Commissioned and published by the Humanitarian Practice Network at ODI

Number 44December 2003

Network Paper

About HPNThe Humanitarian Practice Network at theOverseas Development Institute is an independ-ent forum where field workers, managers andpolicymakers in the humanitarian sector shareinformation, analysis and experience. The views and opinions expressed in HPN’s publications do not necessarily state or reflectthose of the Humanitarian Policy Group or theOverseas Development Institute.

Overseas Development Institute111 Westminster Bridge RoadLondon SE1 7JDUnited Kingdom

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

HPN e-mail: hpn@odi.org.ukHPN website: www.odihpn.org

Britain’s leading independent think-tank on international developmentand humanitarian issues

HPNHumanitarian Practice Network

Managed by

Humanitarian Policy Group

SSiimmoonn NNaarrbbeetthh aanndd CCaalluumm MMccLLeeaann

Livelihoods and protection Displacement and vulnerable communities in Kismaayo, southern Somalia

In brief

• This paper is one of the first attempts to put livelihoods and protection into practice as a thoroughly integrated framework. Many of the problems facing internally-displaced people (IDPs) are related to the protective environment. In such circumstances, assistance must be designed in such a way that it will promote the protection of vulnerable groups without adding to their existing burden.

• This paper is based on the findings of field research on livelihoods, protection and IDPs conducted inKismaayo and the Lower Juba Valley inMay 2003, under the auspices of OCHA-Somalia. The research aimed toobtain a clear understanding of the situation of IDPs and other vulnerablecommunities in Kismaayo, and of theoperating environment; and to developan operational plan to better protect and assist the internally displaced and other vulnerable groups.

• This paper begins with a brief description of the livelihoods and protection framework that informed theresearch, and the methodology theresearch employed. It then provides a briefoverview of the findings and describes theconcept of a phased operational plan.

NP 44 crc cover 12/1/03 4:42 PM Page 3

Humanitarian Practice Network (HPN)Overseas Development Institute111 Westminster Bridge RoadLondon, SE1 7JDUnited Kingdom

Tel: +44 (0)20 7922 0331/74Fax: +44 (0)20 7922 0399 Email: hpn@odi.org.ukWebsite: www.odihpn.org

Layout and production: Publish-on-Demand LtdPrinted and bound in the UK

ISBN: 0 85003 699 2Price per copy: £4.00 (excluding postage and packing).© Overseas Development Institute, London, 2004.

Photocopies of all or part of this publication may be made providing that the source is acknowledged. Requestsfor the commercial reproduction of HPN material should be directed to the ODI as copyright holders. The NetworkCoordinator would appreciate receiving details of the use of any of this material in training, research or pro-gramme design, implementation or evaluation.

About the authors

Simon Narbeth is Senior Humanitarian Affairs Officer for UN-OCHA Somalia. He has worked with WFP Somalia,as a consultant for UN-OCHA Somalia, and as a research-extension officer for the University of California.

Calum McLean is Chief of UN-OCHA Somalia. He has worked extensively in the Horn of Africa over the last 18years, mainly with the international NGO community.

NP 44 crc cover 12/1/03 4:42 PM Page 4

i

Chapter 1 Introduction 1

Chapter 2 The analytical framework: livelihoods and protection 3

What is a livelihoods approach? 3

What is humanitarian protection? 4

Livelihoods and protection: bringing the two together 5

Chapter 3 Methodology 7

How to ‘do’ livelihoods and protection research 7

The fieldwork 7

Who is an IDP? 9

Participation and the research environment 10

Further issues and constraints 10

Chapter 4 Livelihoods and protection: study findings 11

Population movement and displacement in Somalia 11

Population movements into the Lower Juba and Kismaayo: a multi-clan topography 11

The Kismaayo complex 12

Conditions for IDPs 13

The protection environment 13

The implications for aid 14

Chapter 5 An operational plan of action 17

Chapter 6 Conclusions 21

Appendix 1 Camp committee/focus group questions 23

Appendix 2 Household questionnaire 25

List of boxes

Box 1: Defining livelihoods 3

Box 2: Defining protection 4

Box 3: The language of displacement 11

Box 4: Minority status and vulnerability 12

Contents

NP 44 crc 12/1/03 4:43 PM Page i

ii

NP 44 crc 12/1/03 4:43 PM Page ii

Population displacement is a feature of many conflicts.People may flee violence or human rights abuse, or theymay become displaced because the minimal requirementsfor life are unmet – for instance, during drought or flood, oreconomic upheaval. The displaced often face specialdifficulties not shared by other groups touched by conflictor disaster. The displaced are often disadvantaged in termsof their access to public facilities, compared to a host orindigenous community. Their location may influence theiraccess to humanitarian assistance, and their ability tosurvive and regain their economic security.1 Thehumanitarian challenge is to deliver assistance andprotection in what are often unfavourable environments,especially when the authorities are unable or unwilling toact.

Under hostile and predatory conditions, many of theproblems facing internally-displaced people (IDPs) andother vulnerable groups are related to the protectiveenvironment, and the potentially exploitative relationsbetween them and local authority structures and hostcommunities. A major dilemma in these situations is toensure that humanitarian assistance – resources such asfood aid, as well as activities such as healthcare orschooling – is accessible to the most vulnerable, and hasthe greatest beneficial impact. Assistance must bedesigned in such a way that it will promote the protectionof vulnerable groups without adding to their existingburden.

Although livelihoods and protection have been broughttogether elsewhere, at least in theory,2 this paperrepresents the first attempt to put livelihoods andprotection, as a thoroughly integrated framework, intopractice in Somalia, and to our knowledge anywhere.Conditions within the UN aid community are currentlyconducive for an approach of this nature. Respect forhuman rights and protection constituted one of the fourmain principles of the UN Somalia Country Team for 2003(the others were HIV/AIDS, education and the provision ofbasic services). Furthermore, the 2003 ConsolidatedAppeal stated the importance of ‘assisting in theintegration and protection of internally displacedpopulations, minorities, refugees and returnees and othervulnerable groups by enhancing protection efforts aimedat: building of local and national protection capacity;participation in governance; increased humanitarianaccess; awareness raising among populations and localauthorities; and the development and promotion ofdurable solutions’.3

In June 2002, the UN Office for the Coordination ofHumanitarian Affairs (OCHA)-Somalia convened aworkshop in Hargeisa, Somaliland, to raise the profile ofvulnerable populations in Somalia, and to develop a

strategic framework for UN agencies and the internationalcommunity to engage with them. The workshop wasattended by Somali nationals representing all of thecountry’s regions, as well as representatives from UNagencies, international NGOs and donors. The resultantframework promotes a cross-cutting livelihoods approach,rather than the more traditional sectoral approach(education, food security and rural development,governance, health and water and sanitation andinfrastructure) used by the UN and other actors within theSomalia Aid Coordination Body (SACB), the entityestablished to bring together donors, UN agencies andNGOs engaged in aid work in Somalia.

The June workshop once again highlighted the particulardifficulties faced by IDPs and minorities in Somalia. It wasnoted that, while vulnerable communities face similarchallenges across the country, IDPs in southern Somaliaare particularly vulnerable in terms of their livelihoodsecurity and access to basic services due to the generallypoor protective environment. While Somali refugees inother countries benefit from return and resettlementprogrammes, little has been done for people displacedwithin Somalia beyond meeting short-term needs.

An estimated 320,000–350,000 IDPs are distributedthroughout Somalia.4 The largest concentrations are inMogadishu and Kismaayo, with an estimated 150,000and 15,000 respectively. While Mogadishu is consideredtoo insecure to allow any meaningful work, Kismaayowas ‘reopened’ to international personnel of the UN bythe Security Coordinator for Somalia (UNSECOORD) inMarch 2002 after the Juba Valley Alliance (JVA), a clan-based factional alliance, had established a reasonabledegree of security in the town. It was therefore thoughtfeasible by the Somalia aid community to startnegotiating access to IDPs and other vulnerable groupsin the city, and to start planning interventions. In January2003, the UN Resident Humanitarian Coordinator andthe head of OCHA-Somalia met JVA leaders in Eldoret,Kenya. A month later, the Resident HumanitarianCoordinator, accompanied by representatives from theUN Children’s Fund, the World Food Programme, the UNDevelopment Programme and OCHA, held meetings inKismaayo with the JVA authorities, elders and religiousleaders and civil society groups (local NGOs, women’sand youth groups). This, and renewed interest by anumber of agencies in Lower Juba, has provided anopportunity to address some of the concerns raisedduring the June 2002 workshop.

This paper is based on the findings of field research onlivelihoods, protection and IDPs in Kismaayo and the LowerJuba Valley, conducted in May 2003 under the auspices ofOCHA-Somalia. The research had three main aims:

1

Chapter 1

Introduction

NP 44 crc 12/1/03 4:43 PM Page 1

livelihoods and protection: displacement and vulnerable communities in Kismaayo, southern Somalia

2

• to obtain a clear understanding of the situation of IDPsand other vulnerable communities in Kismaayo, andthe issues that they faced;

• to obtain a clear understanding of the operatingenvironment in Kismaayo and the areas from which thedisplaced originated; and

• to develop an operational plan to better protect andassist the internally displaced and other vulnerablegroups. Crucially, this plan was elaborated within alivelihoods and protection framework. Ideas for puttingthe framework into practice were initially developed bythe research team, but this work also draws uponplenary discussions during a workshop in Nairobi on 11June 2003 to disseminate and discuss these findings,sponsored by OCHA-Somalia.

The sensitive nature of the information contained in thisreport means that the names of informants and nationalstaff participating in the fieldwork have been withheld.

The paper begins with a brief description of the livelihoodsand protection framework that informed the research(Chapter 2), and the methodology the research employed(Chapter 3). Chapter 4 provides a brief overview offindings. Although this report is critical of the current JVAadministration, many of the same issues applied underprevious occupations. Chapter 5 describes the concept ofa phased operational plan. Lastly, conclusions are drawn inChapter 6.

NP 44 crc 12/1/03 4:43 PM Page 2

3

What is a livelihoods approach?

Several developmental and humanitarian agencies anddonors are exploring the livelihoods concept and what itmay mean for their policies and practice. These include theWorld Bank, the UK’s Department for InternationalDevelopment, UNDP, the Food and AgricultureOrganisation and WFP, as well as NGOs such as Oxfam,Save the Children (UK) and CARE.5

The key element in a livelihoods approach is its over-ridingemphasis on how people actually live.6 The principles thatunderlie the approach are:

• it puts people at the centre of analysis and action; • it is holistic, recognising that there is a multiplicity of

actors, influences, livelihood strategies and outcomes;• it recognises that livelihoods and the factors that

influence them are dynamic, which means that it isinherently flexible;

• it is multi-level, in that it tries to bridge the gapbetween micro and macro factors and incorporate anumber of different activities; and

• it is cross-sectoral in its approach to programmingthese activities.

Livelihoods approaches draw attention to the way peoplelive, why they live that way, and why and how this way oflife changes. The questions it asks indicate likely entrypoints for interventions – both short- and long-term – andtheir likely impact on people’s lives.

To date, livelihoods approaches have been predominantlyused in academic analysis and NGO practice under theguise of rural development in peaceful settings. Theiradaptation for use in situations of chronic political

instability, where violence and poverty have becomeentrenched and where people’s livelihoods are persistentlyor purposefully threatened or undermined, is moreproblematic.8 In terms of adapting the framework to makeit more appropriate for these situations, the emphasis hasbeen on the concept of vulnerability, components ofconflict and power relations and their temporaldimensions. The contextual nature of vulnerability is nowcentral to the livelihood model. This involves integrating apolitical economy approach into the livelihoodsframework, looking at vulnerability both as a factor ofmaterial assets, and in terms of powerlessness, and howthis changes over time.9 It also requires an understandingof these processes at the micro level, from the household(however defined), through to the macro-level processesthat shape vulnerability and livelihood strategies.

This study includes notions of humanitarian protection inthe analysis (see below) precisely as a way of describingthis aspect of vulnerability. It is used also as a means ofidentifying the human rights claims of claim-holders andthe obligations of duty-bearers, as well as the immediate,underlying and structural causes of exploitative or unequalrelations and the non-realisation of rights. This shouldthen indicate potential solutions to address this non-realisation of rights, solutions which go hand in hand withinterventions designed to support material needs.

Moving beyond saving lives to adopting a livelihoods-oriented approach amid chronic conflict is clearly fraughtwith difficulties across all phases of activity, frominformation gathering to the provision of assistance.10

Investigations of livelihoods are investigations into theworkings of human society, and human societies arecomplex – so complex and diverse that they easily breakout of any attempts to confine them within neatly-drawnframeworks, categories and definitions. They are alsodynamic, in a state of constant change. The approach is sobroad that problems may arise in identifying the mostimportant issues. Add to this the difficulties of working inconflict environments like southern Somalia, and it is clearthat putting this framework into practice is notstraightforward.

There is also the problem of overcoming the ‘relief-to-development’ mentality (and the so-called ‘livelihoods’gap) that bedevils assistance in Somalia and elsewhere.11

In Somalia this led to a somewhat artificial division of thecountry into three zones, in theory reflecting levels of‘political maturity’ and donor budget lines: crisis (in thesouth of the country), transition (the middle) andrecovery (the north, basically ‘Somaliland’ and‘Puntland’). Although this has been to a degree replacedby a distinction between the stable northern and theconflict-prone southern regions, the opportunities in the

Chapter 2

The analytical framework: livelihoods and protection

A widely-accepted definition of livelihoods was given byRobert Chambers and Gordon Conway in 1991:

a livelihood comprises the capabilities, assets (stores,resources, claims and access) and activities requiredfor a means of living: a livelihood is sustainable whichcan cope with and recover from stress and shocks,maintain or enhance its capabilities and assets, andprovide sustainable livelihood opportunities for thenext generation; and which contributes net benefits toother livelihoods at the local and global levels and inthe long and short term.7

Box 1

Defining livelihoods

NP 44 crc 12/1/03 4:43 PM Page 3

livelihoods and protection: displacement and vulnerable communities in Kismaayo, southern Somalia

south for moving beyond relief-only activities, tosupporting livelihoods not just saving lives, and toaddress the dynamics of vulnerability beyond meetingimmediate material needs, have remained largelyunexplored.

There is here the potential for bringing together therelatively separate realms of humanitarian and developmentdiscourses, but in a more dynamic manner than reiteratingvague platitudes about disasters being merely an indicationof underdevelopment.12 An understanding of howhouseholds combine survival and risk-mitigation strategiescan reveal the fallacies and limitations of (non-)linkagesbetween humanitarian assistance and developmentcooperation, and in the case of situations of chronic politicalinstability, the simplistic and now-contentious andinappropriate ‘relief to development continuum’ model thatarose in the mid-1990s.

Livelihoods approaches are complex and difficult toimplement. But in situations of chronic political instabilityand recurring crises such as Somalia, can we afford not totake such a comprehensive approach? In place of short-term and frequently ad hoc humanitarian relief, livelihoodapproaches provide a contextualisation of need that isoften lacking in relief needs assessments.

What is humanitarian protection?

There is increasing recognition that protection, or rather itslack, lies at the heart of conflict-induced humanitariancrises. This has led to a renewed interest in internationalhumanitarian, human rights and refugee law.13 The UN hasadopted several resolutions on the protection of civiliansin armed conflict, and on the protection of children andwomen. UN Secretary-General Kofi Annan has introducedthe notion of the ‘Right to Protection’, and the SecurityCouncil has debated the scope of the internationalcommunity’s obligation to protect civilians in the face ofhuman rights violations.14 The ICRC – the onlyhumanitarian organisation formally mandated by theGeneva Conventions to intervene on behalf of civilians orsoldiers hors de combat in armed conflict – engages in avariety of protective actions. The Inter-Agency StandingCommittee (IASC) has also adopted guidelines and policydocuments on the protection responsibilities ofinternational humanitarian agencies.15

One such operational instrument, the UN Guiding Principleson Internal Displacement, combines existing internationalhuman rights and humanitarian law to provide guidance forthe assistance and protection of IDPs.16 Without being aconcrete action plan for humanitarian organisations in thefield, the Guiding Principles have at least become a usefulframework for the implementation of humanitarianprogrammes, as well as strengthening and complementingthe protection of displaced people. The Principles cover allphases of the IDP problem, including how to preventdisplacement, how to provide protection and assistanceonce people have been displaced, how to achieve their safe

return with dignity, and alternative resettlement andreintegration.

The primary goals of humanitarian work are to protect lifeand health, prevent and relieve suffering, and ensure thathuman beings are treated with dignity. The right to receivehumanitarian assistance – and the right to offer it – arefundamental principles underpinned by international law.18

In situations of chronic political instability, humanitarianneeds are inextricably related to processes of violence. Inthese instances, violence is more than usually associatedwith the failure of national or other public actors includingthe international community to regulate violations offundamental human rights. But protection has to be muchmore than an arid legalistic concept founded ininternational humanitarian law and human rightsinstruments, and divorced from the reality of humanitarianpractice.19 Humanitarian protection in practice mustaddress violations at different levels.

Protection is also closely associated with the fundamentalidea of responsibility and the associated concept ofaccountability.20 This takes two forms. The protection ofvulnerable communities is, first and foremost, theresponsibility of the state and national authorities. Insituations of internal conflict, this same obligation adheresto non-state actors such as insurgent or armed groups.There is increasing recognition that international and non-governmental agencies bear a responsibility to identifyand deliver appropriate assistance to enable betteroutcomes for the recipients of aid, and further theirprotection.21 This means not adding to the burden of thedisplaced and exacerbating tensions or violence in analready-contested environment. It also meansendeavouring to adhere to the principles and standards ofhumanitarian action, such as impartiality and neutrality.22

The Guiding Principles do not seek to create a privilegedcategory of person: they are based rather on theassumption that the displaced have the same rights andobligations as anyone else.

4

One definition of protection (and the one used here),elaborated in an ICRC-sponsored workshop, is from theIASC. Protection activities are defined as:

Any activity – consistent with the above mentionedpurpose – aimed at creating an environmentconducive to the respect for human beings,preventing and/or alleviating the immediate effects ofa specific pattern of abuse, and restoring dignifiedconditions of life through reparation, restitution andrehabilitation.17

Box 2

Defining protection

NP 44 crc 12/1/03 4:43 PM Page 4

chapter 2 The analytical framework

5

Protection and the provision of assistance have bothinternal and external aspects. For external actors, thismeans developing an understanding of governance (in itswidest sense) in any given displacement context: what arethe mechanisms of power and the structures within andbetween different stakeholders and stakeholder groups?Identifying who is responsible for protection at all levels isthe first step in making them accountable, both tovulnerable groups and to the international community. Byimplication, in order to achieve protection in practice ashift in power relations is required.

Livelihoods and protection: bringing the two together

In many ways, livelihoods and protection can beconsidered two sides of the same coin. In this approach,humanitarian action and humanitarian protection informedby livelihoods analysis are not seen as separate activities,but as a complementary and synergistic approach, whichintegrates the ‘software’ of international humanitarian lawand human rights advocacy with the ‘hardware’ of materialresource provision.23 Ideally, this would mean reducing (orpreferably eliminating) the negative impacts of aid. Fine-grained analysis through the acquisition of knowledge andinformation should lead to fine-grained or more sensitiveprogramming and resource provision.

This requires an understanding of the nature of conflictand the place of resources – including aid – within it.24

Sensitivity to the context and contestability of aid isimportant if the international community is to reducecriticisms that aid fuels conflict.25 Without this holistic andcontextual understanding, the provision of resources mayupset the already fragile (or exploitative) social balance, ormake things worse than they already are, including, atworst, leading to an increase in violence and ultimatelyadding to the humanitarian and protection burden of thosethat we set out to assist.

In order to achieve humanitarian protection, it is necessaryto analyse the connections between those in authority andspecific target groups, so as to explore why a group hasbeen deprived of its rights. The livelihoods approachfocuses more on the constraints that prevent people from

realising their rights as a prerequisite for promotingpeople’s livelihoods. Although these rights are notspecifically outlined in livelihoods approaches, they are anintrinsic component of them.

Livelihoods approaches help direct protection activities inpractice. Adding protection to livelihoods analysis helpsfocus (or direct) the investigative process of fieldwork.Every IDP situation is different. Livelihood approachesallow us to contextualise these differences and identify thesocial, political and economic structures among IDPs, thehost community and those in authority. The livelihood andprotection framework helps both to meet basic needs, andto address more underlying protection-related problems.

This study suggests that the added value of a livelihoodand protection framework is six-fold:

1. The identification of entry points over the short andlonger term.

2. The more efficient use of resources.3. An increased positive impact on beneficiaries.4. The reduced risk of doing harm through the provision

of aid.5. The opportunity to build a more conducive environ-

ment for human rights.6. The increased opportunity for actually doing good.

What can be achieved in practice depends very much uponthe development of protection responses. After all,protection in theory (or policy) is quite different fromprotection in practice. This in turn depends upon thewillingness of people and agencies to get involved in thefield. It depends upon the context of conflict and theenabling environment within which protection activitiesare undertaken. Even though problems may be acute andimmediate, they may take months or years to address.

The combination of humanitarian protection and anadapted livelihoods framework, which includes a politicaleconomy component, thus provides the opportunity forinterventions that meet more than basic needs: it altersand shapes the functions of aid, and the relationshipsbetween humanitarian organisations and those who areresponsible and accountable for protection.

NP 44 crc 12/1/03 4:43 PM Page 5

6

NP 44 crc 12/1/03 4:43 PM Page 6

Protection in theory is one thing; operationalisingprotection and incorporating protection into humanitarianresponse is quite another, especially when authorities areunwilling or unable to live up to their responsibilities. In acontested environment where it is believed that protectionissues are not foremost in the minds of the authorities,research in this area is highly problematic. Similarly, wherethe provision of humanitarian assistance is perceived byone or more sides to the conflict to be politicallymotivated, or when the resources themselves arecontested, problems may arise. These concerns dictatedthe research methodology employed here, and thesubsequent dissemination of information, where nationalstaff involved in the research process and key informantsmay face threats to their personal safety.

How to ‘do’ livelihoods and protectionresearch

In the literature, there is little consensus about how to ‘do’livelihoods research in situations of chronic conflict andpolitical instability.26 While livelihoods models provide aconceptual background, there is no guidance on thespecific questions that need investigating; genderanalysis, for example, is not an explicit component of mostlivelihoods models. Gaps are left to the researcher orresearch team to fill.

In Kismaayo, the approach used was governed bypragmatic issues of access, as well as questions aroundthe types of data that could be collected, and from whom.One of the inherent contradictions of livelihoodapproaches is that, to be able to ask the right questions,researchers need to have prior knowledge of the issues tobe investigated. Paradoxically, this may mean knowing theanswers whilst endeavouring to abandon anypreconceptions before entering the ‘field’. As in allresearch, it is important that the methods used to obtaininformation and data are clearly explained in order tovalidate findings. This study is no exception. While it wasthought that predominantly qualitative methods were thebest way to engage with the complexity of issue-basedlivelihoods research, the techniques employed weredictated by the need for methodological flexibility toovercome these and other constraints.

While questions remain over the acceptability (or evenvalidity) of qualitative issue-based research (such as thedifficulty of replication), there is a clear need to movebeyond ‘fieldwork by chatting’, or a set of randomly-executed interviews. Rather, it demands a rigorouslythought out and structured framework for the explorationof displacement and the operating environment, whichgets around potential ‘gatekeepers’ (those with interests

in distorting information), and illuminates the hiddenvoices of the vulnerable or marginalised. The triangulationof information is necessary in these settings.

The use of secondary as well as primary informationsources is particularly important in contestedenvironments where information is sensitive (for example,on illegal activities), the protection environment is poor orthere is clear evidence of bias. A literature review wasundertaken of several key themes and regions: Kismaayo,the Lower Juba, IDPs (the history of UN and NGOengagement, the successes and failures in Somalia andelsewhere), and on the theory and application oflivelihoods, and the theory and application of protectionand humanitarian assistance. In Nairobi, interviews wereconducted with representatives from organisationsengaged in the Lower Juba, or with an interest in re-engaging; with individuals with an interest in livelihoodapproaches; and those with knowledge of the Lower Jubain general and Kismaayo specifically. This first phaseprovided the background information from which the semi-structured questionnaires were derived.

Theoretical and practical objectives were balanced againstother constraints such as time, the availability of staff(especially suitably qualified female national staff ),personal safety and health, and restrictions in movementand access to particular locations and potentialinterviewees.

The fieldwork

Fieldwork, sponsored by OCHA-Somalia, was conductedfrom 11 to 26 May 2003, with the full consent of the JVAleadership. A team of six national staff (four from OCHA-Somalia and two from the Food Security Assessment Unit(FSAU), all male) was brought together for this phase. Itwas hoped to include a specific component on the linksbetween gender and protection (with a national femalestaff member from UNICEF Somalia and one from MuslimAid UK) within the wider study, but this was cancelled dueto increasing insecurity in Kismaayo during the visit.

After an initial team-building exercise, which included abriefing on the objectives and approach of the study and asuperficial ‘ground-truthing’ to ensure that the Englishused in the field questionnaires had a correspondingmeaning in Somali, an initial broad overview andsubsequent breakdown of the IDP camps by relative sizeand clan composition was completed. Interviews were thenundertaken in what were considered representative campsby teams of two enumerators. Semi-structured focus groupdiscussions were held with IDP camp committees,whenever possible including women. After mapping the

7

Chapter 3

Methodology

NP 44 crc 12/1/03 4:43 PM Page 7

livelihoods and protection: displacement and vulnerable communities in Kismaayo, southern Somalia

8

camp’s social and economic structures, further interviewswere conducted with representative households. This two-tier approach was used in an effort to negate or bypass theinfluence of potential aid ‘gatekeepers’. (For a furtherdiscussion of the role of these gatekeepers see thefollowing chapter, and for the questionnaires see theappendices.) For obvious reasons, interviews in all campswould have been preferable, but time and the securitysituation did not permit this. A number of camps were notvisited in the Galjeecel clan area of Kismaayo due tofighting between rival militia groups; in all, in-depth focusgroup and household interviews were conducted in 11camps. Fifteen household interviews were completed.Debriefings were held at the end of each day to discuss thefindings and problems, to adjust the approach if necessary,and to reach a common analysis of displacement, theprotective environment, and potential activities.

In all previous studies of IDP camps, no mention has beenmade of the timing of interviews. This study suggests thatthis is a major weakness, with important implications forthe demographic profile of interviewees, the perceived andobserved demographic make-up of camps and theprovision of humanitarian assistance. Income generationtakes precedence over answering questions – thelikelihood is that those working or seeking employmentwill be out of the camps during working hours. There arealso opportunity costs involved in participating ininterview sessions. To minimise disturbance and todetermine who would be available for interview sessions,the roles and daily work burden of men, women andchildren were established, and the schedules for campcommittee and household interviews were arranged toaccommodate these activities and routines. This oftenmeant conducting interviews early in the morning or late inthe evening, before or after work. This limited the numberof interviews that could be conducted during this two-week period.

Although the focus of this study was Kismaayo, it was feltthat the IDP issue could not be fully understood without atleast some attempt to conduct research in other settings ofdisplacement. Two villages from the east and west bank ofthe Juba river near Jamaame were chosen because theywere accessible, and judged by the team members to berepresentative of the Juba Valley more widely.

One of the main differences between this work andprevious studies or assessments is the way that a ‘camp’was defined. This study used a definition that includedseveral households of displaced persons in a formergovernment or publicly-owned building, such as a bank, aswell as much more immediately recognisable structures,such as walled camps like Dhumaase, with over 350households. Perhaps understandably, and given thedynamic nature of displacement in the Lower Juba, thetotal number of camps identified in this study (23) is atodds with previous assessments, where even rudimentarydefinitions are not supplied: 26 (ICRC) and 17(UNCU/OCHA).

A further weakness of much livelihoods research is aninability to combine micro- with macro-level analysis of thewider dynamics of conflict (including its political economy)and, for IDP research in particular, the multiple dimensionsof displacement and the wider processes of vulnerability ingeneral. Displacement should be analysed in a widercontext to consider what has led to it, what its conditionsare, and how or why this is likely to change. Displacementis a symptom of profound problems, but it may also be anoutcome of positive choice. Generalised discussions aboutdisplacement do not illuminate the micro-level processesof livelihood patterns. Ultimately, therefore, this researchis grounded in the experiences and perspectives of thepeople affected by displacement: it is a ‘view from thedoorstep’, an explicit recognition of the complexity anddiversity of vulnerable people themselves.

Similarly, most programmes for IDPs are based on theassumption that they always want to go home (in Somalia apoorly understood notion in itself ). Although this mayindeed be the case, it cannot be taken for granted. Nor canit be assumed that IDPs are disconnected from the worldoutside. The idea that the IDP camp (or for that matterrefugee camp) is somehow a ‘sealed container’ plagues theliterature, but we should not assume automatically that tiesto ‘home’ (wherever it may be and in whatever form theymay take) have been broken. Moreover, displacement maynot necessarily be a temporary interruption to normal life –it may also be an opportunity for long-lasting change. Thisprompted several questions put to the displaced: where ishome, why did you choose Kismaayo, why this particularcamp, why have you moved, why don’t you return, can youreturn, if you return what are the potential problems thatyou face, what connections do you have with home?

The answers to these questions have repercussions for theways in which assistance is conceived and implemented:two extremes would suggest that the displaced may beassisted at the site of refuge with the knowledge that thisis where they will permanently settle; or there is arecognition that they are in limbo, until such time asconditions are ripe for return so that their immediate basicneeds are met and little else. An approach somewhere inthe middle would meet basic needs, but would also makean investment in future return by raising the levels of the‘mobile’ livelihood assets of human capital (education andskills training, for example). Answers to these questionswould also acknowledge the role of aid resources, byencouraging further displacement because of theattractive force of aid, encouraging return by undertakingresource activities in ‘home’ areas, or even by encouragingpermanent settlement.

These levels of analysis are brought together in Figure 1. Byusing a series of porous circles, this figure endeavours toportray the dynamic nature of displacement and break awayfrom the notion of the IDP camp as sealed container. Startingfrom the centre of the large circle, the diagram describes therelationship between individuals (small solid circles) withina household, between households within a camp, and then

NP 44 crc 12/1/03 4:43 PM Page 8

9

between camps within Kismaayo. It also shows the possibleconnection to home areas and to the diaspora (for examplethrough remittances), and the relationship between IDPsand the host community, the authority (in this case the JVA)and the aid community. These relationships are, in turn,shaped by the security, political and business environment,in Kismaayo and beyond.

Who is an IDP?

Despite the widespread promotion of rights and protectionfor IDPs, the definition of the term ‘IDP’ itself, at least inpractice, is fuzzy. This lack of clarity poses problems forresearch, analysis and subsequent action, as previousstudies in Somalia have indicated.27 Internally-displacedpeople are, according to the Guiding Principles:

persons or groups of persons who have been forced orobliged to flee or to leave their homes or places ofhabitual residence, in particular as a result of or in orderto avoid the effects of armed conflict, situations ofgeneralized violence, violations of human rights ornatural or human-made disasters, and who have notcrossed an internationally recognized State border.

How useful for this study is this definition? One of thecentral problems is that the debate on internaldisplacement is dominated by the actors who ‘invented’the IDP category in the first place. Endeavouring to render

the complexity and disorder of a disaster into a coherentsystem and an enabling environment, internationalbureaucracies define institutional tasks and create spacesof action. This creates and gives form to a particularfunctional interpretation of social and humanitarian reality.Ultimately, this will have implications for who is includedand excluded from this process. Operational clarity andeffectiveness depends upon the ability to identify, andthen select and deliver resources to, those identified to bemost in need on the basis of how they fit somepredetermined criteria, in this case ‘IDP’. These labelsbecome defining terms for aid agencies, creating divisionsof labour as actions become compartmentalised, with theirown set of defining terms.

It is clear that the IDP label as it is applied in Somalia lackscohesion and is plagued with difficulties, not least becauseof the complex and dynamic patterns of populationmovement within the country (see Chapter 4). In southernSomalia in particular, these labels become part of thecommon vocabulary of the disaster-affected in a bid to gainaccess to resources. It may be obvious, but in reality IDPsdo not carry signifying markers of identification. They areoften, but not always, indistinguishable from other groups,such as the urban poor. In Kismaayo, the urban poor ofteninhabit the same areas as those considered better off.What the homogenising label of IDP obscures may be moreimportant than what is brought into focus. Given thedifficulties in identifying and then locating urban poor and

chapter 3 Methodology

Figure 1 The Kismaayo IDP complex

Host

Security

Aid

BusinessDiaspora

‘Home’

?JVA Political

NP 44 crc 12/1/03 4:43 PM Page 9

livelihoods and protection: displacement and vulnerable communities in Kismaayo, southern Somalia

other vulnerable groups, this study, by necessity and likethose before it, focused on so-called ‘camps’ and thoseliving in them.

While the label IDP serves its purpose in drawing attentionto the plight of the displaced, and in the political sense ofmobilising resources on their behalf, its utilitarian value asa prescriptive operational label for the assessment andmonitoring of needs and the provision of resources isdubious. Obviously, clear distinctions need to be drawnbetween movement and the forced or stress migrationevident in the outpouring of people from areas of conflict ordrought as described in the Guiding Principles. While intheory the people-centred focus of livelihoods approachesenables this, in practice the practicalities and trade-offs ofthis research meant that the camps became the centralfocus. Therefore, those outside of the camps were generallyexcluded from the information-gathering exercise.

What remains, especially in terms of protection, is adetermination of the degree of access to justice,employment, health and education facilities, clean water andsystems of governance of the people in the camps. Thedisplaced are, by the very fact that they no longer live in theirhome, and by inference clan, areas, especially vulnerable interms of protection. Vulnerability and the degree ofintegration into the host community then become thedefining terms of study. This, in turn, widens the information-gathering exercise, and starts to bridge the potential dividebetween host and displaced communities.

Participation and the research environment

Although livelihoods approaches are consideredparticipatory, the type and level of overall participation inthis study was determined by the context in which theresearch took place. This included not favouring one groupover another, or one clan over another. One extremeexample of this methodological inclusiveness was evidentin interviews held with representatives from the twohuman rights organisations present in Kismaayo: onefavouring the Marexaan-led JVA alliance and the secondthe former Harti-led occupation of General Mohamed SaidHersi ‘Morgan’. But it meant, at times, excluding those whowould hinder information flows; so that elders of ‘minority’clans could talk freely, meetings were held in neutralvenues and behind closed doors. Efforts were made toinclude IDPs as much as possible, but businessmen,religious leaders, staff from local NGOs and other lessprominent figures were also heard.

Interviews with multiple actors and between multiple sitesallowed for a crosscheck or triangulation of findings,although this did not necessarily clarify who was being

truthful or not. It did, however, help to make sense of thecontradictions and ambiguities both within and betweenindividual accounts. In this way, a more complete andsubtle picture of displacement was pieced together thanwould have been possible with answers taken solely fromthe displaced (especially in the presence of aidgatekeepers). The likelihood that the prospect of aid mayinflate answers is an unfortunate element that must beconsidered in future studies.

Further issues and constraints

The use of the livelihoods and protection framework as theanalytical approach met with several obstacles. Teammembers from different backgrounds each faced learningchallenges. This was particularly noticeable for the OCHAstaff when introduced to the livelihoods framework. Allteam members found the rights-based approaches toprotection novel.

Second, despite our best efforts, given the short periodof time to build trust and break down barriers, it was feltthat certain discussion themes – the numbers of IDPs, thedeplorable state of sanitation, lack of education andhealth facilities, or population estimates – were beingdeveloped at least partly for effect, and the teammembers were being addressed as potential benefactors.In addition, a constant theme was interview fatigue anddisappointed expectations – after numerous studies ofIDP camps in Kismaayo over the years with little concreteto show in terms of resources, there is evidentdissatisfaction with agencies. Efforts must be made tocoordinate information-gathering exercises to minimisethis.

Third, although efforts were made to conduct interviews inmultiple sites, the lack of geographic spread throughoutthe Lower Juba is a noticeable weakness of this study.Ideally, more time would lead to greater coverage(including the Bajuni islands, which appear to be ignoredin assessments).

The last problem is to do with language. Although it ispossible to function at certain levels using only Af-Soomaali (or Af-Maxaatiri), the official language duringcolonial and pre-war administrations, Bantu client groupsspeak Af-Maay, Bajuni islanders speak Swahili-basedKibajuni, and the Mushunguli of the Juba river speakMushunguli. Effective and nuanced communicationrequires some knowledge of the appropriatelanguage/dialect if any type of socio-economic,monitoring or extension work is to be successfullyundertaken. However, only one member of the teamspoke Af-Maay.

10

NP 44 crc 12/1/03 4:43 PM Page 10

What emerges from this research is a picture of complexity(and the difficulties of undertaking activities in Kismaayo).This is perhaps not surprising given the researchframework. But this does not mean that the picture is socomplex that nothing can be done. It raises a number ofissues and potential entry points for further action.

Population movement and displacement in Somalia

The history of migration and conquest in Somaliainfluences everything, from dialects and language to thenature of ethnic or clan identity, land tenure and ownershiprights, and the fluidity of political and clan alliances.Population movement is a common and often ‘normal’feature of Somali society, notably among pastoral andagro-pastoral economies, where population displacementsare by no means random or limitless. In its grazing,permanent cultivating and trading centres, and above all inits wells and water points, every clan and group possessesa series of points between which movement rotates.Migration and household splitting do not necessarilyreflect an exception to the normal patterns of society. Inagricultural and agro-pastoral populations, householdmembers do not even necessarily live togethercontinuously year-round: the numbers present at any giventime depend largely upon the season, the nature of theproductive assets available to the household, economicand employment opportunities elsewhere (most likely tobe found in urban centres), kinship ties and the strength ofsocial networks, and the nature of the shock or stressexperienced.

At the height of the conflict in the early 1990s, over onemillion Somalis are estimated to have fled to neighbouringcountries in the region and outside of Africa.28 Peoplecontinued to leave southern Somalia in large numbers until1995. Others chose to move to safer areas within Somaliainhabited by their kin and extended families. Waves ofdisplacement, and even multiple displacements, of varyingmagnitude continue in some parts of southern Somaliadue to localised conflict, droughts and floods. Rapidurbanisation (or rural-to-urban drift in search of new orbetter livelihoods), refugee return and reintegration, andthe continuing return of people to their home areas due toa generally improving security situation make the picturestill more dynamic and complex.

Population movements into the Lower Jubaand Kismaayo: a multi-clan topography

In recent decades, like the rest of Somalia, settlementpatterns in the Juba Valley and Transjuba have beenanything but static. The Juba Valley in particular hasexperienced dramatic population movements over the last

150 years. A significant theme has been the migration ofSomali clans from the more arid central, north-eastern andOgadeni plains southwards in search of better pasturageand water. In the 1970s and 1980s, large-scale state farmsand agro-industrial and refugee resettlement projectsattracted settlers from other regions of Somalia.29

More recently, there have been distinct waves of migrationto Kismaayo (and to the camps), a consequence of conflictin the early 1990s, the El Niño floods in 1997, and conflictagain in 2002 and 2003. The main areas of origin havebeen the Juba Valley and Gedo region to the north,Mogadishu and, further afield, Galgaduud region to thewest, and Bay region to the north-east. The ‘pull’ ofpotential aid resources, social connections, economicopportunities and the ‘West Bank’ effect (thelegitimisation of occupation), all have drawn migrants tothe town. There is also continuous movement betweenKismaayo and the Juba Valley to market farm produce,seasonal migration for land preparation, and the familialmigration of resource sharing.

This migration has fundamentally altered the ethniccomposition of the area, affecting everything from access toagricultural land and aid resources to the politics ofadministration. The issue of guri (‘local’) versus gelti(‘outsider’) is a profoundly important undercurrent in theregion’s politics and in the competition over resources. Ithas also been problematic for the important issue of landtenure and access to agricultural resources. Although landgrab is not a new phenomenon during the past decade,Bantu communities on the west bank of the river Juba inparticular (notably between Kamsuuma and Kismaayo)have seen their land occupied by the Habr Gedir andGaljeecel and, around Buale, by Absame clans. Agriculturalimplements and water pumps have been looted, watermanagement systems have fallen into disrepair, and where

11

Chapter 4

Livelihoods and protection: study findings

Several Somali words hint at the complexity and subtletyof population movement, even in the absence of conflict:the term kiinaan describes the voluntary movement ofpeople in search of resources, from Bakool to relativesliving in Bay region; hayaan describes a long journey;and qaxooti describes someone who has travelled agreat distance and is destitute. Perhaps barakac comesclosest to a Western understanding of IDP – it describessituations where people are obliged to move after havinglost all their possessions.

Box 3

The language of displacement

NP 44 crc 12/1/03 4:43 PM Page 11

livelihoods and protection: displacement and vulnerable communities in Kismaayo, southern Somalia

land is no longer farmed it has reverted toscrub. Although some patron–clientsharecropping arrangements exist on landunder ‘occupation’, the collapse ofplantation agriculture has dramaticallyreduced income opportunities.

Alongside the dramatic decline in employ-ment opportunities due to the collapse ofplantation agriculture, there has been aconcomitant increase in economicmigration to the urban market centre ofKismaayo. This has resulted in themovement of large numbers of clan militiaand livestock:

1. Marexaan An estimated 4,000–5,000households have arrived from twomain directions: Awudwaaq district ofGalgaduud region and Gedo region.

2. Harti Since January 2003, an estimated 800–1,000families have arrived from refugee camps in Kenya(including Dadaab), from Mogadishu and from villagesof the eastern and coastal areas of Kismaayo district. Afurther 500 former militia from Baidoa district are alsothought to have arrived.

3. Absame From villages west of Jilib, from Buale andDobley and from some areas of Afmadow and Hagar,100–200 households have arrived. A further 200–300former militia have arrived in Kismaayo seekingemployment.

4. Ormaale As a consequence of fighting in Buale, 20families moved to Kismaayo.

Most observers agree that Kismaayo will be one of the lastplaces in Somalia to experience peace. This is attributable,among other things, to the large number of clans (between18 and 20), and competition for its relatively rich naturalresources and the strategic and economically important

seaport and airport. Although the rudimentary JVAadministration in place in Kismaayo has provided order inthe town (though some interviewees suggested that theJVA had at most 80% control of the militia) this seems to bedevised primarily to regulate the benefits of occupationand to prevent in-fighting within the dominant Marexaanand Habr Gedir clan duopoly.

The Kismaayo complex

Ultimately, the stability of Kismaayo depends upon afragile mix of political, militia and business actors thatshare a common interest – generating and using incomefrom the ‘taxation’ of port and airport activities. The controlof these resources is highly politicised (clan-based) andhighly contested. The sums at stake are relatively large: inApril 2003, for example, 23,000mt of sugar was imported,generating estimated ‘tax revenue’ of $202,400; 364,90325kg sacks of charcoal were exported, generating a ‘taxrevenue’ of $183,000. The over-exploitation of naturalresources, including offshore fisheries, is much inevidence, and customary management mechanisms havebeen over-ridden in the interests of profit. Many areas ofthe Juba Valley have been stripped of acacia trees forcharcoal production, the largest of the natural-resourceexport trades. Although this level of deforestation andenvironmental degradation such as soil erosion is notsustainable, wood biomass will remain the only availablesource of fuel for most Somali households. The charcoaltrade is tied to the survival strategies of some of the mosteconomically marginalised households, as well as topowerful vested interests. The trade providesopportunities for the urban and rural poor (cutting andproducing charcoal, processing and re-bagging inKismaayo, and portering and loading activities), as well asfor groups who have an interest in maintaining the trade(transporters, brokers, exporters, the JVA and the militia).

For all its problems, the urban centre of Kismaayo remainsattractive for those seeking employment and income

12

Somalia is one of the few virtually ethnicallyhomogeneous countries in Africa, if not the world.Nonetheless, it has its cultural and ‘ethnic’ splits. Thesubtleties and implications of socio-ethnic stratificationwithin Somali society have remained largely ignored. Anumber of minority groups exist, such as the Bantu (inthe Juba Valley divided into the Shanmbara and theMushunguli), Benadiri, Midgaan, Bajuni, Eyle, Tumal,Yibir and Galgaala, characterised not only by physicalappearance, but also by the stigma of occupationalstatus and long-forgotten grievances of ‘ritual un-cleanliness’. The latter are often indistinguishablephysically from the majority clans.

Box 4

Minority status and vulnerability

A charcoal store: wood biomass is often the only available source of fuel

NP 44 crc 12/1/03 4:43 PM Page 12

chapter 4 Livelihoods and protection: study findings

13

opportunities. Men are largely dependent on low-paid andshift-based work in the port (other activities include marketportering, the production of lime, quarrying, buildingconstruction, seasonal farming activities and charcoalproduction), and women upon low-paid work as domesticservants, selling water, collecting and selling firewood,house-mudding and seasonal farming activities. Children relyon shoe-shining, begging and the collection of discarded qatleaves for resale. The sub-contracting of labour is commonbetween stronger clans, such as the Marexaan, and the‘minorities’, and there are frequent accounts of the non-payment of wages. For Bajuni fishermen, gatekeepers mostlycontrol the revenue from fishing.

Conditions for IDPs

The IDP camps of Kismaayo are crowded, and mostshelters are rudimentary structures made of scavengedmaterials. They lack adequate sanitary facilities, and theincidence of communicable diseases appears to be high,although conditions vary between camps. Many latrineshave collapsed or are simply full, and have beenabandoned. In some instances, communal areas outside ofthe camps are used as informal latrines. The domestic useof contaminated water is a major contributor to poornutritional status. For groups along the Juba Valley unableto purchase clean water, the river provides for their needs.This greatly increases the risk of water-borne disease.

There is a common and strong desire for educationamongst IDPs and other vulnerable groups in Kismaayo (aselsewhere in Somalia). In Kismaayo, where there is no freeprimary or secondary education, access is determined bythe ability to pay for private facilities (Quranic schools costup to SShs10,000 a month per child, and non-religiousschools SShs20,000–40,000/month/child). For minoritygroups, including many in the camps, the ability to pay ispartly determined by socio-ethnic status.

The protection environment

Many of the problems facing IDPs and other vulnerablegroups in Kismaayo are related to the protectionenvironment. Individuals from weak and powerless clanssuch as the Bantu, Bajuni and Galgaala (including those inthe IDP camps) rarely enjoy the protection afforded toothers. These entrenched socio-ethnic divisions affectaccess to economic capital (such as employmentopportunities); this in turn affects the degree of access (orreinforces the lack of access) to education and healthfacilities.

There are numerous accounts of gender-based violence: ofsexual abuse in IDP camps and in the workplace. Generalabuse takes the form of theft, beatings, the non-paymentof wages and the constant reinforcement of socio-ethnicstatus (through the use of terms such as adoon or slave).When human rights violations take place, in the absence ofany properly functioning mechanisms for the rule of law,individuals from ‘minority’ or weak clans in Kismaayo oftenhave little recourse to systems of justice – whether throughcustomary law (xeer) or religious law (sharia). Thesubtleties of political, economic and social discriminationremain largely hidden to outsiders.

While it is easy to criticise a particular clan group orauthority structure for the treatment of minority groups inKismaayo, and Bantu and Bajuni groups have beenparticularly persecuted, the reinforcement of socio-ethnicstatus is not unusual in Somalia. Furthermore, during the1990s conflict between the factions led by Colonel OmarJess, Hussein Aideed and General Mohamed Said Hersi‘Morgan’ became the defining feature of the Lower Juba, asthe region endured some of the worst fighting in the civilwar. The area and the town of Kismaayo changed handsnumerous times. Each time, retreating militia looted andpillaged, and incoming militia did likewise.

Aid gatekeepers are an importantelement of the protection environment.These were seen as a potentialobstacle in the research process. Interms of aid, they are generallyconsidered a negative influence,positioning themselves prior to thedelivery of assistance to the camps totake advantage of potentially valuableresources. One interviewee suggestedthat the gatekeepers took as much as75% of the aid delivered to the camps.

At the same time, however, gate-keepers offer some degree of pro-tection to camp residents, especiallythose from weak or minority clans. ASomali proverb describes this situationwell: Ama buur ahaw ama mid ku-tiir-sanaw (‘Either be a mountain or leanon one’). In other words, members of

Kismaayo’s camps are crowded, and shelters are built of rudimentary materials

NP 44 crc 12/1/03 4:43 PM Page 13

14

weak clans seek protection from a strong one, in terms ofsheegata (adoption and client status) or, moreimmediately, the protection of a gatekeeper. Even thoughliving conditions are extremely poor, many of the displacedfeel safer as part of a group in a camp, receiving‘protection’ from a Marexaan gatekeeper. Moreover,economic migrants might consider living in a camp as low-cost housing: rents for an arish (a wooden shack) are in theregion of SShs40–60,000 per month, whereas a room in astone house would cost SShs100,000 a month.30 In returnfor this ‘protection’, and in lieu of ‘rent’, the gatekeeperwill receive a portion of the assistance packages allocatedto the displaced.

The implications for aid

One of the key issues raised during this researchconcerned the potential negative impact of aid. InKismaayo, aid resources are likely to be a source ofdispute, destabilising an already-fragile environment andreinforcing the existing socio-ethnic divisions and patternsof power. This raises a number of issues (and challenges)outlined below. One interviewee asserted succinctly that ifthese are not addressed, it would be like putting ‘fuel intothe fire’. Rather than simply ‘rushing in’ in an ad hocmanner, this suggests a more tentative and coordinatedapproach to committing resources, which incorporatesanswers to these challenges in an operational plan.

1. How does the international community work with a defacto authority like the JVA without conferringcredibility on its occupation and illegitimateadministration? Although this issue is not new inSomalia, the perceptions of other (excluded) groupsremain important.31 If this is ignored, the aidcommunity runs the risk of adding to the security threatfaced by international and national field staff.

Some interviewees suggested that potential aid flows intoKismaayo and the Lower Juba will be insignificantcompared to the revenue collected from the sea andairport. What is sought by the JVA is not materialassistance per se, but a reinforcement of their claim ofauthority in the wider Somali political arena. By workingwith the JVA clan duopoly in Kismaayo, the internationalcommunity is, in the eyes of the clans inside and outside ofKismaayo town (whether guri or gelti), legitimising thisclaim, albeit inadvertently. The international community isno longer perceived to be neutral in its engagement in theregion. This is reinforced by concentrating assistance inKismaayo to the exclusion of other clans and clan areas.

To counter this, a much more transparent and inclusiveapproach to engagement in the region, not just inKismaayo, is required. All clan parties in the region, guriand gelti, must be approached and brought intonegotiations on the provision of assistance. This is a muchhigher tier of engagement than that normally considered inSomalia (where negotiations are held with local authorities

only). However, this may lead to accepting some degree oftrade-off between identified needs and operationalpragmatism.

2. The international aid community in Somalia fallssomewhere between operational pragmatism and aprincipled approach (‘principled pragmatism’). But howpragmatic (or unprincipled) is the internationalcommunity prepared to be? There must be a bottomline below which agencies suspend activities orwithdraw. This is rarely defined, and the negativeimpacts of aid are rarely monitored. Consequently, theboundaries of operational pragmatism continuallyshift. Although this lack of clarity and consistencyallows the international community to continueworking in contested environments, this comes at acost: it allows those seeking to manipulate theresources of aid to operate between these fuzzyboundaries.

3. International assistance must be sensitive to the multi-clan topography of Kismaayo and the Lower Juba ingeneral. This means ‘thinking multi-clan’ at several levels:

• humanitarian agencies have a responsibility to ensurethat, in the hiring of national staff, vehicles and securityguards, clan affiliation is taken into consideration. Forexample, the hiring of vehicles from only one clan in amulti-clan environment will exacerbate underlying clantensions, increase the competition for resources, andreinforce existing power and wealth structures withinthe community. In Kismaayo, and in other parts ofSomalia, this lack of sensitivity has led to threats ofviolence against national and international staff;

• determining clan ‘boundaries’ in geographic terms andin terms of the clan make-up of, for example,administrations, local NGOs and community-basedorganisations to inform initial access negotiations (seeabove) and in the provision of assistance; and

• multi-clan projects should be considered, following theexample of several private enterprises in Kismaayo (inthe telecommunications and remittance sector and inthe provision of private health facilities). These haveproven less prone to looting during conflict as theinterests of several clans are affected.

4. Programmes must integrate protection and assistance.The incentive of humanitarian assistance provides anopportunity to put across more assertively the need forprotection for vulnerable groups and to improve theprotective environment. But how does the internationalcommunity work with or around entrenched socio-ethnicdivisions to ensure that aid resources most benefit thosetargeted? Similarly, how does the internationalcommunity work with or around aid gatekeepers?Ultimately, we can either accept some degree ofdiversion (the pragmatic approach) or enforce a zero-tolerance policy (a principled approach). However, bothrequire that effective monitoring and evaluation

livelihoods and protection: displacement and vulnerable communities in Kismaayo, southern Somalia

NP 44 crc 12/1/03 4:43 PM Page 14

chapter 4 Livelihoods and protection: study findings

indicators are applied and enforced. Furthermore, as theGuiding Principles suggest, the international communitymust be careful not to accord special privileges to targetgroups for fear of creating (or exacerbating) tensionsbetween them and the host community.

5. The focus of humanitarian assistance is on thoseconsidered most vulnerable. To encourage local‘ownership’ of project activities, the internationalcommunity should endeavour to use the ‘pillars’ ofSomali society (such as the ulamaudiin or religiousleaders, clan elders and the dynamic businesscommunity) to work for the benefit of the vulnerableand the implementing agencies. Using these ‘pillars’ toprovide resources (such as cash or credit), act asguarantors, or simply to facilitate access, transfers the

responsibility and risk of potential asset loss to theseintermediaries.

6. Due to the fragility of the security situation and themulti-clan topography of the region, ‘mobile resources’such as mobile health clinics or veterinary services maybe more appropriate than ‘fixed’ facilities such ashospitals. These could operate a routine multi-locationand multi-clan itinerary (Tuesday in Buaale, Thursday inAfmadow, and so on) and could incorporate healtheducation literature (on HIV/AIDS, female genitalmutilation and sanitation). In case of renewed fighting,they could be ‘evacuated’. This approach has the addedbenefit of addressing two common complaints: why doall the resources focus on Kismaayo; and why do theytend to focus on urban areas?

15

NP 44 crc 12/1/03 4:43 PM Page 15

16

NP 44 crc 12/1/03 4:43 PM Page 16

17

This paper argues for an approach to programming that isincremental, based on local realities and knowledge-basedoperations, and where individual activities, such asagreement on the principles of engagement anddisengagement or the provision of clean water andimproved shelter, are integrated within one holisticoperational plan. Because it seeks to address issues in aholistic and integrated manner, the operational frameworknecessitates a collaborative approach by a number ofagencies (from the UN and the NGO community) withdifferent, albeit overlapping, mandates and expertise.While this is a challenge in itself in terms of agency buy-inand effective coordination, it also provides the opportunityto phase the activities of agencies. Phasing is important fortwo reasons: first, as part of a logical process ofsequencing interrelated components for strictlyprogrammatic purposes; and second, to avoid the suddenarrival of large amounts of resources in an area that ispolitically fragile and potentially highly volatile, and wherecompetition for aid resources frequently results in conflictbetween rival groups.

Figure 2 (overleaf ) shows suggested phases and potentialactivities.32 The vertical thick line on the left shows initialre-engagement. The thick horizontal lines, for example thecreation of humanitarian space and basic needs, show themain phases within which potential activities (the thinlines) take place. Hashed vertical lines show where aphase or activity is initiated, for example an agreement onthe creation of humanitarian space and the start ofactivities to meet basic needs.

These cannot be seen in simple linear terms, with oneactivity starting after another has finished. In this plan,activities across several phases may take placeconcurrently in the same humanitarian space. What isimportant is the relative temporal position of activities andtheir connection to one another.

In Kismaayo, as in other contested environments, thecreation of an enabling and conducive environment for theprovision of material assistance and the promotion ofprotection must be developed to an acceptable levelbefore these resources are committed. In addition, acoherent approach by the international community is asine qua non for the success of principles of engagementand disengagement. These two aspects, together withskilful programming by well-informed individuals, areessential to develop a process for intervention that avoidsmany of the pitfalls associated with fragmentedapproaches with poor follow-up. This especially concernsincreasing the competition for resources, leading togreater instability and ultimately decreasing the impact ofaid. It should also be made clear that interventions in theoperational plan would be required to be implemented in

the ‘home areas’ of Lower Juba, rather than just inKismaayo. Equally, material assistance has to be providedto a wider community than just the IDPs, in terms ofneutrality and impartiality on the one hand, but also as apragmatic means to reduce the potential for conflictbetween groups, or the targeting of aid providers.

The first phases of engagement, then, would largelyinvolve ‘software’ activities that build a conduciveenvironment for more resource-based interventions. Indifficult operating environments, the first object of theseinitial activities would be to establish very clear groundrules, particularly with the authorities that control territoryand ‘public’ assets, such as the port in the case ofKismaayo. Ground rules would ideally identify modes ofcommunication through accepted focal points to avoid aplethora of individuals acting without the knowledge ofothers; agree on the independence of agencies, forexample in recruiting staff without interference fromauthorities or clan leaders; contextualise the normal‘privileges and immunities’ that are accorded UN agenciesand staff, such as the tax-free import of goods forhumanitarian programmes; develop complaintsprocedures; and agree on what level of reporting isrequired to maintain an acceptable level of transparency.As a central part of the development of ground rules, theroles and responsibilities of the various parties need to beclearly defined to reduce the risk of misunderstandings inthe future. It is important here for the various ‘non-stateactors’ to commit to ensuring a level of security thatenables the agencies to function. Internationally-recognised humanitarian principles would underpin theessence of the ground rules, and the process provides anopportunity to build the capacity of authorities andleadership on these principles.

The development of ground rules can lead logically to thefirst elements of building a protective environment. Oncethe ground rules have been developed, a process forbuilding the capacity of duty-bearers can begin, typicallyby promoting the Guiding Principles on InternalDisplacement. This would normally be led by OCHA, incollaboration with other agencies according to capacity.This can appear to be wildly optimistic in the kind ofpredatory operating environments like Kismaayo, andespecially where there is likely to be resistance to‘Western-driven’ instruments of international law.However, the engagement of religious leaders and arbitersof customary law (xeer) and secular law may place theGuiding Principles within a more appropriate culturalfabric. Of course, there is the potential for discrepancies inthe detail between the three bodies of law operating inSomalia (i.e., sharia, xeer and secular law) andinternational law, but the essence or spirit of the laws arecompatible and complementary. This process is also

Chapter 5

An operational plan of action

NP 44 crc 12/1/03 4:43 PM Page 17

livelihoods and protection: displacement and vulnerable communities in Kismaayo, southern Somalia

18

Figu

re 2

O

pera

tion

al p

hasi

ng a

nd p

oten

tial

act

ivit

ies

Tran

slat

ion

of a

ll do

cum

ents

Prot

ecti

ve e

nvir

onm

ent

Adv

ocac

y fo

r pr

otec

tion

Prom

otio

n of

HR

into

the

Con

stit

utio

n

Stru

ctur

al n

eeds

Agr

icul

tura

l reh

abili

tati

on

Inco

me

gene

rati

on

Dev

elop

alt

erna

tive

s to

the

cha

rcoa

l tra

de

Impr

ove

heal

th fa

cilit

ies

Mon

itor

ing

and

Eval

uati

on

Coor

dina

tion

Impr

ove

shel

ter

Prov

isio

n of

cle

an w

ater

Trai

ning

of p

olic

e

Reac

h ag

reem

ent

on t

he p

rinc

iple

s of

eng

agem

ent

and

dise

ngag

emen

t

The

crea

tion

of h

uman

itar

ian

spac

e

Stre

ngth

enin

g of

the

judi

ciar

y

Stre

ngth

en t

he v

oice

of t

he d

ispl

aced

Inte

grat

ion

of la

nd t

enur

e/la

nd r

ight

s in

to t

he C

onst

itut

ion

Dev

elop

an

unde

rsta

ndin

g of

the

deg

ree

of c

ompl

imen

tary

bet

wee

n xe

er, s

hari

a, IH

L, a

nd H

R

Bas

ic n

eeds

Cons

truc

tion

of p

it la

trin

es

Skill

s tr

aini

ng

NP 44 crc 12/1/03 4:43 PM Page 18

important in the gradual facilitation of better governancestructures, driven ultimately by Somalis.

Once the ground rules have been established, and workhas been initiated to build a protective environment, it isimportant to embark on more tangible programmes on theground. In Kismaayo, IDPs have very poor access to basicservices, and their living conditions have been describedas sub-human. Work with communities on improvingaccess to these services can make a real difference to howpeople live, particularly their physical well-being. Whileproviding services in Kismaayo, it would also be importantto balance assistance in the wider community, especially inthe ‘home areas’, as a way to reverse the pull-factor of thetowns. Relatively low-cost activities, such as theconstruction of pit latrines, the provision of clean waterand improved shelter, should be undertaken first. Onlywhen these are monitored and evaluated should moreresource-intensive activities follow. In Kismaayo, UNICEFwould be expected to play a lead role in the provision ofbasic services, working both directly and through localpartners such as Muslim Aid (UK). These kind of activitiesare high-impact and very visible, while requiring relativelylow levels of resources. They are also relatively safe frompotential manipulation. This can have a number of benefitsin addition to the obvious, especially in building theconfidence of the community. Having tangible programmeson the ground can also help in defending the software sideof the process from critics.

More ambitious programmes centred on the rule of law asa means to directly build the protective environment canbe phased in at this point. In Somalia, UNDP has an activeprogramme (Rule of Law and Security or RoLS)incorporating reform of the judiciary, police reform/training and disarmament, demobilisation andreintegration (DDR) components, all of which have greatpotential to address some fundamental issues affectingSomali society, especially vulnerable communities. HavingIDP issues at the heart of the programme would enable afocus on aspects such as improving access to the law forIDP women, or incorporating IDPs into the police forces,and including the Guiding Principles on InternalDisplacement as well as human rights law as an integralpart of police training. Again, it is important to emphasisethat activities related to the rule of law need to beextended to other parts of Lower Juba (and of courseelsewhere), as the protective environment for minorities isno better in rural areas than in Kismaayo. This level ofengagement obviously requires a high level of trust andgoodwill between the authorities and the internationalcommunity, and this can only be built up over time.Improvements in the protective environment would benecessary to enable the next level of assistance aimed atlivelihood support and development to occur with anymeaning. As discussed earlier, many of the causes oflivelihood insecurity are related to the protectiveenvironment and the inability of minorities/IDPs to accessservices and income opportunities as equals.

The next phase would focus on more structural aspects ofcommunities’ livelihood insecurity and disenfran-chisement. Projects linked to income generation (micro-credit, business skills training, literacy/numeracy) wouldaim to make it easier for vulnerable communities to gainaccess to income opportunities. At the same time, thisphase would also look at the integration of IDPs into themainstream of society, either in the place where they havechosen to live, or back in their home areas. The latterwould require work related to agricultural rehabilitationand development in the riverine areas. It would alsorequire research into land issues, which would need to beresolved if a return ‘home’ was to be feasible. Thecompetition for resources such as agricultural land is oneof the most intractable problems of the Somali conflict. Thehistory of settlement, migration and conquest has createda complex mosaic of land occupation and ‘ownership’within and between clan groups. There is little point inintroducing much-needed resources for the rehabilitationof agricultural land before these issues are tackled (see, inFigure 2, the intersection of the vertical line that indicatesthe completion of land tenure activity and the start ofagricultural rehabilitation). These factors may well providea pull-factor back to rural livelihoods, and would at leastgive IDPs a real choice about where they wish to live andinvest in the future. Certainly, the creation of betterconditions in the home areas would point the way to moredurable solutions than previous attempts to bus the IDPshome with a return package. This phase of the plan wouldalso have to address difficult issues such as the negativeeffects of the charcoal trade, especially on the environmentof the area, and look into ways in which alternative incomesources could be developed.

Some phases underpin others: the protective environment,monitoring and evaluation, and coordination. Theachievement of a protective environment involvescapacity-building and advocacy in their broadest sense.They require technical, material and financial support toequip local institutions and individuals with the resourcesand skills to become coherent advocates for change.Possibilities include bringing together hitherto separatecamp committees and incorporating them into theprotection dialogue and local systems of governance; andsample testing of rights through the use of questionnairesand focus groups, workshops, and training days. However,attitudes to socio-ethnic status require much more thanquick-fix solutions.

A monitoring and evaluation component is central to theefficacy of this approach. This paper has argued that manyof the problems facing IDPs and other vulnerable groups inKismaayo relate to the linkage between socio-ethnic statusand access to resources. This infers that monitoring andevaluation must move away from implementationchecklists, noting the number of wells constructed,tarpaulins delivered or people fed. What is needed inaddition are more subtle indicators that depict the realimpact of activities. These indicators could include the

19

chapter 5 An operational plan of action

NP 44 crc 12/1/03 4:43 PM Page 19

livelihoods and protection: displacement and vulnerable communities in Kismaayo, southern Somalia

20

degree of access to resources based on clanrepresentation (and particularly the degree of access thatminority clans can develop); the presence or absence ofaid gatekeepers and the degree of resource diversion; thescale of the security threat – to beneficiaries, staff and thewider community – or even changes in port activity. As thiscomponent underpins basic and structural needs phases,and indeed should provide the impetus for continuingengagement or disengagement, a systematic reportingsystem that is transparent and involves all stakeholdersshould be established.

Livelihoods approaches have clear implications forcoordination mechanisms. This paper suggests that,ideally, there would be three interrelated components.First, a multi-agency, multi-sectoral information-gatheringphase; second, a common analysis and the developmentof an agreed operational framework; and, lastly, theimplementation of activities, including the negotiation ofhumanitarian space, in a cohesive, phased andcoordinated manner. There are risks in the creation of

expectations among communities that are not quickly met,and difficulties associated with agency buy-in andencouraging joint programming. However, the question iswhether we can afford not to invest the time and effortrequired to develop programmes that deliver and havehigh short- and long-term impact.

Finally, livelihoods approaches challenge the sectoralmodel of coordination that currently exists in Somalia. Thispaper suggests that multi-agency zonal or regionalcoordination structures are more appropriate, though themobility of IDPs, and for that matter pastoralists and agro-pastoralists, means that the validity of rigidly applied zonaland regional units are problematic. Appropriate assistancecan only be designed by looking beyond boundaries,whether these are cartographic (district, regional andnational) or social (the porous nature of extendedhouseholds, for example), or even to do with mandate.One implication of adopting the livelihoods approach morewidely in Somalia would be a radical rethink of the currentcoordination structures, which are driven from Nairobi.

NP 44 crc 12/1/03 4:43 PM Page 20

In their synthesis study of livelihoods in situations ofchronic political instability, Catherine Longley and DanMaxwell raise three important questions.

1. How can livelihoods analysis be applied to conflictsituations in order to understand the impact of conflicton livelihoods? How should livelihoods frameworks beexpanded to incorporate conflict situations?

2. How can livelihood assessment tools be adapted andapplied in conflict situations? How can questions ofpower relations, war economies, human rights andhumanitarian principles be addressed?

3. What innovative approaches to livelihoodsprogramming can be developed by agencies insituations of chronic conflict?

While the study described in this paper did not set about toanswer these questions, in many ways it has addressedthese same issues in the complex political andhumanitarian environment of Kismaayo and the Lower Jubamore generally.

The study adapted the conventional livelihood wisdom byincluding a political economy approach to betterunderstand the impact of conflict on people’s livelihoodsand the power relations between IDPs and various othercommunities and external forces. In addition, theincorporation of ‘humanitarian protection’ considerationsdeepened this analysis by looking at power andvulnerability within a rights-based perspective, againwithin the overall umbrella of livelihoods analysis. The

combination of these approaches enabled thedevelopment of an operational framework that integratesmaterial (‘hardware’) support to livelihood needs with the‘software’ approach of protection through the promotion ofinternational humanitarian law, human rights advocacyand the rule of law. Ultimately, we are better able toadvocate for the needs of IDPs and the vulnerable ingeneral.

Evidently, the component elements of the operational planare not innovative; what is new is the integrated,collaborative and phased nature of their implementation.There is the potential here for an impact that issignificantly greater than that achieved by isolated, ad hocand localised approaches. Lastly, the adapted livelihoodapproach, in combination with the protection of humanrights, provides us with a dynamic understanding of theoperating environment within which the impact of aid canbe more accurately predicted and monitored.

How the operational framework is actually implemented inKismaayo and in the Lower Juba Valley, and the impact ofthe overall intervention, remains to be seen. There will bemajor challenges both within the aid community and in theoperating area. Certainly, since this approach transcendsspecific agency mandates, ownership by implementingagencies and by donors will need to be fostered if this is tobe accepted as an appropriate way forward. What can beachieved in practice depends very much upon the degreeof buy-in to the process. Careful monitoring and analysiswill be required to fully learn the lessons of this initiative.

21

Chapter 6

Conclusions

NP 44 crc 12/1/03 4:43 PM Page 21

22

NP 44 crc 12/1/03 4:43 PM Page 22

23

Appendix 1

Camp committee/focus group questions

INTERVIEW DETAILS

Name of interviewer(s)

a. Location of interview (name of camp or area)b. Date and time of interviewc. Composition of group