 countries (2011) findsthat natural disasters create vulnerabilities instate-society relationships that can be exploitedby rebel groups. Nel and Righarts (2008) find thatnatural disasters significantly increase the risk of violent civil conflict in the short and medium termin low and middle income countries that havehigh inequality and sluggish economic growth.11 In contrast to other studies, however, they foundthat the severity of the disaster was relativelyunimportant in terms of the impact on security.This has been contested by Slettebak (2012) onthe ground of technical inaccuracies. Slettebakconcludes that the analysis should show thatcountries experiencing disasters in the sameor the previous year are less likely to have anoutbreak of civil conflict; he therefore cautionsinternational agencies and other actors againstautomatically prioritising security concerns in thepost-disaster period.

There is another wave of research looking atsub-national patterns of conflict and disasters(see Buhaug and Lujala, 2005; Buhaug

and Rd, 2006; Raleigh and Urdal, 2007).While these present a more nuanced picturethan national level studies, they have beencriticised for assuming that the effect of localenvironmental conditions are limited to theimmediate area, whereas migration muddiesthis picture (Hendrix and Salehya, 2012).Hendrix and Salehya (2012: 35) find that veryhigh and very low rainfall years are associatedpositively with violent events such as cross-border and inter-communal violence.

2.1 Combining indicesTo demonstrate the way risks intersect, figures2-5 illustrate the global picture of differenttypes of risk. Figure 2 highlights the distributionof failed states, as indicated by Foreign Policys(2012) Failed States Index 2012. Figure 3 isthe World Banks (2005) map of global droughtmortality risk. Figure 4 is Maplecrofts (2012)Climate Change Vulnerability Index 2013.Figure 5 is UNISDRs (2009) map of globalmortality risk from earthquakes, cyclones,floods and landslides. These maps appear toshow a high level of concurrence of fragility,climate change vulnerability and droughtmortality risk. There is less concurrence,however, when considering mortality riskassociated with other natural hazards. Thisobservation is cautioned with a number of caveats, including the fact that mortality riskdoes not represent impacts on livelihoods or morbidity, that some of the indicators usedto create the indices are repeated, and thatthe indices themselves are highly contested.More work is needed to statistically analysethe geographic relationship between disas ter,conflict and climate change risks.

While acknowledging these caveats, indices canbe overlaid to create a list of countries that areconsidered fragile, vulnerable to climate change,have high levels of poverty, and are at high riskof natural disaste rs. Table 1 uses the FailedStates Index 2012 (Foreign Policy, 2012), theUNU-EHS World Risk Report 2011 (UNU-EHS,2011), the OPHI Multidimensional Poverty Index2011 (OPHI, 2011) and the CGD Climate Chan geVulnerability Index 2011 (CGD, 2011). Data for some countries is missing, which skews theoutcome to some degree (for example, there isnot enough data to include South Sudan).

11. Based on analysis of 187 political units with populations larger than 150,000.



19/68

Figure 3: Global Distribution o Drought Risk - Mortality

source : world bank , 2005

1st - 4th

Risk decilesDrought mortality

5th - 7th

8th - 10th

Figure 2: Fragile and confict-a ected states, an aggregated list

source : original map based on source material from oecd , 2010 in sumner , 2010

Indicates countries that appear in one or two (but not all three)of the lists outlined above.

Indicates countries that appear in the Brookings Index of State Weakness in theDeveloping World, Carleton University Country Indicators for Foreign Policy project, andthe World Banks Country Policy and Institutional Assessments List (in Sumner, 2010).

7


20/68

Figure 4: Climate Change Vulnerability Index 2013

source : maplecroft , 2012

Figure 5: Global distribution o multiple hazards mortality risk

source : unisdr , 2009

Sierra Leone

Guinea-Bissau

BangladeshPhilippines

Cambodia

ZimbabweMadagascar

DR CongoEthiopia

Haiti

Low risk Extreme risk

Rank Country Category

1 Haiti Extreme

2 Bangladesh Extreme3 Zimbabwe Extreme4 Sierra Leone Extreme5 Philippines Extreme6 Cambodia Extreme7 Madagascar Extreme8 DR Congo Extreme9 Ethiopia Extreme10 Guinea-Bissau Extreme

10 High

Tropical cyclones,floods, earthquakes,landslides

Multi-risk

9876543

21 Low0 Unknown exposure



21/68

The OCHA Global Focus Model12 is another useful illustration, systematically analysingnatural hazard and conflict risks. Yet as Lawry-White (2012: 15) highlights, while other modelsof risk exist, the tendency to focus on naturaldisaster risk and leave out conflict meansthe results often bear little comparison. For example, while the UNU World Risk Indexcriteria may not sound markedly different toOCHAs Global Focus Model, only one countryappears in the top 20 of both lists; similarly, of the World Bank GFDRRs 31 focus countriesonly 4 appear in the top 20 of the Global FocusModel. It is interesting to note that of the 20countries listed in Table 1, all except Timor-Leste appear in the top 50 countries of OCHAsGlobal Focus Model.

Table 1: Ranked list o countriesdemonstrating high levels o ragility,disaster risk, poverty and climatechange vulnerability 13

1 Somalia 11 Zimbabwe

2 Afghanistan 12 Ethiopia

3 Niger 13 Central AfricanRepublic

4 Guinea-Bissau 14 Bangladesh

5 Burundi 15 Liberia

6 Chad 16 Sierra Leone

7 Sudan 17 Timor-Leste

8 Congo, Dem. Rep. 18 Burkina Faso

9 Guinea 19 Burma/Myanmar

10 Haiti 20 Rwanda

12. For OCHA, data is analysed in four areas: hazards, vulnerability,capacity and demand for humanitarian coordination support.

13. The ranking has been produced by combining data from theFailed States Index 2012, the UNU-EHS World Risk Report 2011,the OHPI Multidimensional Poverty Index 2011 and the CGDClimate Change Vulnerability Index 2011. The Failed StatesIndex was given additional weighting, so represented 2/5 of theoutcome, to adjust for the fragile and conflict-affected statesfocus of this analysis and to balance the other three indices (of which the CGD index also includes extreme event vulnerability).The indices were chosen on the basis of data accessibilityand the fact that The World Risk Report includes drought as ameasure many others do not.

9


22/68

The impact onatural disasters

on confict This section examines the evidence or how natural disastersa ect confict and ragility. The evidence is organised around

three drivers o confict grievances, opportunities andeasibility commonly used in peacebuilding and statebuilding

rameworks (DFID, 2010). Further details o the analyticalramework employed here are ound in Annex 2.


3


23/68

SummaryOn balance, the evidence suggests that natural disasters exacerbate pre-existingconflicts by contributing to grievances, opportunities and feasibility . There area limited number of cases where natural disasters have supported peacebuilding andled to the resolution of conflicts. More often, natural disasters reduce some conflictdrivers while exacerbating others. Detailed empirical analysis is required in each caseto understand the net impact.

Grievances can be deepened by natural disasters causing greater resourcescarcity or more acute imbalances between areas of scarcity and abundance .However, areas with longstanding experience of droughts and floods have shown thatlocal institutions can mediate effectively and reduce tensions when periodic resourcescarcity occurs.

Grievances can also increase in intensity through the unequal distribution of ex-post humanitarian aid or ex-ante protective assets by governments, internationalagencies or non-state actors . Weak government responses can also contribute toconflict. In a few cases, a combination of factors have led to disaster diplomacy,where disasters have created opportunities for dialogue between opposing parties Aceh in 2004-2005 is a celebrated (and much debated) case.

The disruption caused by natural disasters can present economic opportunitiesfor criminal activity or lead individuals to join armed groups . In some cases,however, a lack of post-disaster economic opportunities can reduce violence, andaccess to reconstruction aid can increase the opportunity cost of conflict.

Disasters can create opportunities for advancing military objectives . Militaryspending might be increased or troops might be deployed to sensitive areas in thename of disaster management. International responses to disaster may also createpolitical opportunities for increased conflict when humanitarian aid is manipulated.

On the other hand there are cases where military activity is deemed politicallyinappropriate when natural disasters have adversely affected rival groups.

The feasibility of conflict can be affected by natural disasters, either becauseone side in a conflict is weakened or strengthened or because aid is beingappropriated . New balances of power may imply either a reduction or exacerbationof conflict, depending on the circumstances. Moreover, labelling complex politicalemergencies as natural disasters may help to draw attention away from underlyingconflict-related causes of humanitarian crisis and provide cover for human rights abuses.

11


24/68

3.1 GrievancesThere is a substantial body of evidencesuggesting that natural disasters ignite or exacerbate conflict by deepening grievances.More limited evidence highlights that naturaldisasters can also affect grievances in ways thathelp to ameliorate conflict. Most of the evidencerelates to resource scarcity and to formalresponses to shocks and stresses.

Natural resource and environmentalscarcitySome research suggests that environmental

changes associated with natural disasters canfuel armed conflict by increasing migration andundermining the economy (notably agriculture) byfuelling tensions within and between communities,and by weakening institutions (particularly stateinstitutions charged with promoting security,health and education) (Homer-Dixon & Blitt,1998; Kahl, 2006). It has also been argued thatenvironmental scarcity exacerbates the likelihoodof civil conflict through increased competitionover scarce natural resources (Homer-Dixon,1991, 1999; Kahl, 2006; Peluso and Watts,2001). Natural disasters associated with droughtare often singled out. Conditions of drought canincrease tensions over natural resources, leadingto confrontation between different land users(i.e. farmers and pastoralists). Such tensions areoften exploited or exacerbated by governmentsseeking to gain political advantages (Flint andde Waal, 2005). Groups or individuals who haveexperienced a major squeeze on their livelihoods(e.g. as a result of climatic change, discriminationor violence) are more likely to be recruited intoviolent processes, which can further jeopardisethe livelihoods of others (Keen, 1994; Younget al., 2009). In post-war Burundi, recurrentdrought and food insecurity, coupled with unevenfood distribution, have led to violence betweenmigrants and host communities over access toland (Heijman et al., 2009).

The notion that natural disasters generate conflictby making resources scarce has intuitive appeal.Yet a number of studies have portrayed a morecomplex reality (de Soysa, 2002a, 2002b; Urdal,2005). Hendrix and Saleyha (2012) suggestthat where grievances and resource competitiondo lead to conflict, this is likely to be relatively

disorganised violence that does not tend toinvolve the state. Another study found thatin poor countries a shortage of land, pastureand renewable (non-mineral) resources wasassociated with lower levels of armed conflict(de Soysa, 2002a, 2002b) one rationale beingthat, under these conditions, local institutionshave adapted to effectively deal with issues of scarcity. Research in the West African Sahelhas suggested that resource conflicts may bekept in check by traditional conflict resolutioninstitutions that have evolved over generations(Nyong et al., 2006). Lind has noted that local-level arrangements to share control and useof resources in some contexts can serve asthe foundation for a broader peace betweencompeting groups (2003: 317).

Other evidence suggests it is not simplyresource scarcity that drives conflict,but somecombination of scarcity and abundance. Fertileand resource-rich areas may attract conflict,raiding or quasi-legal attempts at expropriationfrom outsiders, including groups that occupy lesswell-endowed areas but perhaps have superior access to arms or political power (Keen, 1994).Such dynamics may be accentuated by a natural

disaster that creates an even greater imbalancebetween areas of scarcity and abundance(Keen, 1994). Southern Somalia, for example,has persistently been subject to famine despitehaving many fertile areas in part because theseareas attract interest from violent groups whooften inhabit less fertile areas. The southernSomali region of Lower Juba has not onlyattracted large numbers of migrants (includingthose fleeing the 1997 El Nino floods) due toits relative fertility, but has also attracted clanraiding, land grabs and skirmishes betweenrelatively well-armed and powerful clans and sub-clans (Keen, 1994; Narbeth and McLean, 2003).

In some cases governments have initiatedforced resettlement schemes on the basis of questionable claims that areas are ecologicallyunsustainable or too exposed to natural hazards.Keen (1994) highlighted this in the 1980s inEthiopia, where people were moved from Eritreaand Tigray, sowing the seeds of new grievances.Eriksen and Linds work on the Turkana region

of Kenya found that conflict is far from inevitablein marginal areas: carefully managed boreholeaccess can reduce conflict over water andgrazing resources (2009: 28). Where conflict did



25/68

occur in Turkana, there appeared to be complexpatronage by powerful government officialsrather than more straightforward inter-communalviolence over scarce resources (ibid.).

Resource allocation pre- andpost-disasterThe impact natural disasters have on conflict andstability can depend on the way a governmentresponds (Olson and Gawronski, 2003), or howit has invested in disaster preparedness or prevention (Cohen and Werker, 2008; Oliver-Smith,1996). Grievances have increased when naturaldisasters have been met with a weak or corrupt

government response (Olson and Gawronski,2003; Olson and Drury, 1997). For example, acyclone in East Pakistan in 1970 was met with apoor relief response from West Pakistan and half amillion people died, giving a major impetus to civilwar (Olson and Gawronski, 2003).

In Nicaragua, there was massive governmentcorruption in relief and reconstruction followingthe earthquake of 1972 (Olson and Gawronski,2003). This allowed the Sandinista rebelsto capitalise politically and begin a military

campaign in 1975 (ibid.). After the 2005 Pakistanearthquake, a slow government responseopened space for independent Islamic aidagencies to provide relief and criticise thegovernment for its failings (ibid.). Walsh foundthat areas receiving this relief saw a growth inanti-governmental Pakistani nationalism (2005,in Nelson, 2010: 162).

Natural disasters can often exacerbate existinginequalities in access to government services(Albala-Bertrand, 1993; Cuny, 1983) or further accentuate such inequalities in ways that helpto spur political change (Cu ny, 1983; Birkmannet al., 2010). Political commentators suchas Klein (2007) have also noted that naturaldisasters have been used as an opportunityfor governments to seize valuable land for reconstruction, removing the disaster affectedpopulation and appropriating their assets.

Evidence also suggests that government actionfollowing natural disasters can sometimes

create a climate of empathy that helps toresolve conflict (Quarantelli and Dynes, 1976;Kelman 2012; Brancati, 2007; Nelson, 2010;Slettebak, 2012). The most celebrated and

debated of such cases is the impact of the2004 South Asia tsunami on the conflict in Aceh,Indonesia. Waizenegger and Hyndman (2010:795) note that in the immediate aftermath of the tsunami Acehs secessionist Free AcehMovement committed to a unilateral ceasefireand the Indonesian military was instructed toexercise restraint while relief operations werebeing carried out. Waizenegger and Hyndman(2010: 794) also suggest that the relief andreconstruction processes opened the regionup to international audiences, ending thegovernment-imposed isolation and invisibility(ibid.: 794). The majority of their informantsmaintained that, without the tsunami, conflictwould have gone on for years. Other studies,though, play down the relative importance of the tsunami for bringing about peace, arguingdialogue between the conflict parties pre-datedthe tsunami (as did the drafting of a peaceagreement) and that a change in government inJakarta was more significant (Fan, forthcoming).

While there are a handful of disaster diplomacycases, reconciliation following a disaster can easilybe thrown off course. Akcinaroglu et al. (2011) findthat disasters can catalyse peacemaking between

rival states in the short term, but that this is easilyderailed by inter-communal violence and low levelcross-border conflict (see also Evin, 2004; Nelson,2010: 167-168).

International aid responses can also exacerbategrievances, where actors in the conflict perceivethat there is unequal distribution of resources(e.g. Buchanan-Smith and Christoplos, 2004;Waizenegger and Hyndman, 2010). Significantsteps have been taken however, to increase theuse of conflict sensitivity tools in humanitarian

responses in complex political settings(Zicherman et al., 2011). Most agencies adhereto humanitarian codes and guidelines and valuebroad inclusion in all stages of disaster riskreduction (DRR) (Sphere, 2004).

13


26/68

3.2 OpportunitiesNatural disasters play a role in creating andshaping economic and political opportunitieswhich, if seized, can lead to violence.

Economic opportunitiesNatural disasters can affect the distributionof economic resources, encourage theappropriation of resources by some groups,and present opportunities for criminal activitieswhich can lead to violence. For example, in2006-2007, drought in Afghanistans northernBalkh region forced downstream villagers tolook for new livelihood opportunities, whichfor some young men meant joining armedgroups (Heijman et al., 2009). The conditionsof drought were exacerbated by the unequaldistribution of water rights by elites following thefall of the Taliban rule in 2001 (Heijman et al.,2009: 34). Furthermore, diminished economicopportunities during a natural disaster canlead to a state relying on violence to sustainitself.14 In Sudan, a combination of drought,economic crisis and rising international debtfrom the 1980s onwards arguably encouragedthe government in Khartoum to support militiaraiding against the south that was designed,in large part, to secure access to its oil (Keen,1994). In other examples, increased resourceflows into a country for reconstruction canincentivise different forms of violence. For example, Waizenegger and Hyndman (2010:800) note that in Aceh an abundance of tsunamiaid is easy prey for many former combatantsin an environment with low legal enforcement.Moreover, local tensions have been raised by

the perception that more has been done for tsunami survivors than for ex-rebels and their families, while in reality many people have beenleft with little or nothing (ibid.).

Conversely, the economic opportunitiesassociated with violence may actually be reducedwhen a natural hazard means that there is lessto steal as when crops wither or the deathof livestock reduces the quantity of livestockthat can be raided.15 Equally, a substantialreconstruction effort after a disaster (whether natural or not) may increase the opportunity costof going to war (Collier, 2003).

Political opportunities A natural disaster, and the international response to thatdisaster, can create political opportunities which maylead to increased con ict. In Sri Lanka, a perceivedthreat to sovereignty from tsunami relief operationscreated signi cant political opportunities for nationalistparties and helped Mahinda Rajapaksa win thepresidency (Kleinfeld, 2007; Le Billon and Waizenegger,2007). International aid agencies faced considerabledif culty and outright Sri Lankan governmentintimidation when they tried to report on the extent of the humanitarian crisis, especially during the military'end game' against the Liberation Tigers of TamilEelam (LTTE) in 2009 (Keen, 2009). The rather chaoticinternational response to the tsunami16 contributed totense relations between aid agencies and Colombo,adding to the dif culties of responding to con ict-relatedemergencies thereafter.17 In some contexts, the statesrole in addressing disasters and reconstruction mightlegitimise sustaining a large military.18 In Sri Lanka, themilitary has justi ed its expansion after the war throughits proposed role in reconstruction.19

Disaster relief can be seized as a politicalopportunity to exercise power and can easily beused as a political tool by manipulating distributionand redirecting relief in ways that reward supportersand punish opponents (Williams, 2011: 17).Francken et al. (2008) argue this was the case inMadagascar where communities supportive of thepresident in the 2001 elections were 65% more likelyto receive relief following Cyclone Ga lo in 2004. A more extensively documented case is therepeated manipulation of food aid by the Sudanesegovernment (Keen, 1994; Flint and de Waal, 2005).

On the other hand, natural disasters mightreduce the political opportunities arising from conflict.

For example, a wave of public sympathy for thevictims of a disaster might make it politicallydisadvantageous to wage war on them (Kelman,2012). There seems to have been an element of this in Aceh after the 2004 tsunami.20

14. David Keen, independent analysis for this report.15. Ibid.16. See, for example, Tsunami Evaluation Coalition, 2007.17. David Keen, independent analysis for this report.18. Internationally, humanitarian missions arguably became part of the

just ification for maintaining high mili tary spending in the West after the Cold War (David Keen, independent analysis for this report.).

19. David Keen, independent analysis for this report.20. Ibid.



27/68

3.3 Feasibility As with grievances and opportunities, the feasibilityof violence is shaped by natural disasters indifferent ways.

Natural disasters may make insurrection easier by weakening or distracting the governmentapparatus or strengthening the legitimacy of rebel groups. As Nel and Righarts (2008: 162)have noted, natural disasters can weaken statecapacity and legitimacy, creating opportunities for the disgruntled to engage in violent resistance.In the aftermath of floods in Pakistan in 2010,militant groups reportedly used the disruption tocarry out attacks (Abbas, 2010; Waraich, 2010).

Mampillys study of Sri Lankas LTTE rebelssuggested: Following the [2004 tsunami] disaster,the LTTEs expansion of its governance capacitybrought it greater international respect andfinancial resources, convincing rebel leaders thatoperationally, de facto secession was viable. Inessence, what the tsunami did was to break thedelicate financial dependency of the insurgentson the Sri Lankan state (2009: 316).

Studies suggest that during the famine of 1984-1985the Ethiopian government was able to use relief aid to fund the war (Keller, 1992). Relief supplieswere used to pay soldiers and militia and to lurepeople into locations where they were recruited intothe military or subjected to forcible resettlement(Keller, 1992; Africa Watch, 1991). When nationaland international actors labelled the complex crisisa drought, this contributed to the impunity of the government forces who were at once activelyfuelling conflict and humanitarian disaster.21

In Sudan in the 1980s and 1990s, drought similarly

served as cover for government-sponsored violence,and again the label drought-induced famine wassometimes accepted internationally. In 1994,international assistance in response to the Rwandangenocide was focused primarily on the relief of hunger and a cholera epidemic among those whofled to eastern Zaire; again, there was a tendencyto treat a man-made disaster as if it were a naturaldisaster, and the human rights abuses underpinningthe crisis went largely unaddressed (Eriksson etal., 1996). In 2011-2012, the tendency to label thehumanitarian crisis in the Horn of Africa a droughtdistracted attention from some of the underlyingprocesses of violence that were creating famine,particularly in Somalia (Keen, 2012).

Enia (2008) suggests that evidence for hownatural disasters change the feasibility of conflictis ambiguous. Insurgents and governmentscan use natural disasters to strengthen their own position and this can lead to an escalationof violenceor to its diminution and cessation(Enia, 2008). For example, natural disasterscan speed up or slow down a war depending onthe overarching trajectory of conflict (ibid.). If rebels are losing and are further weakened bya natural disaster, then the war is likely to beshortened (Keen and Wilson, 1994). The droughtin Mozambique in the early 1990s, for instance,appears to have weakened the already strugglingRenamo rebels and helped to end the war (Keenand Wilson, 1994).

This section has underlined the need toconceptualise vulnerability as a dynamic processand understand how different vulnerabilitiesintersect. In contrast to conventional approachesto measuring vulnerability to natural hazards,which tend to focus on a single hazard over adiscrete period of time, more research is neededto understand how vulnerability changes before,during and after disaster events (Birkmann,2008). Vulnerability assessments which

account for such changes would help inform our understanding of the role of natural hazards indynamics of violence, conflict and fragility.

21. Ibid.22. Ibid.

15


28/68

The impact o confict onnatural disasters

This section considers the evidence about how confict

and ragility a ect peoples resilience to natural hazards,and explores the question: Does confict make people

more or less exposed and vulnerable to natural hazards?


4


29/68

SummaryThere is strong evidence to suggest that conflict and fragility increase theimpact of natural disasters. Conflict can increase disaster risk by displacing peopleinto areas more exposed to hazards, such as to informal settlements in exposedlocations. Conflict also increases vulnerability to natural disasters through the impactit has on physical and psychological health, basic service provision and securelivelihoods. Conflict can drive individuals to sell assets or to use valuable naturalcapital, which in turn increases disaster risk.

Conflict can undermine the capacity of government and non-governmentactors to provide adequate protection from natural hazards . Fragile and conflict-affected states, even if willing, often have difficulty implementing basic early warningsystems, devising and implementing building codes, and gaining access to highquality data on risks.

Governments can exacerbate post-disaster suffering by inhibiting aid onsecurity grounds or (mis)appropriating humanitarian aid to support politicalobjectives . Some countries also refuse international help, delay communicatingtheir need for relief, make ambiguous statements about the severity of their situation, or have difficult relationships with international actors.

It is useful to draw a distinction between fragile and conflict-affected statesthat are willing but unable , and those that are unwilling and unable to reducethe vulnerability of populations to disaster risks and impacts . Disaster riskmanagement tends to assume a positive state-society social contract exists wherethe state adopts the management of risk as a public good. But in some statesdisaster risk management is treated as a benefit available to political supporters.Intervention strategies (discussed in section 5) therefore need to be tailored tosuit the context.

17


30/68

4.1 Impact o confict on thecauses o disastersDisasters result from populations being subject to

combinations of vulnerabilities and exposure tonatural hazards. Vulnerability and exposure areshaped by household, local and national capacityto manage risk in the context of sets of structuralcauses like poverty, land access and socio-culturalstatus. Conflict and fragility affect vulnerability todisasters, and measures to manage disaster riskare more difficult to operationalise in fragile andconflict-affected states.

Violence, conflict and insecurity destroy thelivelihoods, infrastructure and basic servicesthat make a population resilient. The roleconflict plays in undermining disaster resilienceis well noted by Kelman (2012: 1): A war-weary population with reduced physical andpsychological health is more susceptible toa pandemic. A government focusing on war might neglect promulgation, monitoring, andenforcement of earthquake-related buildingcodes. Conflict frequently interferes with or cuts

essential supplies such as food, medicine, andbuilding materials, making it more difficult for people to keep their homes and communitiesprepared for floods or storms.

Longer-term disaster resilience is underminedwhen states fail to provide a set of corefunctions necessary to prevent and managenatural disasters and conflict. These functionsmay include effectively managing revenues andbuilding the capacity for accountable and fair social service delivery inherently linked withgood governance (DFID, 2010). Conventionaldisaster literature assumes that an effectivesocial contract exists between the state and itscitizens, or that those governments have a duty

of care towards their citizens to create the rightconditions for effective disaster resilience (HFA,2005; Wisner et al., 2004). Clearly, this may notbe the case in areas affected by armed conflictand fragility (see Box 3).

Box 3: Building resilience in ChadThe Sahel Working Group, an informal inter-agency network, recognises that fragile states pose particular challenges to humanitarian and development cooperation.

In Chad, for example, state fragility and corruption affect governance and the ability to facilitateinternational aid and development. Chadian politics is characterised by frequently shifting alliances: tomaintain a broad coalition of support, leaders of armed groups are included in government and politicalstructures. As alliances and support for these leaders shift, so too do the coalitions. This has resultedin a lack of continuity in government policy. As acknowledged by the governments own National Growthand Poverty Reduction Strategy, corruption remains a major problem. This has posed challenges for government spending, notably with regard to revenues from oil reserves.

Challenges to strengthening disaster risk management in Chad include:

A failure to institutionalise early warning systems (EWS): The Chadian government does not consider EWS to be a priority in its strategy for prevention and management of the food crises. This led todonors withdrawing funding for the development of effective EWS, which had been in developmentsince 1986, after the project evaluation highlighted a critical lack of national ownership.

Poorly developed systems for food security and crisis management: This includes poor structures for collecting information and consequently unreliable and incomplete data. Moreover, the governmentsnarrow focus on food availability over and above more nuanced conceptualisations of food securitymeans international food aid has often been considered the most appropriate response to addressfood crises. During the 2010 food crisis, for instance, the National Office for Food Security lacked thelogistical resources to transport food from major towns to the most severely affected rural areas.

source : gubbels , 2011



31/68

A particular facet of this problem is conflict-induced displacement. In Colombia, insecurityhas forced people to migrate to informal urbansettlements located on flood plains and unstableslopes (Williams, 2011: 24).

Fragility as well as violent conflict or war canmake disasters more likely. Institutions andgovernance structures may have limited capacityto mitigate, prevent, prepare and respond tonatural shocks and stresses (Kostner and Meutia,2011; World Bank, 2011). States may not bewilling or able to enact early warning systems,mitigate risks or launch effective response.

In Haiti successive unstable governments havefailed to provide either adequate regulation or economic opportunities to those living in chronicpoverty. In rural areas distress coping behaviour has caused unmanaged deforestation andsoil erosion, heightening the risk of landslidesand flash f looding (UNDP, 2011: 16). Theconcentration of poor and vulnerable communitiesin low-lying urban areas increases their exposureto the risk of flooding and storm damage (ibid.).This is coupled with an environment of criminalityin areas prone to storms and hurricaneswhich deters individuals from responding toevacuation warnings for fear of looting and theftin their absence (UNDP, 2011: 22-36). In bothexamples, the weak political voice of those ininformal settlements and the absence of politicalincentives to spend money on disaster riskreduction (DRR) have perpetuated high levels of exposure to natural hazards.

In some situations, the government may be partyto a conflict, inherently politicising its decision-making about how it manages disaster risk (de

Waal, 1997). In certain contexts, governmentsneglect particular regions or ethnic groups,making them more vulnerable to the effects of adisaster (ibid.).

Where states have shown the willingness andcapability to protect people from a varietyof risks, the relatively protected part of thepopulation is likely to be those deemed morepolitically influential. Individuals or communitieswho are well connected (politically or through theprivate sector), have greater economic assets,or are connected to urban centres are likelyto solicit more risk-reduction measures thanthose who are politically weak or marginalised,voiceless, poor and often rural. In Indonesia,higher DRR investment occurs in well-connected, high population density areas suchas Western Sumatra and Central Java (Williams,2011). In Colombia, seismic DRR in major urban areas such as Bogota contrasts starklywith underinvestment in flood-prone La Mojana(ibid.). Williams (2011) argues this illustratesthat residents in Bogota have more access to,and influence over, decision-makers than thepoorer and more remote fishing communities of La Mojana (ibid.). Conflict dynamics also play arole: the lack of government control in La Mojanadue to an active rebel group and communitiesweak political voice possibly explain a lackof investment.

Table 2 considers the challenges conflict andfragility present to building disaster resilienceto natural hazards through the practicalimplementation of risk management measures.

19


32/68

Table 2: The practical implications o undertaking disaster risk management in ragileand con lict-a ected states

Risk

managementcomponent

Role in building

resilience to naturalhazards

Challenges of fragile and

con ict-affected states

Example

Hazard pro le A crucial component of reducing disaster risk ismonitoring and analysingnatural hazard-relatedinformation to informpriorities for action andearly warning (Priority twoHFA, 2005; Twigg, 2004).This requires technical,scientific and institutionalcapabilities, investment indata collection and skilledtechnical staff.

Contexts of armed conflictpresent significant barriers tounderstanding the natural hazardprofile of an area. There maybe a lack of investment in datacollection, destruction of data,inaccessibility of local monitoringstations, lack of financial supportor suspicion over the use andcollection of information.In contexts where investmentand capability is present,data about the risks of naturalhazards may be embeddedwithin the security apparatus andthus not used for DRR.

For hazards that can be forecast, suchas hurricanes or volcanic eruptions,there are strong political incentivesfor ensuring early warning exists andinforms preparedness measures. Thefailure to ensure adequate preparednessagainst such predictable risks wouldindicate obvious negligence on the partof government, expose leaders to heavycriticism and thus create a very highpolitical cost (Williams, 2011: 23).

Disseminationof information

Dissemination of information plays a vitalrole in enabling disaster resilience, enablingindividuals and groups toprepare for a hazard andenact disaster plans.

In highly sensitive, fragile andconflict-affected contexts themode, style and content of messages associated withdisaster risks must be treatedwith caution to avoid becomingmanipulated or politicised.

Action taken on the basis of information disseminated isnot straightforward as disaster risks are weighed-up in relationto other risks associated withconflict and insecurity.

There is evidence of governmentswithholding information about the extentof the impacts of a disaster from theinternational community for variousreasons. In Burma the government failedto provide neither warning of CycloneNargis nor the humanitarian crisis that

followed (South et al., 2012: 10). Thesituation was further exacerbated bymany communities being denied accessto humanitarian aid by internationalagencies due to government restrictionson entry to the country.

Infrastructureand buildingcodes

Effective DRRrequires enforcementof infrastructure andbuilding codes, qualityconstruction, enforcedsustainable land usepractices, and penaltiesfor non-compliance(Wilkinson and Mitchell,2012). Governments andstate infrastructure are, or should be, regulators of these standards. Whereconflicting priorities are notmonitored or regulated,disaster risk can beinadvertently exacerbated.

In contexts with poor regulatory systems and lackof enforcement, particularlywith regard to the privatesector, construction industryand land management, theinfrastructural standards andincentive structures required toenforce DRR may be lacking.Wilkinson and Mitchell (2012: 4)note how political disincentivesexist where powerful interestgroups are likely to opposethe regulation of private sector activity, even where this maybe in the interest of reducingdisaster risk.

Drawing on case studies from Afghanistan, Democratic Republicof Congo, Nepal and South Sudan,international support to infrastructurein fragile and conflict-affected statesexperiences an intensified form of theproblems of underinvestment, lack of maintenance and weak institutionaland policy framework that apply acrossmost low income countries (Jones andHowarth, 2012: 2).

23. Basic development infrastructure includes access routes, water and sanitation, health facilities and a functioning banking system.Structures specifically designed to manage risk include hazard-related building codes and regulations, emergency responseequipment and facilities, insurance and reinsurance, and earlywarning systems for monitoring and dissemination of information.



33/68

Risk

managementcomponent

Role in building




Example

Environmentalprotection

UNEP (2009) hasdocumented a rangeof case studiesdemonstrating howwidespread unsustainableenvironmental practicesand weak enforcement of environmental protectioncome at the expense of environmental preservationand the exacerbation of disaster risk (UNEP, 2009).

The role of (politically) powerfulindividuals and their associationwith private companies means that,in some contexts, lobbying power,political donations and positionin patronage networks createconditions to enable damagingenvironmental practices to continueto the nancial advantage of asmall group of elites (Williams,2011: 19). Confronting powerfulprivate industry is perceived tobe beyond the ability (or in somecases will) of some governments.

Insecurity can inhibit effective environmentalprotection, which in turn increasesexposure. In Atrato Media, Colombiaunchecked exploitation of forests on slopesincreases the risk of oods and landslides(Buchanan-Smith and Christoplos, 2004).The authorities inability to access the areaowing to insecurity means that clearanceand maintenance of river blockages was notpossible, exacerbating the risk of ooding. As a result of armed con ict, the area alsolacked institutional capacity to enforceregulations and political incentives to protectcitizens from increased risk.

Diversion of resources

In societies affected byarmed con ict the diversionof resources to con ict-related activities can beat the expense of broader infrastructural development.The fact that no fragile andcon ict-affected state hasachieved a single Millennium

Development Goal istestament to this (WorldBank, 2011). Effective riskreduction for natural hazardstakes basic developmentinfrastructure as its startingpoint (with the addition of special provisions relatedto the hazard pro le of agiven area).23

Violence, con ict and insecuritydestroy livelihoods, infrastructureand basic service provision. Therole this plays in underminingdisaster resilience is well noted:A war-weary population withreduced physical and psychologicalhealth is more susceptible to apandemic. A government focusing

on war might neglect promulgation,monitoring, and enforcement of earthquake-related building codes.Con ict frequently interferes withor cuts essential supplies suchas food, medicine, and buildingmaterials, making it more dif cultfor people to keep their homes andcommunities prepared for oods or storms (Kelman, 2012: 1).

The prolonged complex political crisisin Darfur, Sudan has hindered thedevelopment of long-term investment inlivelihoods, services, infrastructure andcapacity. Despite a substantial amountof humanitarian (and some development)spending in the region, developmentprogress has been severely stunted bythe on-going conflict, and its associated

challenges.

Revenuecollection,taxes andcorruption

Revenue collection andtaxes are crucial to fundingservices which supportcommunal DRR. However,revenue collection andtaxes can also contributeto a lack of enforcedbuilding regulations asprohibiting development inhazard-prone areas meanslosing revenues fromdevelopment charges andproperty taxes two vitalsources of funding for localgovernments (Henstra andMcBean, 2005: 111).

Lucrative financial gains canbe made through corruptionin public construction, whichundermines the enforcementof building codes, quality andcontrol of materials and design,and building management (WorldBank, 2010). Corruption further undermines risk reduction effortsand has been witnessed acrossgovernmental scales, from localpoliticians to large-scale nationalprojects (Williams, 2011).

The relationship between corruptionand disaster mortality has been studiedusing a range of econometric indicators.Corruption and avoidance of safety andbuilding codes are believed to be major factors undermining DRR (Escaleras et al.,2007). Where there are weak incentivesfor government to provide public goods, itis often questioned whether the less visiblerisk reduction activities actually take placeat all, such as inspections, high-quality riskassessment and environmental protection(Williams, 2011: 18).

Corruption and perceptions of corruptionalso in uence the way international agenciesdeliver aid to a country. Concerns over government corruption led donors to deliver Malawis 2001-2002 disaster responsethrough international organisations (Darcyand Hofmann, 2003, in Harvey, 2009).

Table 2: The practical implications o undertaking disaster risk management in ragileand con lict-a ected states (continued)

21


34/68

Risk

managementcomponent

Role in building




Example

Contractingsystems

Fair and transparentcontracting systems arerequired to ensure safetyis not compromised onthe basis of cost saving,bribery or corruption.

Longer-term disaster resilienceis undermined by weaknessesin the contracting system for construction projects. In situationswhere transparency is lacking anda black market or bribery occurs,the enforcement of practices suchas DRR which require additionalresources is found wanting.

In Aceh, Indonesia, procurement ruleswere undermined in post-tsunamireconstruction. The pressure to mobiliserelief and reconstruction reportedlyto lead to a relaxing of the normalprocurement rules and existence of negligent practices including the inflationof contract values and bribery (seeWilliams, 2011: 18).

Coordination The coordination of sectors, programmes andpolicies across a rangeof issues is essential todeveloping a coherent andconsistent environmentfor building disaster resilience. The state has acrucial role in coordinatinginternal and externalassistance to ensurecomprehensive approachesto preparedness andresponse.

Where states do not have thecapacity and ability to act, thecoordination of different actorsmay be severely limited, resultingin the duplication of efforts or the exclusion of some recipientgroups over others.

Guatemala passed a law in 1996 obligingall private and state bodies to cooperatewith the national disaster managementsystem (Picard, 2007, in Harvey, 2009).In practice this faces many challenges.Coordination problems remain includingthe bypassing of national coordinationstructures.

Legalisationandregulation

Disaster risk policies andpractice require supportand guidance throughan effective legislativeand regulatory system.In practice, efforts toestablish or improvethe legal framework for disasters occurs ex-post.For example, Pakistanestablished the NationalDisaster Management Authority after the 2005earthquake.

The legal and regulatory systemof a country can help or hinder the ability of international actorsto work in a disaster context anddetermine (or at least influence)their engagement with statestructures.

In Pakistan the 2005 earthquakeprompted the establishment of theNational Disaster Management Authority.Similarly, after the 2004 tsunami, SriLanka enacted new disaster managementlegislation (Harvey 2009: 7).

Security andphysicalpresence of internationalactors

Enabling DRR requireseffective and sustainedpresence of agencieswith the ability to supportcapacity development inDRR, across a range of scales and geographicalareas.

Insecurity can put staff severelyat risk, limiting the abilityof agencies to undertakerisk reduction programmes.Inaccessible communities areoften the worst affected, whereconflict and insecurity inhibits allaspects of risk reduction.

The expulsion of 13 international NGOsfrom Sudan following the indictment of Sudanese President Omar al-Bashir bythe International Criminal Court is anexample of international agencies' abilityto work in fragile and conflict-affectedstates being challenged by the influenceof external events.

Table 2: The practical implications o undertaking disaster risk management in ragileand con lict-a ected states (continued)



35/68

4.2 Impact o confict ondisaster responsePowerful groups in fragile and conflict-

affected states (which can include government)sometimes make it difficult for national andinternational actors to work with populationsvulnerable to natural disasters. The most typicalexamples concern the impact of conflict ondisaster response, where humanitarian space isrestricted.

Following Cyclone Nargis in 2008 the Burmesegovernment allowed only limited humanitarianaccess, and only after substantial delays.Some agencies were denied access and thosethat were permitted entry were subject torestricted movement (Asia-Pacific Centre for theResponsibility to Protect, 2008: 2). The delaywas exacerbated following a defensive militarydeployment of army units to affected areas inresponse to warnings by foreign governments andactivists that aid would be imposed unilaterally,and attempts to mobilise the responsibility toprotect doctrine (South et al., 2012: 10). Thusthe already stretched resources of the state weredeployed on a security setting, rather than in relief

and recovery efforts (ibid.). The army deploymentdiscouraged relief to some vulnerable, cyclone-affected communities as international agenci

Documents

When Disasters and Conflicts Collide