Moving Up or Moving Out? A Rapid Livelihoods and Conflict Analysis in Mieso-Mulu Woreda, Shinile...

8/8/2019 Moving Up or Moving Out? A Rapid Livelihoods and Conflict Analysis in Mieso-Mulu Woreda, Shinile Zone, Somali R

1/100

Strengthening the humanity and dignity of people in crisis through knowledge and practice

A p r i l 2 0 1 0

Moving Up or Moving Out?

A Rapid Livelihoods and Confict Analysis in Mieso-Mulu

Woreda, Shinile Zone, Somali Region, Ethiopia

Andy Catley and Alula Iyasu


2/100

2010 Feinstein International Center and Mercy Corps. All Rights Reserved.

Fair use o this copyrighted material includes its use or non-commercial educational

purposes, such as teaching, scholarship, research, criticism, commentary, and news

reporting. Unless otherwise noted, those who wish to reproduce text and image fles

rom this publication or such uses may do so without the Feinstein International

Centers express permission. However, all commercial use o this material and/or

reproduction that alters its meaning or intent, without the express permission o the

Feinstein International Center, is prohibited.

Feinstein International Center

Tuts University

200 Boston Ave., Suite 4800

Medord, MA 02155

USA

tel: +1 617.627.3423

ax: +1 617.627.3428

fc.tuts.edu
https://wikis.uit.tufts.edu/confluence/display/FIC/Feinstein+International+Centerhttps://wikis.uit.tufts.edu/confluence/display/FIC/Feinstein+International+Center


3/100

Acknowledgements

The authors would like to thank the ollowing Mercy Corps sta who

assisted with organisation o the study or who acilitated ocus group

discussions in Mulu: Mesn Ayele, Mohammed Haji, Niama Ibrahim,

Berhanu Esehete, and Yigezu Solomon. Other Mercy Corps sta

provided helpul advice on the design o the work or commented on

the drat report, including Abdi Aden, Fasil Demeke, Raael Velaquez,Nigist Tilahun, and Sarah Gibbons. Demeke Eshete o Save the

Children UK provided ood economy reports or Shinile Zone. We

are also grateul to participants in ocus group discussions in Mulu and

to the various government sta who provided inormation and joined

the livelihoods analysis training. Yacob Aklilu and Berhanu Admasu o

Tuts University also provided useul comments and advice. The study

was unded by the Foreign and Commonwealth Oce Bilateral Fund,

British Embassy, Addis Ababa.


4/100


5/100

Summary 6

1. Introduction 7

1.1 Livelihoods and confict in pastoralist areas 7

1.2 Background to the analysis in Mieso-Mulu woreda, Shinile Zone 7

1.3 Methodology and process 8

2. Using the Livelihoods Framework to Integrate Livelihoods and Conict Analysis 10

2.1 The livelihoods ramework 10

2.2 Policies, institutions, and processes: the PIPs analysis and confict 13

2.3 Confict and livelihoods programming 16

3. Livelihoods Analysis in Mieso-Mulu Woreda 17

3.1 Assets 19

3.1.1 Financial assets 19

a. Livestock herds and household income 19

Background notes 19

Income rom livestock and other sources in Mieso-Mulu woreda 20

b. Livestock herds and human oods 22

Background notes 22

Livestock herds and human oods in Mieso-Mulu woreda 22

c. Markets and trade 23

Background notes 23

Markets and trade in Meiso-Mulu woreda and beyond 23

3.1.2 Natural assets 25

a. Environment, rainall, and mobility 25

Background notes 25

Environment and mobility in and around Mieso-Mulu woreda 25

3.1.3 Human assets 31

a. Indigenous skills and knowledge 31

Background notes 31

Indigenous skills and knowledge in Mieso-Mulu woreda 31

b. Basic serviceseducation 31

Background notes 31

Education in Mieso-Mulu woreda 32

Contents


6/100Mercy Corps and Feinstein International Center2

c. Basic health services 35

Background notes 35

Health services in Mieso-Mulu woreda and Shinile Zone 35

d. Maternal health care and child vaccination 38 3.1.4 Social assets 40

Background notes 40

Social capital in Mieso-Mulu woreda 41

3.1.5 Physical assets 42

Background notes 42

Physical assets in Mieso-Mulu woreda 43

3.1.6 Political assets 43

Background notes 43

Political representation in Mieso-Mulu woreda 44

3.2 Vulnerability context 44

3.2.1 Seasonality 44

a. Seasonality o production and market patterns 44

b. Seasonal variations in nutrition and the impact o drought and confict 46

3.2.2 Shocks 46

a. Confict 47 Background notes 47

Conict and livestock raiding in and around Mieso-Mulu woreda 47

b. Reduced access to ormal and inormal trade and markets 51

Background notes 51

Market and trade issues in Mieso-Mulu woreda and beyond 51

c. Livestock diseases 52

3.2.3 Trends 53

a. Human population growth 53

b. Confict histories and trends 54

c. Bush encroachment and rangeland degradation 55

d. Enclosure o land or crop production 55

e. Increasing impact o normal dry seasons and drought 55

. Terms o trade 59

3.3 Policies, Institutions, and Processes 61

3.3.1 Regional policies and legislation 62 3.3.2 Pastoralism, policies, and institutions in Ethiopia 64

Contents (continued)


7/100



Abbreviations

ACTESA Alliance or Commodity Trade in Eastern and Southern Arica

BCG Bacille Calmette-Gurin tuberculosis vaccine

CAHW Community-based Animal Health Worker

CCM Community Case Management

CEWARN Confict Early Warning and Response Mechanism

CHA Community Health Agent

CJTF Combined Joint Task Force

COMESA Common Market or Eastern and Southern Arica

CSA Central Statistics Agency

CSOs Civil Society Organisations

DFID (UK) Department or International Development

DPPA Disaster Preparedness and Prevention Agency

DPT Diphther ia, Pertusis, and Tetanus

DRMFSS Disaster Risk Management and Food Security Sector (o the MoARD)

EB Ethiopian Birr

EHNRI Ethiopian Health and Nutrition Research Institute

ELTAP Ethiopian Land Tenure Policy and Administration Program

EPRDF Ethiopian Peoples Revolutionary Democratic Front

FAO Food and Agriculture Organization

FEWSNET Famine Early Warning Systems Network

FMOH Federal Ministry o Health

GAM Global Acute Malnutrition

GCC Gul Cooperation Council

GDP Gross Domestic Product

HEW Health Extension Worker

HSDP III Health Sector Development Plan

HSEP Health Sector Extension Programme

IGAD Intergovernmental Authority on Development

Ksh Kenyan Shilling

MDG Millennium Development Goal

MMR Maternal Mortality Rate

MoARD Ministry o Agriculture and Rural Development

MoE Ministry o Education

MoFA Ministry o Federal Aairs

MoFED Ministry o Finance and Economic Development

NGOs Non-governmental OrganisationsONLF Ogaden National Liberation Front

ORT Oral Rehydration Therapy


9/100Moving Up or Moving Out? A Rapid Livelihoods and Confict Analysis in Mieso-Mulu Woreda, Shinile Zone, Somali Region, Ethiopia 5

PACAPS Pastoral Areas Coordination, Analysis and Policy Support (Program)

PASDEP Plan or Accelerated and Sustained Development to End Poverty

PIPs Policies, Institutions, and Processes

PRA Participatory Rural Appraisal

RFE Rainall Estimation

SCFUK Save the Children Fund UK

SCUK Save the Children UK

SCUS Save the Children US

TBAs Traditional Birth Attendants

TFG Transitional Federal Government

UNICEF United Nations Childrens Fund

UNOCHA United Nations Oce or the Coordination o Humanitarian Aairs

USAID United States Agency or International Development

WHO World Health Organization

WSLF Western Somalia Liberation Front



This report describes a rapid, combined

livelihoods and confict analysis in Shinile Zone,

Somali Region o Ethiopia, conducted in March

and April 2010. An underlying question or the

analysis was the extent to which aid actors should

integrate peace-building and livelihoods

programming as part o long-term development

strategies or the Zone. A strategic ramework

with explanation is presented in section 4 o the

report and includes the ollowing issues:

TheIssapastoralistsystemextendsbeyond

Shinile Zone, so that changes to the system have

impacts in Aar and Oromiya Regions. The

system is determined by the need or pastoralists

to access natural resources, and it cuts acrossadministrative borders, ocial or otherwise.

Essentially, this is the system in the Rit Valley

between the Hararghe highlands and the eastern

escarpment. A harmonized cross-regional state

approach, involving Somali, Aar, and Oromiya

Regions, is needed to deal with both resource-

based conficts and conficts arising rom regional

structural/border disputes.

Peace-buildingandeconomicintegrationacross

regions are mutually supportive approaches,and are the core part o the ramework; a

10- to 15- year time rame is proposed.

Rationalizationoflandtenureandlanduse

policies is central to both economic integration

and peace-building. Policy and legal support to

pastoralist communal land use, with protection

o pastoral lands rom appropriation by other

users, is a crucial complementary process to

capacity-building support in peace-building to

government and community actors.

Moresupportivelandtenurearrangementswill

contribute to livestock development and

marketing, and so mainly assist those people

who stay in the pastoralist system. Currently

this includes households which are not only

staying, but are probably slowly moving up

economically as they maintain or increase their

herds and supply livestock markets.

Commercialization trends are clearly evident in

other Somali pastoralist areas and seem likely to

take hold in Shinile Zone. Educationandhealtharefundamentalto

economic and social development. While there

may be increasing government commitment to

improving health and education in Somali

Region, the service delivery strategies will need

urther adaptation i basic health and education

indicators are to improve. Both sectors seem to

lack specic strategies or dealing with the

cultural barriers to accessing services aced by

women and girls. Education is particularly

important as a means to support economic

diversication and integration, especially or

women and youth, but also or other people

who may opt to move out o the pastoralist

sector.

Servicedeliverystrategiesneedtoexamine

possible cross-regional state border

arrangements that might help to overcomesome o the practical diculties o service

delivery through government alone. The

proximity o urban centres in Oromiya Region

to Shinile Zone and the position o Dire Dawa

indicate that basic service providers in parts o

the Zone could be supplied rom these areas.

To some extent, this is already happening. The

private sector is oten able to work more easily

across administrative borders than government

is, and ederal-level health and veterinarypolicies support private sector involvement in

service provision.

Long-termdevelopmentstrategiesneedto

anticipate natural disasters such as drought.

Government guidelines and policy support

livelihoods-based responses to drought such as

commercial destocking, targeted livestock eed

supplementation, and veterinary voucher

schemes. These approaches require private

sector involvement, indicating that economic

integration within the ramework covers both

development and relie activities, and can

include many o the same private sector and

community partners.

Although commissioned by Mercy Corps, the

study did not examine or evaluate Mercy Corps

programmes. Mercy Corps own strategies could

draw on this report, but also might use

independent evaluation o their current

programmes, analysis o core organisationalstrengths and technical experiences, and dialogue

with donors to assess unding opportunities.

SUMMARY



1.1 Livelihoods and conict in

pastoralist areas

The Somali Region o Ethiopia is characterised

by many o the same development issues aecting

other mainly pastoralist areas o Ethiopia, and the

Horn o Arica more widely. A very longstanding

and core problem has been confict, which, in

part, relates to the limited and variable natural

resources in pastoralist areas and competition or

resources between groups. However, confict also

has many other dimensions in these areas due to

actors such as changing national political

ideologies and structures, limited government

capacities in more remote areas, and the requent

cross-border aspects o confict. During the last15 years or so, humanitarian and development aid

organisations and donors have supported an

increasing array o peace-building, confict-

resolution, or similar programmes in pastoralist

areas. When implemented by non-governmental

organisations (NGOs), these programmes oten

ocus on confict resolution activities at the levels

o local government and communities, including

eorts to build local capacities to prevent confict.

A common nding rom these programmes acrossdierent countries is that, while progress is oten

evident during implementation, local reductions

in violent confict are later undermined by

higher-level actors and processes. Oten running

parallel to these programmes, and sometimes

implemented by the same agencies with the same

unding sources, are livelihoods programmes.

These vary considerably in their approach in

pastoralist areas but, in one way or another, oten

ocus on livestock and related areas such as water

development, livestock marketing, or natural

resource management. Variations in programming

partly relate to diverse interpretations o the word

livelihood and equally varied analysis around

how livelihoods can be protected or strengthened.

In aid circles, livelihoods thinking is associated

with the emergence o livelihoods analysis in the

late 1990s and an analytical approach called the

sustainable livelihoods ramework. When applied

in confict-aected pastoralist areas, use o thelivelihoods ramework quickly highlights the

major impact o confict on livelihoods and how,

1. INTRODUCTION

in certain circumstances, peoples responses to

confict lead to behaviours or activities which

perpetuate confict. The ramework also reveals

the actors at multiple levelsrom community to

international processes and vice versathat

contribute to confict. With these issues in mind,

livelihoods analysis in confict-aected areas

should automatically include confict analysis,

leading to programming strategies in which

livelihoods work and peace-building work are not

separate entities but combined, mutually-

reinorcing approaches.

1.2 Background to the analysis in

Mieso-Mulu woreda, Shinile Zone

Shinile Zone in Somali Region has been

characterised by confict or many years, dating

back to the period o imperial rule in Ethiopia

and beore. Historically, the main Somali clans

gained and occupied territory by orce, and the

Issa clan in Shinile Zone is no dierent. The

boundaries o its current position in Shinile are in

part an outcome o the clan receiving arms rom

the Italian army in the early 1900s, which allowed

them to expand their territory rom around theEthiopia-Djibouti railway into the Allighedi plain

and towards the Awash River. While intra-clan

disputes have occurred, as they do in other

Somali pastoralist areas, the Issa have also been

involved in longstanding confict with the Aar to

the west and with Oromo to the south. Both

conficts were relatively localized and resource-

based, with dierent groups aiming to secure

access to water and grazing or, in the case o the

Oromo, land or cultivation.

Following the introduction o regional autonomy

in Ethiopia in the early 1990s, additional tensions

were superimposed over the local competition or

land, grazing, and water resources. These tensions

arose rom delineation o the border between

Oromiya and Somali Regions, and Aar and

Somali Regions, and were also related to regional

governments making claims over key strategic or

economic locations. Violent clashes between

Somali Issa and Hawiya groups in Mieso-Muluworeda occurred as recently as February 2009.

They represent a distinct shit in confict


12/100



Corps based on a drat report, covering the

livelihoods analysis and strategic ramework.

These activities were designed with the intention

o using the livelihoods ramework to organise

and analyze existing data, within the time

available. A key problem here was that data

specic to Shinile Zone or Mieso-Mulu woredawas limited, and requently, the analysis assumed

that inormation rom other parts o Somali

Region was relevant to Shinile Zone or that

otherworedas in the Zone were representative o

Mieso-Mulu woreda. Some o the main reports

used in the analysis were

alivelihoodsstudyinSomaliRegionin2006,

which included pastoralist areas o Shinile

woreda in Shinile Zone (Devereux, 2006);

ahealthandnutritionsurveyconductedin

April to May 2009, which included Ayisha

woreda in Shinile Zone (EHNRI/UNICEF/

SCUS, 2009);

twostudiesonconictinandaroundMieso-

Mulu woreda, both o which cover structural

and governance aspects o confict since the

early 1990s (Ahmed Shide, 2005; Fekadu

Beyene, 2009);

ananalysisoftheAfar-Issaconict(Markakis,

2003).

The study ocused on Mieso-Mulu woreda due to

time constraints and the limited availability o

local government ocials due to the orthcoming

elections in Ethiopia (in May 2010). Focus group

discussions were used to investigate issues which

were not well described in existing reports and to

cross-check the relevance o zonal or other data

to Mieso-Mulu woreda.

The study ocused on pastoralist areas o Mieso-Mulu woreda, while recognising the importance o

relationships with neighbouring agropastoral

Oromo groups in the woreda, and on the strong

linkages between Issa communities in Mieso-

Mulu woreda and those in other parts o the Zone,

and in Djibouti and Somalia. Ecologically,

conditions in the pastoralist areas o Mieso-Mulu

woreda were similar to those in other parts o

Shinile Zone. Thereore, to some extent, certain

parts o the analysis in Mieso-Mulu woreda can be

applied more broadly to the Zone.



risky, asset loss due to confict and within certain

policy or institutional contexts can also be a cause

o confict e.g., by encouraging illegal activities to

acquire assets. Thereore livelihoods analysis

covers both the actors which can cause confict

and how coping strategies can arise that

perpetuate confict in the absence o otherlivelihoods options.

2.1 The livelihoods ramework

The livelihoods ramework enables a description

o local individual, household, or community

assets to be positioned and analyzed against

actors which contribute to vulnerability, such as

seasonality, shocks, and trends. The ramework

also allows examination o ormal and inormal

policies, institutions, and processes that aect the

ways in which people are able to protect or

develop their assets. This part o the ramework

includes sub-national, national, regional, and

international policies and institutions. From the

perspective o dening and responding to

poverty, pastoralist communities commonly

explain poverty by reerence to both their

livestock holdings (nancial assets) and access to

indigenous social networks and support systems

(social assets). The sustainable livelihoods

ramework allows both nancial and social assets

to be examined.

In a confict-aected area, livelihoods analysis

should automatically include analysis o confict

issues. Typically, confict cuts across three core

elements o the livelihoods ramework, viz. assets,

vulnerability context, and policies, institutions,

and processes. While confict might reduce assets

or make certain strategies or managing assets very

2. USING THE LIVELIHOODS FRAMEWORK TO INTEGRATE LIVELIHOODS AND

CONFLICT ANALYSIS

The sustainable livelihoods ramework

A livelihood comprises the capabilities, assets (including both material and social resources) and activities

required or a means o living. A livelihood is sustainable when it can cope with and recover rom stresses

and shocks, maintain or enhance its capabilities and assets, while not undermining the natural resource

base. (Scoones, 1998)



Table 1. The impact o conict on livelihoods assets in pastoralist areas o Arica

Livelihoods Impacts o conict

asset

Human Themainvictimsofcivilandinternationalconictsarecivilians,notarmed

combatantsphysical injury, mental trauma, death

Localizedlivestockraidingpossiblyresultinginviolentinjurytolivestock

keepers and their amilies, the raiding groups, or local police and security orces Womenandchildrenareparticularlybadlyaectedbymanyformsof

confictrape, mutilation, and orced marriage o women and girls are tactics

o war and counter-insurgency; spread o HIV

Useofchildsoldierswithrelatedlong-term,mentaltrauma,andwidersocietal

impacts

Scorchedearthtacticswithviolentremovalofcommunitiesfromresource-

rich areas e.g., oil, natural gas, minerals

Destructionofhealthfacilitiesanddisruptionofbasicpreventivehealth

campaigns e.g., child vaccination

Destructionofeducationfacilitiesordisruptedaccesstoeducation

Destruction/damagetowatersupplies;water-bornediseaseoutbreaks

Injuryanddeathafterconictduetolandminesandunexplodedordnance

Financial Directandviolentdepletionofnancialassetssuchaslivestockisatacticof

war and counter-insurgency

Restrictionsonmovementseasonallabourmigrationandremittances

Marketclosureordysfunction,preventingsaleorexchangeoflivestockfor

cash or grain

Breakdownofveterinaryservicesnopreventiveorcurativeservices,shocks

due to disease epidemics and loss o assets

Limitedprivatesectorinvestmenthighrisksofdoingbusinessandtrade

Wareconomieswithtradecontrolledbyarmedelitesforpersonalgainand

related incentives or maintaining confict

Physical Destructionordamagetoroadsandphysicalinfrastructure

Destructionofgovernmentocesandrecords

Breakdownofcommunicationandtransport

Destructionorcontaminationofwatersourcesasatacticorwarandcounter-

insurgency

Breakdownofpublicsecurity

Natural Restrictedmovementlimitsaccesstograzingareas;overgrazingofaccessible

areas; restricted cross-border movements Baddiversicatione.g.,excessivecharcoalproduction

Nomanslandareasbetweenconictinggroups

Landminesandunexplodedordnancepreventingaccesstograzingareas

Breakdownoftraditionalinstitutionsfornaturalresourcemanagement

Social Forcedmigrationinternalandinternationaldisplacement;breakupoffamilies

and communities

Breakdownoftraditionalsafetynets/socialsupport

Breakdownoftraditionalleadershipandinstitutions

Newandviolentsocialnorms

Political Reducedpoliticalcapitalaspastoralistsareportrayedasinherentlyviolentor

supporting insurgents, opposition groups, or religious extremists

Source: adapted rom COMESA, 2009.



Confict analysis is the systematic study o a

confict in given area and is structured around an

examination o the prole, causes, actors, and

dynamics o confict (Saerworld, 2008). Much o

the content o a typical confict analysis also alls

into the main components o livelihoods analysis,

as summarized in Box 1 below.

Conict analysis1

ProleA confict prole provides a brie

characterisation o the context o a confict,

looking at political, economic, and socio-

cultural contexts, issues that emerge rom

these, and the history o the confict.

CausesCauses o confict may be

structural, pervasive actors built into

the abric o a society; or instance, i there

is unequal access to natural resources or a

discriminatory system is in place;

proximate actors which contribute

towards a climate o violence; or instance,

a prolieration o illicit small arms;

triggers, being single events that may set

o or escalate violence; or example,

elections, coups, or sudden currency

collapses.

ActorsThinking about people is central to

confict analysis. Actors reers to all those

individuals, groups, and institutions

contributing to or aected by a confict.

DynamicsThe interaction between a

conficts prole, its causes, and actors can be

described as that conficts dynamics, how the

confict changes and develops over time.

Understanding a conficts dynamics will help

identiy windows o opportunity or

peace-building and can help organisations

plan uture work.

Box 1. Linkages between conict analysis and livelihoods analysis in

pastoralist areas

1 From Saerworld, 2008.

Livelihoods analysis

Conict proles are covered under

vulnerability context and PIPs, which would

also cover non-confict actors and issues.

Confict histories t naturally under trends in

vulnerability context. Elements o livelihoods

assets, such as natural, nancial, and social

assets can also be very relevant to confict

analysis.

Structural actors all mainly under PIPs,

which include both ormal and inormal

institutional issues and processes.

Proximate actors all under PIPs and/or

trends (vulnerability context).

Triggers all under shocks in vulnerability

context.

Actors are covered by livelihoods analysis

under PIPs and, or community-level groups,

also under social and political capital.

Dynamics are covered by PIPs and trends in

livelihoods analysis.



2.2 Policies, institutions, and processes:

the PIPs analysis and conict

In terms o household or individual assets, one

reason confict is so important in pastoralist areas

is that it can impact negatively, and oten

prooundly, on all types o assets (Table 1).

Understanding these impacts is central to

livelihoods analysis in these areas. Also,

pastoralism is oten noted or its vulnerability, and

a mix o predictable and unpredictable actors

threaten assets. These are outlined in Table 2.

Table 2. Conict and vulnerability contexts in pastoralist areas

Seasonality

Shocks

Slow-onset

disastersdrought

Trends

Markedseasonalityoflivelihoodsinanormalyearduetorainfallpatternsand

seasonal variation in ood production, ood access, and market conditions

Seasonalvariationingrazingandwaterresources;seasonalpeaksinresource

competitionrisk o conict fashpoints

Sudden onset and unpredictable events such as:

LivestockdiseaseoutbreaksHumandiseaseoutbreaks

Floods

Marketbans

Conict and raiding

Borderclosures

Foodpriceincreases

Drought usually involves ailures or marked reductions in two or more

successive rains, and can be expected every ve to seven year or sothereoredrought is predictable and slow- onset.

Drought is also characterised by

highlossoflivestockassets,withlongrecoverytimesforrebuildingherds,

especially cattle and camels;

humandiseaseoutbreaksaspeoplecongregatearoundwatersourcesorrelief

centres;

conict as herders compete or dwindling resources.

Long-term changes including:

Protracted conicts and political instabilitycomplex emergencies

Humanpopulationgrowth

Increasingnegativeimpactofdrought

Decliningpercapitalivestockholdingsamongpoorerhouseholds

Developmentdisplacement

Bushencroachment

Inappropriatewaterdevelopment

Erosionoftraditionalinstitutions

Accesstomodernsmallarms

Commercializationoflivestockrearing



Looking more closely at confict and

vulnerability, many NGO peace-building

programmes work locally and gain an

understanding o tensions between, or example,

conficting ethnic groups and the role o actors

such as local politicians or government. However,

attention to policies and institutions in livelihoods

analysis places local issues in a broader contextwhich can include sub-national, national, regional

(i.e., groups o neighbouring countries), and

international policies and processes. This wider

analysis can be particularly useul in pastoralist

areas, where confict oten has important cross-

border or regional dimensions that in turn can

point to the need or regional responses to

confict. For NGOs and donors in the Horn o

Arica, this regional aspect o confict oten raises

important challenges because, in general, these

organisations are structured and managed

nationally, not regionally. It ollows, or example,

that i confict-sensitive programming includes

balanced inputs to communities across borders,

such programming can be dicult or aid

organisations to design and implement.

Policy and institutional analysis in the livelihoods

ramework also covers inormal policies and

norms, which can reveal disparities between

ormal policies or laws and their interpretation orenactment among dierent actors. This type o

analysis can show how some policies are well

thought out and articulated on paper, but not

implemented. In contrast, bad policies can be

ignored locally, with communities and local

private sector developing systems which t a local

context. Institutions include inormal institutions

such as the traditional social units and leaderships

o pastoralist communities, but also socio-cultural

norms and practices. While peace-making is oten

an important unction o traditional leadership, so

is the mobilization o youth and organisation olivestock raids. Thereore, at one level, livelihoods

analysis can show that people can suer the

impacts o confict while also being a cause o

confict. Among policy makers in government,

pastoralist areas are oten described as problematic

due the apparently violent tendencies o

pastoralists and requent clashes over, or example,

raids or natural resources. At another level, the

same analysis might argue that i one o the main

unctions o a state is to protect its citizens, then

ultimately the responsibility or resolving conficts

and maintaining civil order lies with government.

In terms o vulnerability, ormal and inormal

policies and institutions have a major infuence on

the extent to which seasonality, shocks, and

trends result in vulnerability. Some examples are

provided in Box 2. In summary, confict can be

instigated by pastoralists and prolonged by

strategies to control resources by sel-arming and

use o physical orce in preerence to negotiation.However, decits in ormal government policies,

laws, and capacities create the space or violence-

based strategies and behaviour to take place.

Box 2. Formal and inormal polices and institutions, and pastoral vulnerability

SeasonalitySeasonal, resource-based conficts refect inadequate ormal institutions to

dene and enorce user rights and, in these situations, the tendency or one group or anotherto claim control o resources by physical orce. Thereore ormal institutional weaknesses

increase the risk o resource-based conficts. Government policies and laws may deliberately

marginalise pastoralists in avour o settled armers, or can be well-meaning but evolve rom

limited understanding o pastoral livelihoods. Traditional peace-making mechanisms tend to be

stronger within ethnic groups rather than between groups, and tend to react to confict rather

than prevent confict.

ShocksLivestock raids are a type o shock. For those involved in raiding, risks such as

imprisonment or other punishments are lessened in situations where local police orces are

weak or where commercial or contract raiding is led by actors who can bypass the police or

judiciary. Again, weaknesses in ormal institutions create environments that support raiding. At



During the last 20 years or so, considerable eort

has been expended in resolving and managing

confict in the Horn o Arica, through diplomatic

means, government interventions such as orced

disarmament, and through various confict earlywarning systems and local confict resolution

programmes supported by NGOs and other

actors. Analysis o these interventions points to at

least our important challenges:

Thepersistenceofinternalconictisessentially

a governance issue and relates very much to

government systems and capacities both

centrally and in pastoralist areas.

Inappropriatedevelopmentpoliciesand

strategies can uel conficts e.g., through theconstruction o new acilities such as water

points in confict-sensitive locations or the

unlawul allocation o pastoralist land or

agriculture.

Simplyignoringpastoralistareascanleadto

weak or harmul livelihoods diversication,

including activities which may hurt the physical

environment and social abric o society; when

people respond to limited livelihoods options

by resorting to illegal or environmentally-

damaging activities, this can cause or perpetuate

local conficts.

Thepoliticaleconomyofcertainformsof

the same time, raids are well-planned events which can be condoned or even actively led by

the same traditional leaders who also have peace-making responsibilities. Traditional

leadership, like ormal political leadership, is variable, with some individuals using their

positions or individual economic gain. Thereore there can be important weaknesses in

traditional institutions as well as in governments.

Slow-onset disastersThe well-established, ormal institutional response to drought is oodaid. Procedures within governments and aid organisations ocus on assessing ood aid needs

and delivering ood aid. Drought cycle management and related livelihoods-based

programming, i.e., risk-based drought management, is poorly institutionalized. One result is

excessive loss o core livestock assets by pastoralists, with long recovery times and use o thet/

raiding to rebuild herds.

TrendsVarious trends combine to increase the risk o both seasonal conficts and shocks,

and the impact o drought. Trends such as bush encroachment lead to reduced availability o

useul vegetation or livestock grazing and thereore increase pressure on already-scarce

resources. Similarly, the continuing appropriation o pastoralist lands or commercial arming

oten excludes pastoralists rom key dry season grazing and water resources, thereby increasingthe need to access these resources elsewhere. Thereore ormal land administration and land

use planning processes are central to long-term development planning and indirectly to

confict prevention.

confict indicates that confict can be initiated

and maintained by individuals or personal gain,

and that the same individuals may undermine

attempts to resolve confict in the long term. I

a core unction o government is to ensure thesaety and protection o its citizens, in many

pastoralist areas this unction is not achieved.

This brie list o confict-related policy and

institutional constraints indicates the links

between development policies and strategies, and

confict. Land, environment, and service delivery

policies all impact on livelihoods options and,

when not well-conceived, can directly or

indirectly encourage violent confict. Inpastoralist areas, policy and institutional barriers

also oten occur at dierent levels, rom local to

international. These can include policies on

cross-border movements and trade, policies

around nancial systems which enable (or not)

remittances, and so on. Confict is also oten

multilayered and multidimensional, and includes a

complex and dynamic set o actors and

relationships. Levels o confict vary rom

international or inter-state conficts, internal

political conficts (sometimes linked to inter-state

proxy wars and arming or support to

insurgents), and local ethnic conficts. At times,



and especially in Somali pastoralist areas, dierent

types o confict are interlinked. As indicated

above, an understanding o the political economy

o confict is important, including the economic

incentives or some actors to sustain confict over

time.

2.3 Conict and livelihoods programming

In terms o joined-up confict and livelihoods

programming, confict work is expected to lead

to peace and thereore improved livelihoods by

reducing the negative impact o confict on

various assets, e.g., see Table 1. Similarly,

livelihoods programmes may include confict-

related objectives. For example, a programme

might aim to develop land policy and legislation

that supports pastoral mobility and claries user

rights. Indirectly, this could result in reduced

confict over natural resources. In general,

confict-sensitive livelihoods programmes include

particular attention to the appropriate distribution

o inputs between conficting groups so that

greater acquisition o assets or services by one

group does not then become a cause o confict.

Within this general principle is the notion o

economic inter-dependence and the mutual

benets arising rom cooperation around activities

such as trade or the shared use o certainresources. While intuitively this type o

programming logic makes sense, sustained peace

and related benets probably depend on the root

causes o confict. For example, these causes can

be entirely localized and arise rom reactive,

violent behaviours. In this case, acilitation o

peace-making with traditional institutions, and

with complementary livelihoods support across

groups, may lead to sustained benets.

In reality, however, many confict-prone

pastoralist areas are characterised by diverse causes

o confict, with weak governance rom central to

local levels being a common actor. In part, this

explains the ragility o aid programmes, which

work only locally on confict and livelihoods

because local gains are easily undermined,

consciously or unconsciously, by higher-level

actors, disabling ormal policies, or conusion over

ocial boundaries.



Mieso-Mulu woreda lies in the southwest o

Shinile Zone in the Somali Region o Ethiopia

and is bordered by Aar Region to the west and

Oromiya Region to the south. Shinile Zone

borders Djibouti to the north and Somalia to the

east. The Zone is occupied mainly by Somalis o

the Issa clan, who also predominate in Djibouti

(including the Djibouti government) and parts o

northwest Somalia. The physical location o the

Zone and distribution o the Issa immediately

points to the cross-border nature o livelihoods in

the area. Shinile Zone and Mieso-Mulu woreda

are also occupied by smaller numbers o Somali

Hawiyas and Oromos, who are mainly

agropastoralists ound in the relatively higher and

wetter areas along the southern border o theZone, at the base o the Hararghe highlands.

The geography o the area explains the pastoralist

3. LIVELIHOODS ANALYSIS IN MIESO-MULU WOREDA

Figure 1. Shinile Zone in the Somali Region, Ethiopia

economy o the Issa. Approximately 76 percent o

the rural population in Shinile Zone is pastoralists,

and around 80 percent o the rural population in

Mieso-Mulu woreda (Table 3). Lying at the foor

o the Rit Valley, between the Ethiopian highland

escarpment running south to north and the

Hararge highlands running west to east, Shinile

Zone becomes progressively lower, hotter, and

drier, especially near the border with Djibouti.

Annual rainall in the Zone is 557 mm, but this

varies rom relatively higher rainall woredas o

Mieso-Mulu (678 mm/year) and Dembel (670

mm/year) to the lower-lying, more northern

woreda o Ayisha (404 mm/year) (see Figure 4).

Temperatures in the Zone range rom around 27

to 43C. In the total zonal area o approximately89,200 km2, there are no permanent rivers, but

there are three seasonal rivers running roughly

south to north, down rom the Hararghe hills.

Notes: adapted rom the unocial maps produced by UNOCHA. The heavy blue line delineatesadministrative zones. Shinile is divided into six woredas (distr icts), viz. Mieso-Mulu, Adem, Erer, Shinile,

Ayisha, and Dembel. Mieso town is situated just outside Shinile Zone (Somali Region), to the

southwest o Mieso-Mulu woreda.



Woreda Rural population Pastoralist population

(% o rural population)

Shinile 96,988 77,591 (80%)

Dembel 77,321 30,928 (40%)

Ayisha 50,043 50,043 (100%)

Erer 83,471 66,777 (80%)

Adem 31,991 31,991 (100%)

Mieso-Mulu 45,570 36,456 (80%)

Total 385,384 293,786 (76%)

Figure 2. Physical geography o Shinile Zone

Notes: the Shinile Zone border is approximate and unocial. The gray-blue lines depict major roads.

Mieso-Mulu woreda is situated in the southwest o the Zonesee Figure 1. Some inormants described

an extension o the northwest corner o the Zone westward into Aar Region and covering part o themain road to Djibouti, but this was an unocial expansion o the Zone at the time o writing.

Source: DPPA/SCUK, 2008.

Table 3. Human population estimates or Shinile Zone



Despite its apparent remoteness and harsh

environment, the physical position o Shinile

Zone, adjacent to both Djibouti and Somalia,

indicates its strategic importance or Ethiopia as a

whole. For example, national economies depend

on a countrys transport system. Ethiopia is a

land-locked country, and the Djibouti port is the

major route or imported goods and an importantchannel or Ethiopian exports. The main road

rom Addis Ababa to Djibouti runs around the

western border o the Zone, and, in April 2010,

passed through an unocial extension o the

Zone occupied by Issa communities. At the same

time, the Ethiopian government was planning to

asphalt the road running through Dewele in the

northeast o the Zone, making that road an

alternative route to and rom Djibouti. The

northern Somali ports o Berbera and Bossaso are

also important export routes, especially or

livestock exported unocially rom Somali

Region.

Both Djibouti and northern Somalia (Somaliland

and Puntland) are part o the wider and regional

political context, including Ethiopia-Eritrea

relations, Ethiopia-Somalia relations (both

northern and southern Somalia), and, internally,

Ethiopian counter-insurgency strategies against

the Ogaden National Liberation Front in theSomali Region. The regional politics are complex

and dynamic, both currently and historically. As

previously mentioned, the Issa are the dominant

ethnic group in Djibouti, including in the

Djibouti government. As they also occupy most

o Shinile Zone, they represent an important

cross-border socio-political grouping, which

directly controls the Djibouti port while also

having close proximity to the main Addis Ababa-

to-Djibouti road and railway. Similarly, the Aarare a cross-border group, present in Djibouti,

Eritrea, and Ethiopia.

Although the northern part o Shinile can be

considered remote and inaccessible, more

southern areas are relatively accessible i compared

to other parts o Somali Region such as Warder

or Gode. Two main roads run along the western

and southern boundaries o the Zone, and the

city o Dire Dawa lies around the midpoint o the

southern border, just outside the Zone. The cityo Harar is also relatively accessible. Within the

Zone, Mieso-Mulu woreda is relatively small

(about 60 km rom east to west and 70 km rom

south to north), especially i compared to the

huge tracts o land that characterise many

pastoralist areas o Ethiopia and other parts o the

Horn o Arica. The woreda can be reached in

about our hours by road rom Addis Ababa.

3.1 Assets

3.1.1 Financial assets

Financial assets comprise the inows o cash rom

income, gits, or other sources, as well as stocks and

savings held by a amily or household.

Pastoralists commonly dene wealth in terms o

livestock holdings and elements o social capital,

such as access to amily members or loans, gits,

or other types o support. In any given pastoralist

area, there are households which own relatively

large numbers o livestock, measured in absolute

terms in hundreds or even thousands o animals.

Even households that are characterised locally as

poor may own up to 30 or 40 small ruminants,

and a ew cattle or camels. These minimum

numbers o livestock are needed to enable a

pastoralist way o lie and are sometimes measured

under the concept o minimum herd size. In

terms o development strategies, understandingwealth groups is important because marketing

behaviour varies by wealth group, with poorer

households strategizing to build herds, and

middle-wealth and wealthier households selling

more animals and being the main suppliers o

livestock to markets.

a. Livestock herds and household income

Background notes

For pastoralist communities, the main nancial

asset is livestock. Livestock holdings represent

wealth and animals are used both as a orm o

savings and as assets to be exchanged or cash or

grain as needed. The emphasis on livestock

ownership and production in pastoralist areas is

largely determined by environmental actors and

especially the arid or semi-arid conditions with

marked variability in rainall.

In general, pastoralists keep mixed herds olivestock comprising combinations o large and

small ruminants. The composition o herds in



dierent areas is infuenced by environment and

particularly the drought tolerance o dierent

livestock species. Market demands also infuence

the types o livestock reared, especially by

wealthier households. As a general rule, camels

and goats are prioritized in the most arid areas,

with cattle and sheep also reared, but in areas

with relatively higher rainall. It ollows thatmixed herding is a logical strategy that allows risk

management and fexible nancial management.

In terms o risk, dierent livestock species have

dierent water and grazing needs, and to some

extent, are aected by dierent diseases.

Thereore it is likely that some animals will

survive an adverse event such as a disease

epidemic or drought. In terms o nancial

management, small stock such as sheep and goats

are a convenient asset to be sold to meet basic

needs such as ood, medicine, or school ees,

whereas larger stock represent more long-term

savings.

From the perspective o relatively intensive or

purely commercial livestock production systems,

the types o livestock kept by pastoralists are

sometimes viewed as low producers in terms o

indicators such as milk production or

reproductive perormance. However, production

should be viewed relative to the required outputs

o the system and the cost o inputs. Livestock

breeds in pastoralist areas have evolved over many

hundreds o years and are generally well adapted

to the local environment and disease risks. In part,

this adaptation has been infuenced by selective

breeding by pastoralists to emphasize specic

production characteristics such as milkproduction. Furthermore, seasonal variations in

rainall and grazing, and recurrent drought, mean

that pastoralist livestock such as camels are able to

produce milk in very dry conditions when other

species have ceased production.

Income rom livestock and other sources in Mieso-Mulu

woreda

Income rom livestock sales, renting o pack

camels, and sale o ghee were the main types o

income reported in the pastoralist areas o Shinile

Zone or the year 1998 to 1999 (Table 4). Due to

the contraband trade rom Djibouti through the

Zone, the renting o pack camels was an

important source o income, representing 32

percent, 31 percent, and 48 percent o income or

better-o, middle, and poor households

respectively. No sources o income were reported

or very poor households.

Renting o pack camels is an important source o income or pastoralists in Shinile Zone.



Other sources o income or pastoralist

households were also associated with the close

proximity o the Zone to Djibouti and Dire

Dawa, and ties with Issa relatives there. For

example, remittances rom amily members who

had been sent to work in Djibouti, Dire Dawa, or

other urban centres were reported at less than 10

percent o total income or poor and middlewealth groups, and between 10 to 20 percent or

better-o households.

Table 4. Contribution o livestock to annual household income in pastoralist areas,

Shinile Zone

Wealth group Total annual income (EB) rom livestock/total income (%)

1998-1999 2004-2005

Better-o 6050/7000 (86%) 9300/9300 (100%)

Medium 3978/4350 (91%) 5800/6200 (94%)

Poor 1980/2200 (90%) 2200/3950 (56%)

While the inormation above might apply to

Mieso-Mulu woreda, reports by Mercy Corps and

direct observation indicates the importance o

other income generation activities such as sales o

milk, charcoal, and rewood by those who can

access Mieso town. At the time o writing this

report, such access was limited to Hawiya

communities due to confict between Issa andHawiya; this is described in more detail in section

3.1.1c.

Notes: the data were collected rom dierent pastoralist areas o Shinile Zone and thereore are

not specic to Mieso-Mulu woreda. Sources o income were livestock sales, sale o ghee, and

renting out o pack camels.

Source: DPPA/SCUK, 2002, 2008.

For the year 2004 to 2005, the DPPA/SCUK

report noted a decline in income rom renting o

pack camels due to heightened government

restrictions on the contraband trade. Whereas this

income source accounted or between 31 and 48

percent o annual income in 1998 to 1999

(depending on wealth group), rom 2004 to 2005

it was reported to be absent. During the eld

visits during the preparation o this report inMarch 2010, the use o camels as pack animals or

movement o contraband goods was reported to

be extensive and rental prices were increasing.

Relative to 1998 to 1999, all households earned

higher income rom livestock sales as a result o rising

livestock prices (DPPA/SCUK, 2008).

The sources o income described above indicate

that, to some extent, Issa communities in the

woreda have already diversied their income. Incommon with other pastoralist areas, these

activities can be categorised as good and bad

diversication as ollows (Little, 2009):

Gooddiversicationiscloselylinkedtothe

pastoralist sector and keeps value added in the

region; it includes milk and meat processing,

tanning, trading, retail input suppliers, and

local natural product gathering/processing. For

Mieso-Mulu woreda, good or at least neutral

diversication might include remittances.

Ocially, the use o pack camels to support

the contraband trade is probably seen as a ormo bad diversication by government,

especially i smuggling o arms takes place,

though or pastoralist themselves the practice

represents a logical economic activity with

high returns.

Weakorharmfuldiversicationmayhurt

the physical environment and social abric o

society and, in the long run, can undermine

the main economic activity o pastoralism. For

Mieso-Mulu woreda, it includes excessivecharcoal production and rewood sales, and

livestock raiding.



The strategy o sending amily members to towns

to seek employment would probably all under

harmul diversication in cases where people

were exploited or abused by employers, or

exposed to other risks. For example, girls sent to

work as domestic servants in Djibouti might be at

particular risk o physical or sexual abuse.

Related to economic diversication in pastoralistareas is education and a deliberate strategy o

educating amily members with a view to either

uture income via remittances or greater capacity

to engage in novel income-generating activities.

Education is discussed urther in sections 3.1.3b

and 3.3.2.

b. Livestock herds and human oods

Background notes

An important characteristic o pastoralist systems is

the direct, household-level consumption o

livestock products and especially milk. Animal milk

is a valuable ood containing high-quality protein

plus at, vitamins, and minerals. A basic nutritional

analysis o animal milk shows that two cups o milk

per day (around 500 ml) can meet at least 50

percent o the recommended intake or many

essential nutrients (Sadler et al., 2009). Thereore

the ability o livestock to convert dryland

vegetation into milk is one o the undamental

economic justications or pastoralism. Milk is not

only consumed resh, but preserved in the orm o

ghee, butter, yogurt, and cheese.

A second important characteristic o pastoralist

nutrition is the consumption o cereals and the

use o livestock to sell or exchange or cereals. All

pastoralists groups consume grain and thereore

require markets to exchange livestock or grain.

In terms o child nutrition and ood security, it is

important to note that grain alone is not easily

digested by young children. However, when

mixed with milk, cereal protein is more ully

utilized or growth.

The high reliance o pastoralist groups on milk

explains their good nutritional status during

periods o high milk production, such as the main

wet seasons. However, pastoralistsand especially

their childrenare very susceptible to periods o

milk decit, as occurs during long dry seasons or

drought. This aspect o pastoralist nutrition

explains seasonal and drought-related child

malnutrition.

Livestock herds and human oods in Mieso-Mulu

woreda

The DPPA/SCUK report or 2004 to 2005 or

Shinile Zone noted a decline in household size,

especially among middle-wealth and better-o

households; middle-wealth households declined

rom 9 to 11 to 6 to 8 members, whereas better-

o households declined rom 15 to 17 to 8 to 12

members. This represented a 30 to 40 percent

decrease in household size. This trend was

explained by migration to towns or abroad,

especially during bad years. One result is that it

became relatively easier to eed remaining

household members using livestock-derived

oods. This is urther evidence o livelihoods

diversication in the area.

Table 5. Contribution o livestock to human ood consumption in Shinile Zone, 1998 to 1999

Wealth Food typegroup

Cereals and sugar Milk rom Total ood derivedacquired through own livestock directly or indirectlylivestock sales rom livestock

1998-99 2004-5 1998-99 2004-5 1998-99 2004-5

Better-o ~73% ~58% ~25% ~42% ~98% ~100%

Medium ~65% ~60% ~28% ~38% ~93% ~98%

Poor ~65% ~70% ~15% ~12% ~80% ~82%

Notes: the data were collected rom dierent pastoralist areas o Shinile Zone and thereore are not

specic to MiesoMulu woreda. Decits lled by gits, relie ood, wild oods, etc.

Source: SCUK/DPPA, 2002, 2008.



A health and nutrition survey rom April to May

2009 noted that or Ayisha woreda: The main

sources o ood or adults in the our weeks prior to the

survey were rice (44%), sorghum (38%) and wheat

(16%). The main source o ood was cereal purchase

(69%), own animal production (22%) and relie ood

(12%). The source o income in the last three months

was mainly rom the sale o livestock (86%), ollowedby sale o poles/charcoal (4%) and sale o relie ood

(9.5%). Cash source or the month ahead was predicted

to be rom the sale o livestock (82%, n=120). Access

to markets is challenging in many parts o theworeda,

with modal average distance to main market (on oot,

return journey) ranges rom 4 to 8 hours. The majority

(81%) reported a round trip on oot o 8 hr. walk was

needed to make the return trip to the main market. A

total o 64.5% (n=71) mothers reported eeding animal

milk to their young child in the last 24 hours. It is

suggested that the livelihood o communities in Ayisha

is very dependent on the condition o their livestock or

trade and home consumption o their produce (milk in

particular) (EHNRI/UNICEF/SCUS, 2009).

c. Markets and trade

Background notes

Data on sources o ood and income in pastoralisthouseholds demonstrate that pastoralists are very

dependent on markets, especially as a means to

sell animals and acquire cereals. A vast body o

research shows that certain livestock marketing

behaviours are observed amongst all pastoral

wealth groups, but this depends on seasonal

conditions at the time o sale. In the wet seasons,

there is herd growth and prices generally are

good, and these conditions motivate sellers.

However, pastoralists must attain suciently largeherd sizes to allow them a comortable margin to

liquidate their animals through the market (e.g.,

Barrett et al., 2006). During dry seasons (or

drought) prices are lower, herders oten are

pressed by immediate cash needs and thus do not

have the luxury o timing sales according to

periods when prices are most avourable. Here,

the undamental principle is to market what is

considered surplus at a time when cash need

arises, and price is not such a key actor in

determining sales compared with wet season sales.Conversely, sales could be dictated by desperate

situations when the decimation o livestock

becomes imminent, as in times o prolonged

drought (Aklilu and Wekesa, 2001).

While certain behaviours hold true or all wealth

groups, important dierences exist in marketing

behaviours by wealth status. Research shows that

wealthier households use livestock markets more

requently to sell animals because these householdshave greater cash expenditures. For the poorer

households, the key livelihoods strategy is herd

growth because, in these environments, livestock

provide the best economic returns relative to other

available options.

From a cross-border perspective, many pastoralist

ecosystems are cross-border systems and this

includes marketing arrangements.

Markets and trade in Meiso-Mulu woreda and beyond

As Tables 4 and 5 indicate, pastoralists in Shinile

Zone are highly dependent on markets to sell

livestock to acquire cereals and other items. The 2002

report by DPPA/SCUK gives a good overview:

The main markets or the pastoral community in

Shinile Zone are: Meisso, Beki, Dire Dawa and

Djibouti. Other secondary markets are Bardode and

Gadamaito (Garba Issa) in Shinile/Aar Zone border.

Mieso market is a relatively major market or livestock

sale. Tuesday and Thursday are the main camel and

cattle market day and shoats are sold every day.

Traders rom Addis Ababa and Djibouti buy all

three species o livestock rom Meisso market. Oromo

armers also buy young oxen or ploughing purposes.

Traders are not able to buy livestock rom Dire Dawa

and take to Shinile Zone, as this is considered

contraband. Traders thereore opt to buy livestockrom Meisso and Beki or sale to other destinations

including Djibouti. Beki market days are the same as

Meisso market days. Beki market mainly sells cattle,

shoats, camel and donkeys in that order. Traders buy

livestock rom this market and take them to Djibouti,

Addis Ababa, and Dire Dawa. Dire Dawa is a

major sheep and goat market but not or cattle and

camels.

Djibouti market takes all three species shoats, cattle

and camels. The main source or Djibouti market istraders who purchase their livestock rom Meisso,

Beki, and other markets in the zone. These traders



have a double-advantage by selling in Djiboutirst

rom currency exchange (which avours traders who

purchased in Ethiopian birr), and, second, by

investing in relatively cheaper goods rom Djibouti

(mainly second-hand clothes, and oodstus and other

goods) or exchange with other traders in Shinile

Zone. These goods are considered contraband and

thereore come through unofcial channels. Whenthere is no ban on international livestock export, the

main livestock market becomes Djibouti, which will

absorb all species. However, this market is considered

illegal and traders use unofcial means to export

livestock. Main markets or oodstus and clothing are

Meisso and Dire Dawa. The pastoralists mainly buy

sorghum, sugar, tea, salt and other household goods,

including clothing.

There is a signicant trafc o unofcial cross-border

trade going on in Shinile Zone. This trade is locally

reerred to as contraband implying that it is

unlicensed by the government o Ethiopia. This is

mainly trade in clothing both used and new but

also non-ood household items (known as bagaash),

electronics and ood items may be imported in this

way. The main source or such goods is Djibouti,

transiting through Dikhil in Djibouti, Abdulqadir

and Harirat in Somalia, and the destination is in

Shinile Zone, mainly Adem district. A signicantamount o these contraband items are also transited

through other parts o Shinile Zone, destined or

neighbouring Regions/Zones o Ethiopia. Whether it

is transshipped or expressly passed through the

Shinile Zone, goods imported in this way would then

get to several markets outside the zone.

Furthermore, Mieso market has been an import

source o small ruminants or export abattoirs near

to Addis Ababa (Hailemarium Teklewold et al.,

2009), with agents rom ELFORA and Mojo

oten present. In the past, exporters had also used

Young goats purchased by the ELFORA export company in Mieso market.



the railway to move animals to Djibouti, but at

the time o this study in March 2010 the railway

to Djibouti was not operating.

Metehara and Mieso markets are the two most

important source markets or exportable livestock in

the central Rit valley Mieso is also a secondary

market that sourced livestock rom the surroundingprimary livestock markets, such as Bedesa, Chiro,

Hirna, Beroda, Asebot, Kora and other markets in

the neighbouring Somali lowlands Mieso is located

in a strategic location or livestock export via Djibouti.

Some live animal exporters like SHAG are using

this opportunity especially or cattle and camels.

SHAG has a collection/waiting centre at Mieso

where it collects animals rom Borena, Bale and

Wello, conditions them, and passes them through the

quarantine requirements and transports them to the

port using railway (Getachew Legesse et al.,

2008).

The regular use o the Mieso market by the Issa

to sell livestock also had benets or Oromo

armers, because the Issa were important buyers o

cereals at the market. As noted in the PRA report

o Mercy Corps in November 2009, the Mieso

market had not been accessible to the Issa or

some months, due to confict with Oromo and

Hawiya. Other markets or the Issa include Bikie,Adem, and Gadamayito (Garba Issa) on the

border o Shinile Zone and Aar Region. Due to

their more distant location, Djibouti markets are a

last resort.

Not well covered in marketing reports is the

importance o Mieso market as a source o

camels to other areas o Ethiopia, including the

salt mines in Aar, the Raya plains, Minjar, and

parts o Tigray (Yacob Aklilu, personalcommunication).

3.1.2 Natural assets

Natural assets are the natural resource stocks rom

which resources ow and services useul or

livelihoods are derived.

a. Environment, rainall, and mobility

Background notes

Pastoralists live in areas which are characterised by

marked climatic variability, especially with respect

to rainall. This variability may be seasonalas in

the alternation between wet and dry seasonsor

more serious, as in multi-year droughts. Livestock

mobility is one o the most eective techniques

Arican pastoralists have developed or both

exploiting and coping with regular seasonal

variability and droughts in these semi-arid and aridareas.

Although pastoralist mobility is oten seen as a

nonprogressive and unproductive way o lie,

economic analysis shows that herd movement is

eective in preserving livestock production in

variable climates. Research in this area dates back

to the mid-1970s and produces consistent results

using dierent methodologies in dierent areas

(COMESA, 2009). Although the scientic

evidence shows the eciency o herd mobility in

pastoral areas and the need or relatively large

system boundaries to optimize land use,

diminishing access to grazing areas is one o the

most important long-term trends aecting

pastoral livelihoods. Pastoralists ace competition

rom armers and agropastoralists, have their lands

allocated to commercial rice, sugar, or cotton

schemes, are displaced by large-scale dams or

other development projects, are excluded rom

wildlie conservation areas, and, in some areas,are aected by bush encroachment that prevents

grazing by livestock. These pressures on grazing

land increase tensions and conficts between

groups, which in turn can limit access to

communal grazing areas.

Environment and mobility in and around Mieso-Mulu

woreda

Shinile Zone is characterised by three main typeso rangeland, viz. closed savanna (bushland), open

savanna (bush-grassland), and open grassland

(Table 6). As in otherworedas in Shinile Zone, Issa

herders in Mieso-Mulu woreda practice seasonal

movements o livestock to access water and

grazing, including movements outside the woreda

to other parts o Shinile Zone, to Aar Region, or

to Djibouti or Somalia. Whereas the woreda is an

administrative unit, a more useul way o looking

at pastoralist movements is to understand the

wider ecosystem boundaries that ollow anenvironmental and economic logic, rather than

administrative borders.



Issa dry season or drought movements with their

livestock include

movementswestwardsintoAfarRegion,

especially towards the Awash river during dry

periods, which are confict-prone due to

longstanding animosity between Aar and Issa

over control o the Allighedi plain and Awash

river (e.g., see Markakis, 2003);movementssouthwesttowardsAwash,again,

towards the river and especially during

drought;

movementsnorthintoDjibouti,andfurther

north and east as ar as Zeila on the Somali

coast;

movementsintoAwbereandJijigaworedas o

Somali Region, moving as close as around 20

km rom Jijiga town;

attimesofseveredrought,movementssouth

into Babile woreda o Oromiya, moving as ar

as Faen.

Notes: the dotted red line approximates the boundaries o the most distant movements during dry

seasons or drought. All borders shown are unocial.

Figure 3. Dry season and drought-related movements o Issa pastoralists rom Shinile Zone

These movements within the wider pastoral

ecosystem are aected by actors such as

rainfall,whichdeterminestheavailabilityof

water and vegetation both spatially and

temporally. In general, lower rainall within

the overall system leads to greater movement,

as water and vegetation are less available;

accesstospecictypesofgrazeorbrowsespecies, such as the salty bush species in the

coastal areas o Somalia that are avoured or

camels;

accesstomineralsorsaltsforlivestock;

avoidanceofareasinfestedbyticksorother

parasites;

trendsinaccesstooruseofgrazinglands,

such as

o allocation o grazing areas or crop

production. This is a well-known trend inthe woreda due to expansion o Oromo

agropastoralism, and has been a



longstanding and important cause o

confict; see section 3.2.3b;

o bush or weed encroachment which reduces

the availability and diversity o useul graze

or browse species in a given area. This trend

then increases the need to access alternative

areas, thereby urther contributing to

confict;animportantsetofpolicyandinstitutional

arrangements such as those that

o legally or illegally assign land to one group

o users in preerence to another group;

o create a legal or policy vacuum or vagueness

in terms o ormal recognition o communal

grazing areas and related user rights;

o enable or hinder movements across ocial

boundaries, especially international borders;

o ail to prevent rangeland degradation, or

example, by absence o policies on bush

encroachment or through ailure to enact

policies;

o attempt to provide basic services using

xed-point delivery systems that are not

suited to pastoralist mobility;

o can create incentives or certain actors to

instigate confict as means to sustain undingor confict resolution activities.

The remainder o this section ocuses on rainall

and rangeland issues, whereas policy and

institutional issues are described in section 3.3.

Annual rainall estimates2 or Mieso-Mulu woreda

are shown in Figure 4. The average rainall rom

1996 to 2009 was 676 mm compared to a zonal

average o 557 mm.

Source: FEWSNET (2010) rainall estimates (RFE).

Figure 4. Annual rainall in Mieso-Mulu woreda and Shinile Zone, 1996 to 2009

2 Rainall Estimation (RFE) imagery used by FEWSNET is an automated (computer-generated) product which uses Meteosat inrared data,rain gauge reports rom the global telecommunications system, and microwave satellite observations within an algorithm to provide RFE inmm at an approximate horizontal resolution o 10 km. Further details are available rom Herman et al., 1997.

Mieso woreda

Shinile Zone

Mieso woreda

Shinile Zone



Figure 4 indicates a series o relatively dry years in

the woreda and Zone between 2001 and 2005,

suggesting that mobility issues may have been

particularly important during those years; no

long-terms trends are evident, partly because the

data available covers only a relatively short

12-year period. In both drier years and wetter

years, movement is a normal event rom aroundmid-November to the end o March. These

movements were mainly to the oothills in the

south o the Zone and into Oromiya Region, to

Note the agreement between the actual rainall data in the graph (above) and the rainall patterns

shown on the seasonal calendar (below).

Jijiga Zone, and to sub-coastal areas in Somalia.

Within any given year there is marked seasonal

variation in rainall. Average seasonal rainall

patterns or the period 1996 to 2007 are shown in

Figure 5, together with the seasonal calendar rom

the DPPA/SCUK baseline report, which used a

reerence year o 2004 to 2005. The main rainyseason (karan) alls between July and October, and

the lesser rains (diraa) all between March and

May.

Seasonal variability in rainall in Mieso-Mulu woreda

over the 14-year period between 1996 and 2009 is

shown in Figure 6, and refects typical variations in

pastoralist areas. For example, note the marked

changes in rainall over time during March, April,

May (the diraa rainy season), and July, August,

September, and October (the karan rainy season).

Given the importance o mobility or pastoral

livestock production and the role o rainall in

determining movement, analysis o rainall should

Figure 5. Mean monthly rainall in Mieso-Mulu woreda, 1996 to 2007, and

seasonal calendar or Shinile pastoralist areas

Sources: rainall estimates rom FEWSNET (2010); seasonal calendar rom DPPA/SCUK, 2008.

include the wider ecosystem within which Issa

herders might have access. For example, good

rainall outside o Shinile Zone in neighbouring

areas o Aar Region, Jijiga Zone in Somali

Region, Djibouti, and northwest Somalia would,

in theory, provide options or Issa to move into

these areas should rainall within the Zone all

below normal. Similarly, reduced rainall in

neighbouring areas would lead to reduced

movement options and increased pressure on

resources within the Zone.

Meanmonthlyrainfall

(mm)



In summary, the available rainall data indicate

that, in Ethiopia, the rainall data rom

FEWSNET date back only to 1996, so long-term

trends cannot be analyzed. The 14-year period

rom 1996 to 2009 shows a common pattern

across Shinile Zone and neighbouring areas o

Aar Region and Somali Region, viz. a relative

decline in annual rainall between 2001 and 2005,with an upwards trends towards pre-2001 levels

in 2006 and 2009. In 2009, rainall was relatively

poor and similar to 2004. In Mieso-Mulu woreda,

there was typical variability in monthly rainall

rom 1996 to 2009 (Figure 6). Data rom areas

adjacent to Shinile Zone, such as Awdal in

Somalia (Somaliland), indicate that this variability

tends to smooth out across longer time periods,

rom the 1920s onwards.

The reerence line at 676 mm is the mean total rainall in the woreda over the 14-year period, 1996 to 2009.

Figure 6. Monthly variation in rainall in Mieso-Mulu woreda, 1996-2009

Source: FEWSNET (2010) rainall estimates (RFE).

Livelihood zone maps produced by DPPA/

SCUK (2008) indicate that agropastoralism by

Oromo or Somali Hawiya communities accounts

or around hal o Mieso-Mulu woreda by land

area, suggesting that approximately 50 percent o

the woreda to the south is not accessible to Issaherders. At the same time, Oromo or Hawiya

agropastoralists may try to access pastoralist

grazing areas, especially during dry periods.

Grazing areas in neighbouring Aar Region to the

west may also be dicult to access or Issa due to

confict. This situation explains why Issa mobility

includes Jijiga Zone o Somali Region and

sub-coastal areas o Somalia.

Superimposed on the rainall and seasonal

mobility patterns are issues o rangeland quality,

and, in particular, bush and weed encroachment



in relatively higher and wetter closed savanna

areas (Table 6). These changes indicate that not

only is some Issa pastoralist land being allocated to

crop production by non-Issa groups, but the

remaining pastoralist land is being degraded by

woody species and weeds.

The changes shown in Table 6 are supported bydata on livestock herd composition, recording

changes over 60 years in Erer and Ayisha woredas.

These data show a slow shit towards greater

ownership o camels and goats, and ownership o

ewer cattle. This was explained by declining

grasslands and the capacity o browsers to make

use o encroached bushland (Amaha Kassahun

Gezahegn, 2006). These changes are discussed in

more detail in section 3.2.3e.

Table 6. Rangeland types and trends in weed and bush encroachment in Erer and

Ayisha woredas, Shinile Zone

Type o Description

rangeland

Closed savanna

(bushland)

Open savanna(bush-grassland)

Open

grassland

Notes: see Figure 1 or locations o Erer and Ayisha woredas relative to Mieso-Mulu woreda. The types o

rangeland shit rom closed savanna (bushland) in the higher south o the Zone to open grassland in the

lower north o the Zone (also see Figure 2).

Rainfall300-450mm;temperature27-43C;altitude700-1200m;area

27,700km2 or ~30% o Shinile Zone

Over-cuttingofusefultreespeciesforcharcoalproduction,rewood,and house construction by non-pastoral groups, plus clearing o trees

or crop production

Acacia nubica andA. melliera have been observed not only as

encroaching woody plants but ever expanding across the closed

savannas. Based on the current trend it could be anticipated that the

woody plants would expand to most o the canopy covers o the closed

savannas in Shinile Zone. Both species are ound to be ecologically

unriendly causing a tremendous decline in the productivity o

important grass species.

Rainfall300-350mm;temperature25-40C;altitude600-850m;area25,500 km2 or ~30% o Shinile Zone

GreaterencroachmentofweedsXanthium abyssinica and Parthenium

hysterophorus by 50% more and 20% more relative to open grassland

Acacia nubica categorised as an aggressive woody encroacher, invading

more than 50% o the open savanna

Ingeneraltheopensavannasareunderseriousthreatfrom

encroachment by unpalatable and noxious woody plantswhich may

urther lead rom a partial to a total ailure o the rangelands to produce

enough orage or livestock.

Rainfall300-350mm;temperature27-42C;altitude500-700m;area

36,000 km2 or ~40% o Shinile Zone

Dominatedbygrassspecies,especiallyperennialsofthegenera

Panicum, Dactyloctenium, Eleucine, Cenchrus, Leptochloa, Sporobolus,

Brachiar ia, Cynodon, Cyperus, Phalaris, and Polypogon; around 49 grass

species identied and 11 non-grass species

Tribulus terrestris, Xanthium, and Parthenium encroaching more than 50%

o the grazing lands and a threat to productivity

Source: adapted rom Amaha Kassahun Gezahegn (2006).



3.1.3 Human assets

Human assets represent the skills, knowledge,

ability to labour, and good health that together

enable people to pursue dierent livelihood strategies

and achieve their livelihood objectives.

a. Indigenous skills and knowledge

Background notes

Pastoralists live in some o the harshest and

risk-prone environments in the world, with

severe limitations in the availability o water and

vegetation. Their ability to adapt and surv

Moving Up or Moving Out? A Rapid Livelihoods and Conflict Analysis in Mieso-Mulu Woreda, Shinile...

Documents

Transcript of Moving Up or Moving Out? A Rapid Livelihoods and Conflict Analysis in Mieso-Mulu Woreda, Shinile...