Somali Region

RESEARCH
REPORT 57
ID S_M a ster L ogo
Vulnerable Livelihoods in
Somali Region, Ethiopia
Stephen Devereux
April 2006
About IDS
The Institute of Development Studies is one of the world's leading
organisations for research, teaching
and communications on international development. Founded in
1966, the Institute enjoys an
international reputation based on the quality of its work and the
rigour with which it applies academic
skills to real world challenges. Its purpose is to understand and
explain the world, and to try to change
it to influence as well as to inform.
IDS hosts five dynamic research programmes, five popular
postgraduate courses, and a family of worldclass web-based knowledge services. These three spheres are
integrated in a unique combination as a
development knowledge hub, IDS is connected into and is a
convenor of networks throughout the
world.
The Institute is home to approximately 80 researchers, 50
knowledge services staff, 50 support staff and
about 150 students at any one time. But the IDS community
extends far beyond, encompassing an
extensive network of partners, former staff and students across the
development community
worldwide.
I D S _M a ste r L o g o B l
ack
I DS _ Ma st er
ze
L o g o _M i n i mu m
Si
Mi n i mu m S i z e
mSi
ze
X: 1 5 mm
Mi n
i mu
X: 1 5 mm
For further information on IDS publications and for a free catalogue,

contact:
IDS Communications Unit
Institute of Development Studies
at the University of Sussex
Brighton BN1 9RE, UK
Tel: +44 (0) 1273 678269
Fax: +44 (0) 1273 621202
E-mail: bookshop@ids.ac.uk
Web: www.ids.ac.uk/ids/bookshop
IDS is a charitable company, limited by guarantee and registered in
England (No. 877338).
IDS
RESEA
RCH
REPOR
T 57
IDS RESEARCH REPORT 57
Vulnerable Livelihoods in
Somali Region, Ethiopia
Stephen Devereux
April 2006
1 Institute of Developm
at the University
Brighton BN1 9RE UK
Vulnerable Livelihoods in Somali Region, Ethiopia

Stephen Devereux
IDS Research Report 57
Photo: Andrew Heaven
First published by the Institute of Development Studies in April 2006
Institute of Development Studies 2006
ISBN 1 85864 612 X
A catalogue record for this publication is available from the British Librar
y.
All rights reserved. Reproduction, copy, transmission, or translation of any
part of this publication may be made
only under the following conditions:
with the prior permission of the publisher; or
with a licence from the Copyright Licensing Agency Ltd, 90 Tottenham
Court Road, London W1P 9HE, UK,
or from another national licensing agency; or
under the terms set out below.
This publication is copyright, but may be reproduced by any method
without fee for teaching or non-profit
purposes, but not for resale. Formal permission is required for all such
uses, but normally will be granted
immediately. For copying in any other circumstances, or for re-use in other
publications, or for translation or
adaptation, prior written permission must be obtained from the publisher
and a fee may be payable.
Available from:
Communications Unit
Tel: +44 (0) 1273 678269
Fax: +44 (0) 1273 621202
Web: www.ids.ac.uk/ids/bookshop
Printed by Imagedata Group, Brighton UK
Typeset by Barbara Cheney
IDS is a charitable company limited by guarantee and registered in
England (No. 877338).
IDS
RESEA
RCH
REPOR
T 57
Contents
List
5
List
6
List
7
Acknowledgements
8
Acronyms
9
Executive
11
of
figures
of
tables
of
boxes
summary
Section 1 Context
1
19
Introduction
1.1
19
1.2
20
1.3
20
2
26
Research
Structure
of
this
report
Methodology
Somali
2.1
26
2.2
30
issues
Region:
the
place
and
the
people
Population and demography in Somali Region

The
paradox
of
wealth
plus
vulnerability
Section 2 Livelihoods
3
33
District
3.1
34
3.2
37
3.3
39
3.4
42
4
44
4.1
44
livelihood
profiles
Pastoralist
districts
Agro-pastoralist
districts
Farming
districts
Urban
centres
Livestock
rearing
Livestock
ownership
and
in
marketing
Somali
Region
4.2
48
4.3
50
4.4
55
4.5
56
5
62
Livestock
The
marketing
marketing
system
6
68
Marketing
constraints
7
74
Farming
systems
in
Somali
Region
prices
in
Land
Somali
Region
agro-pastoralism
in
Somali
Region
and
access
ownership
Harvests
Trends
in
crop
production
displaced
persons
Livelihoods
in
Hartisheik
camp
Conditions
in
Hartisheik
camp
Hartisheik
camp
Food
aid
in
Government policy towards the IDP camps

The
way
Incomes
7.1
74
7.2
76
7.3
76
7.4
81
7.5
83
7.6
84
Region
and
Internally
6.1
69
6.2
70
6.3
71
6.4
71
6.5
72
Somali
Livestock
Farming
5.1
63
5.2
63
5.3
65
5.4
66
in
forward?
and
inequality
Household
incomes
Income
Returns
to
inequality
livelihood
Food
activities
consumption
Household
Livelihood
problems
purchases
and
changes
Section 3 Vulnerability
8
86
Livelihood vulnerability and mortality outcomes

8.1
86
8.2
90
Trends
in
self-assessed
vulnerability
Mortality
Drought
and
hunger
Somali
Region
93
9.1
93
9.2
99
9.3
101
Drought
in
Rainfall
Access
trends
to
water
10
Conflict
105
10.1 The complex nature of conflict in Somali Region
105
10.2
Impacts
of
conflict
108
10.3
Conflict
resolution
111
10.4
Compensation
payments
111
10.5
Vulnerability
to
conflict
113
11
Governance
and
political
representation
114
11.1
Political structures and policy processes
114
11.2
Perceptions
of
political
representation
116
12
Gendered
vulnerability
119
12.1
Gender
bias
in
Somali
Region
119
12.2.
Intra-household
discrimination
122
12.3
Khat
consumption
123
12.4
Gendered
division
of
incomes
124
12.5
Women
and
the
future
127
Section 4 Responding to vulnerability

13
128
13.1
128
13.2
128
13.3
131
14
Coping
Coping
strategies
strategies
in
the
literature
Coping strategies adopted in Somali Region

Gendered coping strategies in Somali Region
Informal
transfers
133
14.1
133
14.2
135
14.3
136
14.4
138
14.5
138
14.6
140
15
141
15.1
141
15.2
143
15.3
145
15.4
146
15.5
147
15.6
148
16
149
16.1
149
16.2
150
16.3
153
17
154
17.1
155
17.2
156
17.3
158
17.4
159
17.5
162
Informal transfer mechanisms in Somali Region

Access
Availability
Gender
to
informal
transfers
of
informal
transfers
and
informal
assistance
Zakaat
The
limits
of
redistribution?
Safety
Trends
Food
nets
in
aid
Receipts
food
aid
targeting
of
and
safety
Uses
transfers
food
tastes
Alternatives
distribution
net
of
Beneficiary
deliveries
and
to
aid
preferences
food
Health
Causes
services
of
death
Access
to
Quality
aid
in
Somali
health
of
health
Region
services
services
Education
Literacy
Access
to
education
Costs
of
education
Attitudes
Quality
to
of
education
education
services
IDS
RESEA
RCH
REPOR
T 57
Section 5 Conclusions and implications for policy

18
Somali attitudes to the future of pastoralism
163
19
Implications
for
policy
168
19.1
Is Pastoralism unviable in Somali Region?
168
19.2
How
to
plan
for
unpredictability?
169
19.3
Is
sedentarisation
the
solution?
172
19.4
What kinds of safety nets are needed?
173
Annex:
Household
questionnaire
175
Glossary
192
Bibliography
194
Figures
Figure 2.1 Map of Ethiopia, showing Somali Region
27
Figure 3.1 Somali Region, with survey sites highlighted
33
Figure 4.1 Annual livestock exports from Berbera and
Bosaso ports, 19942004 59
Figure 5.1
Perceived trends in harvests, 19942004
67
Figure 7.1 Lorenz curves of household incomes, Somali
Region 76
Figure 8.1 Trends in self-assessed vulnerability, by district
87
Figure 8.2 Trends in self-assessment index and destitution,
Somali Region 89
Figure
8.3
Deaths
per
year,
19902004
92
Figure 8.4 Deaths per household and by season, 19922004
92
Figure 9.1 Total annual rainfall, Gode, 19572002 (mm)
99
Figure 9.2.
Long-run rainfall trends, Gode (mm)
99
Figure 9.3 Total annual rainfall, Jigjiga, 19522002 (mm)
100
Figure 9.4
Long-run rainfall trends, Jigjiga (mm)
100
Figure 9.5 Sources of water in birka-dependent areas of
Somali Region 101
Figure 9.6 Sources of water in well-dependent areas of
Somali Region 103
Figure 9.7 Sources of water in river-dependent areas of
Somali Region 103
Figure 9.8 Sources of water in urban areas of Somali Region
104
Figure 12.1
Sex ratios in Somali Region
120
Figure 12.2 Missing females in Somali Region (1997
Census data) 121
Figure 12.3 Crude life expectancy estimates (household
survey data) 121
Figure 13.1 A categorisation of household coping strategies
129
Figure 14.1 Most recent payment of zakaat, by district
139
Figure 15.1 Food aid on sale in Jigjiga town, February 2004
142
Figure 16.1 Deaths from natural and unnatural causes, 1992
2004 150
Figure 16.2 Perceived quality of health services, Somali
Region
153
Figure 17.1 Perceived quality of education services, Somali
Region 162
Tables
Table 1.1
Summary of household questionnaire
21
Table 1.2 Sampling frame for the household survey in
Somali Region 24
Table 2.1 Population of Somali Region, 1997 (ranked by
zone population size) 27
Table 2.2 Household size and composition, by district
28
Table 2.3 Males per 100 females in Somali Region, by age
cohort 28
Table 2.4
Absentee household members, by district
29
Table 2.5 Demographic structures in developing countries
and in Somali Region 30
Table 2.6 Poverty headcounts and calorie consumption in
Ethiopia, by region 31
Table 2.7 Anthropometric indicators in Ethiopia, by region
32
Table 4.1
Households owning livestock, by district
44
Table 4.2 Livestock ownership, by district (all households)
45
Table 4.3 Reasons for decreasing livestock ownership, 1995
2005
47
Table 4.4 Gross daily revenue from livestock sales in four
Somali markets, 2005 48
Table 4.5 Time-line for livestock traders in Somali Region,
19922005 49
Table 5.1
Farming and access to land, by district
64
Table 5.2
Most recent harvests, by district
65
Table 5.3 Reasons why recent harvests are worse than in past
years 67
Table 7.1 Average household cash income, by district
75
Table 7.2 Average income by livelihood category, and by
highest and lowest returns 77
Table 7.3 Most prevalent livelihood activities in Somali
Region
77
Table 7.4 Consumption of different food types, by livelihood
category 83
Table 7.5 Purchases of new clothes, by livelihood category
84
Table 7.6 Basic items in the home, by livelihood category
84
Table 7.7
Livelihood problems ranked, by livelihood
category
85
Table 8.1 Mortality in Somali Region, 19912004/5, by
district
90
Table 8.2 Mortality in Somali Region per household, by age
cohort 91
Table 8.3 Mean age at death in Somali Region, by location
and livelihood 91
Table 9.1
Indicators of hunger in 2004, by district
95
Table 10.1 Households affected by conflict in Somali
Region, by district 109
Table 10.2 Impacts of conflict in Somali Region, by district
110
Table 10.3 Institutions for conflict resolution, by district
111
Table 10.4
Compensation payments, by district
112
Table 10.5 Relationships with neighbouring groups, by
district
113
Table 11.1 Perceptions of fairness and effectiveness of
government representation 117
Table 13.1 Household coping strategies in Somali Region
130
Table 14.1 Transfers received in Somali Region, 2004/5
134
Table 15.1 Food aid deliveries to Somali Region, 19952004
141
Table 15.2 Household receipts of food aid, by district
146
Table 16.1
Causes of death in Somali Region
149
Table 16.2
Access to health facilities, by district
151
Table 17.1
Literacy rates, by district and gender
155
IDS
RESEA
RCH
REPOR
T 57
Boxes
Box 4.1
Women livestock traders in Degahbur
55
Box 4.2 Impact of the 1998 Saudi livestock ban on Somali
traders 58
Box 4.3 Government harassment of livestock traders, 2004/5
60
Box 4.4
Case study: contraband trader in Jigjiga
61
Box 4.5 Case study: impacts of currency fluctuations on
Somali traders 62
Box
5.1
Case
study:
riverine
farmer
66
Box 5.2
Case study: agro-pastoralist household
66
Box
6.1
Abdi
Koranes
story
68
Box
6.2
Feinuzs
story
69
Box
6.3
Fatimas
story
70
Box
7.1
Case
study:
charcoal
sellers
78
Box 7.2
Case study: building materials trader
79
Box
7.3
Case
study:
mat-maker
80
Box
7.4
Case
study:
shoe-repairer
81
Box
7.5
Case
study:
tailor,
Gode
town
82
Box 9.1 Food insecurity in Warder Zone, Somali Region,
August 2004 94
Box 9.2
Environmental degradation in Somalia
98
Box 10.1
Conflict in Shilabo District, 19982005
107
Box 12.1 Celebrating sons: intra-household discrimination
in Somali Region 122
Box 12.2 More important than food: womens attitudes to
chewing khat 124
Box 12.3 Trying to forget our worries: mens attitudes to
chewing khat 125
Box 13.1 Case study: household drought coping strategies

131
Box 16.1
Case study: midwife, Gashamo town
151
Box 16.2
Case study: midwife, Shilabo District
152
Box 16.3 Case study: HIV/AIDS womens group, Gode
town
154
Box 17.1 Case study: school headmaster, Gode town
157
Box 17.2 Case study: Koranic teacher, Shilabo District
159
Box 18.1 People of animals? Somali mens attitudes to the
future of pastoralism 164
Box 18.2 No good future in pastoralism? Somali womens
attitudes to the future of
pastoralism
165
Box 18.3 Better to have an education than a hundred
camels? A gendered debate
on the future of pastoralism in Somali Region
167
Acknowledgements
Particular thanks go to the following, who worked on the
study:
Household survey team: Ali Ahmed Kalib (Team Leader
phase I), Mohamid Mohamid Ismail
(Team Leader phases II & III), Ahmed Sheikh Mahamud,
Hassan Aden, Khadir Abdi Fidar,
Naima Mohamed Omar, Mohamed Ali Farah.
Qualitative fieldwork team: Nuur Mohamud Sheekh (Team
Leader), Muhamed Ali Omer,
Fatiya Mohamed Hare, Colette Solomon (gender specialist),
Ali Ahmed Kalib, Abdi Ibrahim
Nuur, Mohmed Ibrahim Mohamed.
Trade and marketing survey team: Abdi Umar (Team
Leader), Ahmed Sheikh Mahamud,
Hassan Aden, Mohamed Abdi Hala, Dahir Ali Ofle, Sirat
Osman, Mohamed Abdi Feder,
Mokhtar Abdi Handulle, Shukri Abdi Noor, Osman
Muhumed Hassan, Bizuhon Legesse
(analyst).
Fieldwork coordination: Ahmed Waqo Hapi (protocol);
Gordon Winthrop (logistics).
Data analysis: Alemayehu Seyoum Taffesse (African Centre
for Economic and Historical
Studies), Bob Baulch, Ricardo Sabates, Rachel SabatesWheeler.
Report reviewing and editing: Ian Scoones; Patta ScottVilliers; Jenny Edwards. Chapter 4
was co-authored with Abdi Umar. Chapter 6 was co-authored
with Colette Solomon.
Thanks also to the following offices for facilitating and
supporting the work:
UNOCHA Ethiopia Office and the UNOCHA Pastoralist
Communication Initiative; the UK
Department for International Development Ethiopia; the
Pastoralist Development
Department, Ethiopia Ministry of Federal Affairs; the
Pastoralist Affairs Standing
Committee, Ethiopia House of Peoples' Representatives; the
Ministry of Livestock
Development, Republic of Somaliland; Somali Regional
State Office of the Regional
Parliament; the President's Office, Somali Regional State;
Gudigga Dhaganka Gurtida (the
Somali Customary Council of Elders), Somali Regional
Pastoral and Rural Development

Coordination Office; the Somali Regional Livestock, Crop
and Natural Resource
Development Bureau; the District Administrations of
Gashamo, Kabridahar, Shinile, Shilabo,
Kebribayah, Doboweyn, Cherati, Kelafo, Dolo Odo, Ferfer,
Jigjiga, Mustahil, Bokh, Warder
and Moyale; the Municipal Administrations of Jigjiga and
Gode; Oromia Pastoralist
Development Commission, Negelle District Administration;
Yabello Zonal Administration;
Save the Children (UK); Hope for the Horn; Ogaden Welfare
Development Association.
IDS
RESEA
RCH
REPOR
T 57
Acronyms
CSB Corn soy blend
DFID Department for International Development (UK)
DPPB Disaster Prevention and Preparedness Bureau
DPPC Disaster Prevention and Preparedness Commission
EPRDF Ethiopian Peoples Revolutionary Democratic Front
FDRE Federal Democratic Republic of Ethiopia
FEZ Food economy zone
GAM Global acute malnutrition
GoE Government of Ethiopia
HICE Household Income, Consumption and Expenditure
survey
ICRC International Committee of Red Cross and Red
Crescent societies
IDP Internally displaced person
IDS Institute of Development Studies
MoFED Ministry of Finance and Economic Development
OCHA Office for the Coordination of Humanitarian Affairs
(UN)
ONLF Ogaden National Liberation Front
OWDA Ogaden Welfare and Development Association
OWS Ogaden Welfare Society
PCAE Pastoralist Concern Association Ethiopia
PCI Pastoralist Communication Initiative
SAM Severe acute malnutrition
SC-UK Save the Children UK
SERP South East Rangelands Project
SNNPR Southern Nations, Nationalities and Peoples Region
SNRS Somali National Regional State
TBA Traditional birth attendant
TLU Tropical livestock unit
UN United Nations
UNHCR United Nations High Commission for Refugees
UNOCHA United Nations Office for the Coordination of
Human Affairs
WFP World Food Programme (UN)
WMU Welfare Monitoring Unit
10
IDS
RESEA
RCH
REPOR
T 57
Executive summary
This report is an investigation into the causes and
consequences of livelihood vulnerability in
Somali Region, Ethiopia. People in this region pastoralists,
agro-pastoralists, farmers and
traders have suffered a series of livelihood shocks in recent
years, some natural (droughts,
livestock disease), others political (a crackdown on
contraband trade, bans by Gulf states on
livestock imports, violent conflict between (sub-)clans or
between Ogaden National
Liberation Front (ONLF) militia and the state). As a result of
these multiple shocks, and
because rainfall in the Horn of Africa has been low in recent
years, questions are being
asked about the sustainability of pastoralism as a livelihood
system, not only in Somali
Region but throughout the Greater Horn of Africa. The
Government of Ethiopia, for
instance, is advocating rural sedentarisation of pastoralists as
one long-term option.
In this highly politicised context, this research study is an
attempt to gather information and
give voice to the people of Somali Region themselves, and to
present this evidence to
policy-makers (community leaders, the regional and federal
governments, international
donors, and international and local NGOs) as an input to their
strategising and decisionmaking. A mixed methodology approach was designed for
our fieldwork, including a
questionnaire survey of 1,100 households in nine rural
districts and two urban centres across
the region, ethnographic methods (key informant interviews,
life histories, community
discussions), market monitoring and semi-structured
interviews with traders, internally
displaced persons (IDPs) and members of government.
Early during the inception phase of this study, the complexity
of livelihood systems in Somali
Region became evident to the research team. Four dominant
livelihood systems were
identified and used as a basis for stratifying the sample,
analysing the data and writing up

the findings. Pastoralists pursue livelihoods dominated by
livestock rearing; agro-pastoralists
pursue a mixed livelihood of both livestock rearing and crop
farming; farmers live in settled
communities and cultivate crops for food and cash income;
urban residents live in towns and
earn their living from formal or informal employment. Two
other categories of people
discussed in this report are traders (and the range of
marketing agents involved, especially in
livestock trading) and IDPs.
Although pastoralists, agro-pastoralists, farmers, urban
residents, traders and IDPs are
treated as distinct livelihood categories for analytical
purposes in this report, the boundaries
between these categories are not always clear-cut, and it is
important to recognise the
interdependence of these various groups. The economy of
Somali Region is a complex,
interconnected system of social networks and political
negotiations, where the sustainability
or vulnerability of each livelihood depends as much on the
individuals interpersonal
relationships, and on international geopolitics, as on his or
her assets and income at any
point in time. In this context, livelihood vulnerability is
affected by processes of social change
(such as a decline in the gift culture due to economic
stresses and shifting social values),
and by political instability (from localised conflicts over
resources to the expulsion of Ethiopian
Somalis from Djibouti). Drought triggers livelihood crises,
but the underlying causes of
vulnerability in Somali Region are social and political, not
natural.
Livelihoods in context
The first section of this report sets the dominant livelihood
systems in Somali Region in context,
by providing livelihood profiles for each of the 11 districts
sampled for the household survey.
Of the three pastoralist districts surveyed, Gashamo was the
epicentre of the 2004 drought.
Paradoxically, given the belief that settling pastoralists may
be the best way forward for
Somali Region, Gashamo has a number of semi-permanent
settlements, but it was the
relative immobility that this introduced that caused livestock
to die in large numbers in
2004. Shilabo is an Ogaden district bordering on Somalia,
which prospers from the large
Ogadeni diaspora and from cross-border trade. Despite being
relatively wealthy in terms of

camels and cash, Shilabo is vulnerable to conflict involving
the ONLF, government forces
and Ogadeni sub-clans. Shinile, in the north of Somali
Region, is rather cut off and isolated,
11
which might partly explain why the people of Shinile are

poorer than pastoralists in
Gashamo and Shilabo. The Issa people of Shinile have close
linkages with Djibouti, which
itself is poor, rather than with wealthier Somalia or
Somaliland.
Agro-pastoralist communities are characterised by
competition for farmland as well as
water and grazing for livestock, since mobility is restricted
by the need to stay near the farm.
Kebribayah District is located close to the regional capital of
Jigjiga, and is a corridor for
trade to and from Somalia. It is also the location of
Hartisheik, the worlds largest refugee
camp during Somalias civil war in the early 1990s, now
home to thousands of IDPs
displaced by drought or conflict. Families in Doboweyn
District became agro-pastoralists
relatively recently, and have found adjusting to a private
property regime problematic.
Conflicts occur over access to farmland and the boundaries
between individual plots. People
in Doboweyn own plenty of livestock, but do not engage in
livestock trading and have low
cash incomes. Households in Cherati District are even poorer.
Farming and livestock rearing
are mainly for subsistence rather than sale, and tiny incomes
are generated by
supplementary activities such as making baskets and mats.
Two of our three farming districts are riverine (Kelafo and
Dolo Odo) and one is rain-fed
(Jigjiga). Farming in Kelafo and Dolo Odo occurs along the
Shabelle and Dawa/Ganale rivers
respectively, the only two permanent river systems traversing
Somali Region. Farmers in
both districts are predominantly ethnic Bantu, rather than
Somali. They live in a subordinate
relationship to local Somalis, and many are sharecroppers on
land owned by wealthier
Somali businessmen. Despite being politically marginalised
and socially excluded, these are
peaceful and stable communities. Farmers in Kelafo are
rather better off than farmers in
Dolo Odo, mainly because Kelafo farmers supply towns in
neighbouring Somalia with
onions, other vegetables and fruits, while Dolo Odo farmers
do not enjoy access to this
lucrative market, so supplement their foodcrop production by
selling baskets and mats made

from riverbank grasses and reeds. Farmers on the Jigjiga
plains are ethnic Somalis, and take
advantage of higher rainfall and a third rainy season (unique
to northern districts of Somali
Region) to grow long-maturing crops like wheat, barley and
maize.
People in the two urban centres of Jigjiga and Gode are better
off than their rural
counterparts in almost every respect. Not only do they enjoy
higher average incomes, their
incomes are less variable and less vulnerable to unpredictable
shocks. They have better access
to education and health services, and to clean water, so they
are more likely to be literate
and are less vulnerable to preventable diseases. They live
longer and healthier lives. Women
have more autonomy in towns than in rural communities.
Urban livelihoods are strikingly
diversified, with many people employed in the public sector
(government administration,
drivers, teachers, nurses), the private sector (retail stores,
hotel workers, builders, security
guards), or as informal service providers (barbers, tailors). In
rural communities, by contrast,
alternatives to livestock and crop production are limited to
Koranic teachers, traditional
healers and minor income-earning activities such as selling
firewood or charcoal.
The next two chapters consider livestock- and crop-based
livelihood activities in Somali
Region. Livestock are owned by most rural households in our
survey. Camels, cattle, goats,
sheep and donkeys are reared for food, for sale, as pack
animals, or as draught power for
ploughing. The majority of survey respondents reported
declining livestock numbers in
recent years, however, due to a combination of recurrent
droughts, livestock disease,
religious or social contributions of animals as zakaat or
dowry payment, or to assist poorer
relatives. Pastoralist wealth is directly correlated with
livestock ownership and exchange, but
dependence on this source of wealth emerges as a source of
vulnerability, given the
susceptibility of livestock to drought and disease. Livestock
are unprotected against disease
because of the absence of veterinary services, while their
owners are unprotected against
livestock losses because of the absence of insurance
mechanisms, except for limited informal
risk-pooling and restocking arrangements among clan
relatives.
Livestock marketing, mainly across the border with Somalia

and into the Gulf states,
generates enormous revenues for livestock owners, traders
and marketing agents in Somali
Region. The informal marketing system is unregulated, but
sophisticated. However, sellers
have little power in the market excess supply means that
prices are low, droughts reduce
12
IDS
RESEA
RCH
REPOR
T 57
the supply and quality of animals offered, and export routes

are subject to unpredictable but
devastating disruption. Three sources of vulnerability in
livestock marketing are explored:
(1) conflict (traders recall the years of civil war in Ethiopia
and Somalia in the early 1990s as
the most difficult they have experienced); (2) bans on
livestock imports by the Gulf states
(following outbreaks of Rift Valley Fever in the Horn in 1998
and 2000); (3) attempts by the
Government of Ethiopia to control cross-border trade (the socalled war on contraband).
Each of these events has blocked exports of livestock to the
most lucrative market Saudi
Arabia sometimes for weeks, at other times for years.
Farming in Somali Region is confined to the banks of the two
permanent rivers in the
centre and south of the region, and a few districts in the north
where rainfall is sufficient
to support rain-fed agriculture. Most farming is low input,
low output. Yields are low,
marketing opportunities are limited though some farmers
are exporting onions and other
cash crops to Somalia and the risk of crop failure in
drought years is high. Sharecroppers
and landless households in farming communities are
especially vulnerable. Food price
seasonality is a particular risk for all farming households.
Poorer farmers with undiversified
incomes are forced to sell some food crops after harvest to
raise cash, and to buy food later
in the year at much higher prices. With no access to fertiliser,
irrigation equipment, input
credit or agricultural extension services, the prospects for
farmers in Somali Region look
unpromising. In this context, and with most available arable
land already allocated and under
cultivation, it is difficult to see how much more
sedentarisation of Somali pastoralists along
the banks of major rivers can be achieved.
Displaced people and refugees, having lost their livelihoods
and property and become
dependent on social assistance, are internationally recognised
as especially vulnerable. Somali
Region currently has an estimated 80,000 IDPs, though the
numbers fluctuate. Fieldwork for

this study included interviews in Hartisheik IDP camp, where
conditions were particularly
harsh. Food aid had been suspended for six months, the
United Nations High Commission for
Refugees (UNHCR) had withdrawn, sanitation facilities were
non-existent, and under-five
mortality rates exceeded emergency levels. The situation is
complicated by the (direct or
indirect) dependence of many IDPs on contraband trading.
The governments recent
crackdown on contraband has elevated livelihood
vulnerability in Hartisheik as much as the
interruption in food deliveries, reinforcing the urgent need to
regulate cross-border trade.
The household survey included questions on incomes earned,
which allowed returns to
different livelihood activities to be compared, and relative
incomes across districts and
livelihood systems to be assessed. The most lucrative
occupation is trading, with contraband
traders enjoying the highest average incomes albeit at some
risk, like all traders in the
region.
Other
high-earning
occupations
(salaried
employment, construction work, carpentry)
are mainly open to urban residents. Conversely, the lowest
returns are to rural activities that
derive from natural products (selling firewood or charcoal,
making baskets or mats,
beekeeping). Urban incomes are three to four times higher
than rural incomes, though the
costs of living (rent, electricity, food purchases) are higher in
towns. Ranking livelihood
activities reveals that rural households derive most of their
incomes from a limited and
fragile environmental base that is dangerously susceptible to
unpredictable fluctuations in
the weather.
Pastoralists and farmers earn similar levels of cash income.
Overall, agro-pastoralism appears
to generate the lowest returns of all livelihood systems in
Somali Region. Pursuing livestock
rearing and crop farming simultaneously does not spread risk,
as both activities are susceptible
to erratic rainfall, and appears to generate lower returns than
specialising in one or the other.
These data on cash incomes are validated by considering
food consumption patterns; which
confirms that agro-pastoralists have the lowest dietary
diversity, while urban households have
the highest. Agro-pastoralists also purchase clothes less
frequently than other groups, and are
least likely to have basic grocery items (salt, sugar, tea,
kerosene) in their home.

These different livelihood systems are not pursued in
isolation from each other. Most
households in Somali Region engage in multiple activities,
and making a living requires
working with others to an extent that is uncommon elsewhere
for example, in farming
communities in the Ethiopian highlands. In a very real sense,
the livelihood of one depends
13
on the livelihood of others. Pastoralists negotiate access to

land and water with
neighbouring herders and farmers, and sell their livestock
into a lengthy marketing chain of
traders and intermediaries. Farmers sell their produce to
urban residents, who purchase
various goods and services from pastoralists, farmers and
traders. Relatives living temporarily
or permanently abroad in the Somali diaspora retain strong
linkages with their families,
sending back remittances which are invested in the rural
economy and urban businesses. The
regional economy is closely linked to the economies of
neighbouring countries Somalia,
Somaliland, Djibouti, Kenya and the Gulf states and any
disruption to the flow of cash,
livestock and commodities, either within Somali Region or
between the region and the
world beyond its borders, constitutes a major threat to many
local livelihoods.
Sources of vulnerability
The next section of this report examines different sources and
outcomes of livelihood
vulnerability in Somali Region. A household survey question
on self-assessed vulnerability
yields graphic insights into the variability of livelihoods
across the region, with rural
households experiencing repeated cycles of accumulation,
collapse and recovery during the
past ten years, as they are hit by drought and other shocks.
Two major covariate shocks
(affecting entire communities rather than individual
households) are revealed as the
droughts of 2000 and 2004. In most rural districts surveyed,
the 1999/2000 drought
emergency is the most significant shock since the mid-1990s,
but in Gashamo and
Kebribayah the more recent drought dominates. The
vulnerability time-lines for Jigjiga and
Gode are smoother and less extreme than for rural districts
urban livelihoods do not
depend directly on variable rainfall, and urban residents
enjoy more stable incomes over
time. Jigjiga and Gode towns are also the only two samples
where more people admit to
doing well than struggling. In every single district, though
(both rural and urban), more
people claim to be struggling today than ten years ago, which

could be a nostalgia effect
but could also reflect increasing difficulties in maintaining a
viable livelihood.
Premature death is the most extreme outcome of
vulnerability, and data collected on
mortality in our survey confirms other patterns of
vulnerability across and within households,
and over time. Rural households suffer almost twice as many
deaths as urban households,
and four times as many children under five die in rural than
in urban households. The
number of deaths recorded since 1990 peaked in the drought
years of 2000, 2001 and
2004. There is also a strong seasonal effect, with two-thirds
of deaths occurring during the
dry season months, when times are hardest. Although infant
and child mortality rates across
the world are directly correlated with poverty, the influence
of seasonality and drought on
mortality in Somali Region suggests that more complex
determinants of vulnerability also
need to be considered.
The following four chapters consider different causes of
vulnerability, starting with drought,
which was identified as the number one risk to livelihoods in
all rural areas. Many people
believe that droughts are more frequent than in the past, but
analysis of long-term rainfall
data does not support this perception. A marginal decline in
rainfall is observed in northern
Somali Region since the 1950s, but in the much drier central
part of the region average
rainfall has actually increased since the 1970s. The defining
characteristic of rainfall in Somali
region is its variability from year to year, and there is no
evidence that the recent sequence
of localised droughts represents a permanent decline in
average rainfall. The impact of
drought on rural livelihoods in Somali Region is
multidimensional. Pastoralists lose their
livestock, their main source of wealth, while farmers and
agro-pastoralists lose their harvest,
their main source of food. Access to food is further
undermined by falling barter terms of
trade, as livestock prices fall while food prices rise. Other
livelihoods that depend on
pastoralist and agricultural incomes, such as traders and
service providers, also report facing
declining incomes in drought years due to falling demand
from pastoralists and farmers for
their goods and services. On the other hand, droughts are part
of the natural cycle in semi-
arid areas, and local livelihoods are sensitively adapted to the

certainty that drought will
come. If vulnerability to drought is increasing, the reasons
have to do with declining ability
to cope, rather than increasingly frequent or abnormally
severe drought events.
14
IDS
RESEA
RCH
REPOR
T 57
Closely related to drought is access to water, and this study

identifies four distinct water
systems in Somali Region. (1) In the arid eastern districts,
people and livestock depend almost
exclusively on berkad (constructed water reservoirs),
especially during the dry season. Many
of these berkad are privately owned, and vulnerability can
arise from high prices or restricted
access to this water. (2) Where adequate groundwater is
available, communities dig shallow
wells which become their main water supply after rainy
season ponds evaporate. These wells
are communally managed and no charges are levied, though
competition can occur over
queuing and access rights. (3) For farming and agropastoralist communities living alongside
the Shabelle and Dawa/Ganale, these rivers are the primary
source of water for people and
livestock. The main problem is the health risk of drinking
untreated surface water that is
shared with animals. (4) Urban centres such as Jigjiga and
Gode towns enjoy virtually the
only access to piped water (from taps and boreholes) in the
region. Unlike rural water
supplies, piped water is clean, cheap and reliable all year
round.
After drought, conflict is the second major source of
vulnerability examined in this report.
Conflict and its consequences in Somali Region has a long
history and takes many forms,
from the waves of refugees during the Ogaden war with
Somalia in the late 1970s and
Somalias civil war in the early 1990s, to the ongoing
struggle of the ONLF for greater
autonomy. On a smaller scale, border disputes with Oromiya
Region created another wave
of displacement around a referendum held in 2004, and
violent deaths occurred around the
regional elections in August 2005. Away from overtly
political conflicts involving armies and
militia groups, frequent disputes occur between clans or subclans over access to resources
land for farming, grazing and water for livestock. Although
indigenous mechanisms of
conflict resolution do exist, if relations break down these
disputes can escalate into violence.

An indirect effect of conflict is aabsi (fear of conflict),
which results in migration routes
being disrupted and large tracts of contested arable or grazing
land lying unutilised for years.
In Shilabo District, the potentially lucrative Calub natural gas
deposits remain unexploited,
because this is an unstable area and control over these
resources is disputed between local,
national and international stakeholders.
Good governance reduces livelihood vulnerability in many
ways notably, by maintaining
peace and security, promoting trade and economic activity,
and providing effective social
protection or safety nets. Unfortunately, the Somali regional
government is weak and
ineffective, and the actions of government agents
(confiscating basic goods from traders, or
withholding food aid from IDP camps) often exacerbate
vulnerability. Respondents to our
household survey feel that government is ineffective and that
their interests are
inadequately represented at all levels local, regional and
federal. Governance in Somali
Region is complicated by a complex relationship between the
state and traditional
institutions such as the Gurti, a regional council of clan
elders. It is not yet clear whether the
ongoing process of decentralisation will enable rural
communities to articulate their
priorities and hold officials and politicians to account, or will
only replicate divisive interest
group politics at the local level.
Vulnerability in Somali Region is highly gendered.
Demographic evidence comes from the
1997 population census and our survey of 1,100 households
both find that sex ratios are
heavily skewed towards males. Among the under-25 age
cohort, the census enumerated
300,000 more males than females, while older men (over60s) outnumber older women by
2 to 1. Girls are more likely to die young than are boys, and
men in Somali Region live
longer than women. Qualitative survey data suggest that
parents favour sons over
daughters, notably in terms of intra-household allocation of
food and access to health care.
Although women have less authority within the home and are
often channelled into lowincome, low-status livelihood activities outside the home, this
is changing. As drought has
decimated livestock herds, so mens status has diminished
and many are reduced to chewing
khat (a mild stimulant), leaving their wives to find income

and food for the family. As their
autonomy and responsibility increase, so womens aspirations
are rising, and many are
looking forward to a brighter future for their daughters.
15
Responses to vulnerability
The next section is titled responding to vulnerability, and
includes chapters on household
coping strategies, informal transfers, formal safety nets
(mainly food aid), health and
education services. Our survey finds that households faced
with a livelihood shock in Somali
Region adopt a similar range of coping strategies to rural
households elsewhere in Africa,
and in the same sequence. Rationing of food consumption is
an immediate and almost
universal response to shortage, as it is costless and easily
reversible. The second most
popular response is to call on support from relatives, which is
apparently more available than
in other African countries, perhaps because of the range of
traditional institutions of mutual
support in Somali culture. There are mechanisms for
redistributing food (grain, meat, milk,
cooked meals), cash (including remittances and zakaat),
animals, seeds and labour, either on a
reciprocal basis or from wealthier to poorer households.
These are vital mechanisms for
pooling and reducing risk, but many people complain that
they are declining in response to
recurrent shocks and social change the perception is that
people are both less able and
less willing to help each other than in the past. The circle of
responsibility may be
shrinking to ones immediate family, raising the vulnerability
of those who are excluded from
(declining) community support mechanisms, but have no
close relatives to assist them.
Formal safety nets in Somali Region are limited to deliveries
of emergency food aid, which
have risen dramatically since the late and inadequate
response by the government and
donors to the drought-famine of 19992000. More than a
quarter of the regions
population has been declared in need of food aid every year
since 19992000, and over a
million tons of grain (mostly imported wheat) has been
delivered. At the aggregate level,
this has been enough to depress local food prices creating
disincentives for local traders
and farmers but for beneficiary households the amount of
food actually received is trivial.
Much food aid is misappropriated or mis-targeted. Since
Somalis dislike the taste of
American wheat, tons of food aid is also sold on local

markets. Food aid policy in Somali
Region urgently needs to be reconsidered. Food aid deliveries
are irregular, unpredictable,
inappropriate and ineffective. Even its beneficiaries recognise
that food aid is doing more
harm than good, and are asking for different types of safetynet assistance (cash transfers,
livestock restocking), or even for developmental needs to
be prioritised instead (support
for livelihood diversification, investment in education and
health services). New thinking and
innovative approaches to social protection in Somali Region
are needed.
The poor quality and limited accessibility of health services
in Somali Region is demonstrated
by high numbers of preventable deaths due to untreated
illness, inadequate immunisation
coverage, malnutrition and hunger-related disease, and deaths
of mothers and infants in
childbirth. Almost one in four deaths recorded in our survey
occurred in childbirth, an
indictment of the lack of reproductive health care in the
region. Health clinics are extremely
rare in rural areas outside district capitals, and pastoralists
have to travel longer distances
than agro-pastoralists or farmers to their nearest clinic. One
solution to low population
density and high population mobility is to deliver services
through mobile facilities, and there
are mobile vaccination teams that have immunised one in
four children in our survey.
Significantly, the lowest immunisation rates were recorded in
the districts that are most
unstable in our survey, which reveals another way in which
conflict and instability raises
vulnerability it is bad for childrens health.
Education indicators in Somali Region are appalling. In our
survey, fewer than one person in
five over 15 years old can read and write, and in rural areas
only 13 per cent are literate.
Literacy is highly gendered men are three times more likely
to be literate than women. In
rural districts, female literacy ranges from 7 per cent to just
1.2 per cent. There is no
difference between settled farming communities and mobile
pastoralists, despite the
governments argument that sedentarisation will facilitate the
delivery of basic services. The
education deficit in Somali Region has both supply-side
and demand-side determinants.
Access to education is constrained by the limited number of
schools and teachers (apart
from informal Koranic teachers) in rural areas. Our

respondents are also very critical of the
poor quality of education provided, especially in rural areas.
On the demand side, mothers
express a stronger interest in educating their children,
especially daughters, than do fathers.
16
IDS
RESEA
RCH
REPOR
T 57
Some parents see little need for pastoralists to be educated,

but many want their children
to have the option to pursue alternative livelihoods, and
mothers believe that education will
give autonomy to their daughters. Some communities are
even raising money to construct
school buildings themselves. This makes a powerful case for
extending access to good quality
education to all children in Somali Region, as part of a
strategy to expand livelihood options.
The regional government is making concerted efforts to
increase education infrastructure
and improve the delivery of services in Somali Region
mobile education services are being
trialled, for instance and these efforts must be supported.
Our research concludes that livestock-based livelihoods in
Somali Region are relatively
lucrative but extremely vulnerable (the paradox of wealth
plus vulnerability), that this
vulnerability is complex and multidimensional, and that the
ability of families and
communities to cope is being eroded and undermined by
multiple interacting stresses. The
ability of the Somali Region economy to generate wealth is
not in doubt. It is the variability
of income generation and asset ownership, and the
consequent vulnerability of groups of
people within the region, that must be addressed.
While the government is putting considerable energy into
new approaches and systems, it
has yet to find effective solutions to Somali Regions
complex natural and political
vulnerabilities. Such solutions will be found not in isolated
projects or programmes, but in
the wider environment and policy spaces within which people
construct their livelihoods.
Facilitating what people are already doing requires building
on the regions assets, as well as
working to reduce risk. Trade, for instance, is arguably the
most lucrative source of income
across all sectors of the regional economy, but legal
constraints are stifling trade rather than
maximising the potential of internal and cross-border trade to
generate both household
incomes and government revenue.
Addressing vulnerability also requires recognising the

linkages between the productive (or
economic) and reproductive (or social) sectors. For
example, investing in education and
health services, often undervalued in the past, is now
recognised as an essential investment
in the productivity and capabilities of people, which expands
their opportunities to diversify
livelihoods and spread risk. Conflict resolution is another
area that is often characterised as
political, but one which has profound implications for
livelihoods, in multiple ways.
Cycles of accumulation, collapse and rebuilding are defining
features of the pastoralist way
of life, but recent shocks and stresses may be stretching
coping capacities in Somali Region
to breaking point. As the capacity of communities to support
their members declines due
to processes of social change as much as economic shocks
so the role of formal social
protection interventions increases. But food aid is unpopular
and is damaging to farmers
livelihoods. Better designed, more innovative and effectively
implemented social protection
is vital, not just to provide a safety net in times of crisis, but
to provide reassurance that the
safety net will be there when needed. Above all, supporting
viable livelihoods in Somali
Region requires planning for unpredictability, expanding
peoples options, supporting their
cooperation and maximising not restricting their physical,
economic and social mobility.
17
18
IDS
RESEA
RCH
REPOR
T 57
Section 1 Context
1 Introduction
This chapter explains the motivation for this study, and the
methodology used in the fieldwork.
1.1 Research issues

In 19992000 and again in 2004, Somali Region in southeastern Ethiopia was struck by
severe droughts that resulted in numerous deaths of people
and livestock, as well as the
destitution and displacement of many pastoralist families.
Though less internationally visible
than the media famines that afflicted Ethiopia, including
Somali Region, in the 1970s and
1980s, these were the latest in a series of devastating
drought-triggered livelihood shocks.
Some observers believe that the current sequence of low
rainfall years constitutes a
permanent decline in rainfall, and some are even predicting
the end of pastoralism in the
Horn of Africa. These assertions are explored in this report. A
more fundamental question,
however, is whether the adaptation mechanisms that
pastoralists have evolved in response
to the predictable threat of drought are collapsing, and what
factors are increasing their
livelihood vulnerability.
Drought may be the most persistent and comprehensive risk
that pastoralists face to their
lives and livelihoods, but the people of the Horn have
adapted their way of life over many
centuries to cope with erratic rainfall moving with their
animals across vast distances,
negotiating access to pasture with neighbouring clans,
establishing sophisticated trading
systems that provide access to lucrative markets in the Gulf
states while spreading risk
among numerous market agents. As a result, pastoralists and
traders in the Horn of Africa,
despite being politically marginalised and living in a highly
vulnerable drought- and conflict-
prone region, have become relatively wealthy.

In recent years, however, a new set of risks has threatened the
economic viability of
livestock-based livelihoods in Somali Region. Government
policies, both domestic and by
other countries, have disrupted trade and blocked vital flows
of cash and commodities.
These policies include a ban by Saudi Arabia on livestock
imports from the Horn of Africa,
and a war on contraband (informal cross-border trade) by
the Ethiopian Government. The
Somali Region economy, being oriented outwards towards
the Middle East rather than
inwards towards Addis Ababa, has few coping strategies in
response to these
unpredictable policy shocks.
This combination of natural stresses and policy shocks
strikes at the two key
components of the pastoralist economy livestock
production and livestock marketing,
respectively and together they have raised the inherent
vulnerability of rural livelihoods in
Somali Region to unsustainable levels. Many pastoralists
have abandoned their way of life,
having lost all their livestock, and fled to towns or internally
displaced person (IDP) camps,
where the men chew khat while women struggle to scrape a
living selling firewood. Many
traders have been forced out of trading, and many businesses
have closed down.
In this context, fundamental questions have been raised about
the future of pastoralism in
Ethiopia. The government has articulated its view that
pastoralism is increasingly untenable,
and that sedentarisation is the only sustainable long-term
option. Opposing this view,
supporters of pastoralism argue that livestock herders and
traders will survive their current
difficulties by applying their ingenuity to negotiate ways
around the new hazards they face,
as they have always done in the past. A third way forward
would be for the government
together with the people of Somali Region to explore
innovative options, which could
involve moving towards a more diversified mixed economy.
This discourse around the future of pastoralism in Ethiopia is
largely rhetorical and uninformed
by empirical evidence. Very few research studies have been
conducted in Somali Region even
the last population census was inadequate and had to be done
twice. The present study is the
19
most extensive to be undertaken in the region for many

years. The overall objective of this
research is to improve the understanding of livelihood
vulnerability in Somali Region, among
people in the region as well as among those in federal
government and international agencies,
in order to contribute to a more informed debate and
improved policy-making.
1.2 Structure of this report

This research report is structured around the analysis of
livelihood vulnerability in Somali
Region. It is divided into five sections.
Section 1: Context provides an introduction to the study.
Chapter 1 describes the
methodology, while Chapter 2 introduces Somali Region in
terms of its demographic
structure, and sets the region in national context by
comparing indicators of poverty across
regions in Ethiopia.
Section 2: Livelihoods examines how the people of Somali
Region make a living. Chapter 3
provides brief district livelihood profiles for each of the 11
districts surveyed in this research.
Chapter 4 discusses pastoralist livelihoods livestock rearing
and marketing focusing on
livelihood risks presented by shocks such as droughts, Saudi
import bans, and Ethiopian
border closures. Chapter 5 discusses crop farming and agropastoralism in Somali Region.
Chapter 6 considers the situation of a particularly vulnerable
category of people internally
displaced persons (IDPs) through interviews conducted in
Hartisheik IDP camp. Chapter 7
presents household survey data on cash incomes and
inequality, and livelihood activities.
Section 3: Vulnerability looks more closely at causes and
outcomes of vulnerability in Somali
Region. Vulnerability outcomes presented in Chapter 8
include mortality statistics and trends
in self-assessed vulnerability by survey respondents. Causes
of vulnerability are divided into
three clusters and examined in Chapters 9 to 12:
environmental (drought), political (conflict,
also failures of governance and representation) and sociocultural (gender bias).
Section 4: Responding to vulnerability disaggregates
various approaches to addressing livelihood
vulnerability into formal and informal responses. Chapter 13

analyses the range of coping
strategies that people in Somali Region adopt when faced
with drought or other livelihood
shocks. Chapter 14 elaborates on one category of coping
strategies, namely informal support
provided by relatives and friends. Chapter 15 switches the
focus to formal transfers provided by
the government and non-governmental agencies (NGOs),
including food aid and other safety
nets. Finally, Chapter 16 and Chapter 17 examine the
provision of health and education
services in Somali Region the serious inadequacy of these
public services is seen as
contributing to raising immediate and long-term vulnerability
in several ways.
Section 5: Conclusions summarises the findings of this
research and offers some implications
for policy.
1.3 Methodology
This section summarises key features of the methodology: the
research methods that were
designed and implemented (a household questionnaire
survey, a trade and marketing survey
and qualitative research), the sampling frame, the conduct of
the fieldwork, and the
consultation process and workshops held during the study.
1.3.1 Research methods

A combination of qualitative and quantitative methods was
developed to undertake this
research. Because so little rigorous fieldwork has been
conducted in Somali Region,
estimating incomes, livestock ownership, mortality and other
quantifiable indicators of
household well-being was an important objective of this
study. On the other hand, the
centrality of conflict, politics, gender and social relations as
determinants of well-being
outcomes necessitated a qualitative and contextual analysis,
as well as an understanding of
trends in vulnerability over time.
20
IDS
RESEA
RCH
REPOR
T 57
The quantitative and qualitative research instruments were

designed in collaboration with
the fieldwork teams in two week-long training sessions in
Jigjiga and Addis Ababa, during
which the methods were discussed, pilot tested in rural
communities, and revised several
times before being finalised. After the first leg of the
fieldwork (in three districts),
preliminary findings were presented at workshops in Jigjiga
and Addis Ababa, where further
issues were raised that were incorporated into a modified
research design for the remaining
legs of the fieldwork.
Table 1.1 Summary of household questionnaire
A.1
Household
profile:
household
structure
(monogamous/polygamous, male- or femaleheaded); sex, age, education and labour capacity of
household members; present /absent
in last week.
A.2 Deaths in the household: since 1991, by age, sex,
year/season and cause of death.
A.3 Child immunisation: BCG, polio, DPT, measles.
B Household livelihood activities: undertaken by each
household member in last 12 months
(from a list of over 65 activities identified during
training and pilot testing); income
earned; livelihood problems; changes in livelihood
activities (past to present, deviation
from parents).
C Crop farming: land ownership and access; water
sources/irrigation; use of
fertiliser/manure; most recent harvest (crops grown,
consumed and sold); comparison
with previous harvests.
D Livestock: access to pasture, water, veterinary services and
drugs; livestock owned
(camels, cattle, sheep, goats, donkeys); changes in
ownership in last ten years; reasons
for changes.
E Water: sources of water for household use and livestock, by
rainy/dry seasons.
F Health and education: nearest clinic/school; school
attendance by children.
G.1 Informal transfers: receipt of remittances, zakaat, other
support from relatives/
community.
G.2 Access to support networks: sources of support (cash,
food, work).
G.3 Zakaat: most recent payment and/or receipt of zakaat.
H Formal transfers: assistance received from government or
aid agencies in last three years
(food aid, food/cash-for-work, faffa (supplementary food
for young children), credit,
seeds/tools, livestock).
I Coping strategies: behavioural adjustments adopted during
livelihood shocks (e.g. drought).
J Impacts of conflict: loss of life, livestock, access to
grazing/water due to conflict since
1991; beneficiaries of recent conflict; conflict resolution
agents; compensation paid
and/or received; fairness and effectiveness of
government representation at
local/regional/national levels.
K Food security: meals per day (adults/children) during most
recent hungry season.
L Clothing and basic items: purchases of new/used clothes in
last three years for
adults/children); presence of basic groceries (salt, sugar,
tea, kerosene) in house at time
of interview.
M Self-assessment of household situation: self-reported
livelihood status (on a 4-point scale)
at the time of interview, for each of the last three years,
and ten years ago.
O Food consumption: food items consumed in the last week,
and source of each food item
(production, purchase, food aid, gift).
21
Household questionnaire survey

The household survey questionnaire was 17 pages long,
divided into 18 sections and
translated into Somali, partly as a reference guide for the
enumerators, and partly for
distribution to officials and elders in the field. The household
survey team, comprising a
supervisor and four or five enumerators, administered this
questionnaire to 1,100 households.
Each interview lasted, on average, 4550 minutes. The full
questionnaire is annexed; Table 1.1
provides a summary.
Qualitative research
The qualitative research team consisted of a Somali
specialist, a gender specialist, a specialist
in participatory methods and a translator. The issues explored
in depth by the qualitative
team were wide-ranging, and included the following:
1. Informal institutions: trends in traditional mechanisms of
social assistance, conflict
resolution and natural resource management; clan
structures, Dia clan groups and
elders (the Gurti).
2. Community-level inequalities: local social structures;
inclusion and exclusion of different
groups within communities from key decision-making
processes (e.g. food aid
allocations).
3. Formal political structures: activities and functions of
regional and local government,
process of decentralisation, local provision of services
(education, health, veterinary,
food aid).
4. Livelihoods: livelihood strategies, trends and constraints;
activities of rich and poor
women and men; impacts of drought and conflict; future
of pastoralism, migration and
urbanisation.
Standard qualitative research methods were used in this
fieldwork: semi-structured
community discussions and focus group discussions, key
informant interviews and individual
case studies. Several participatory methods were also used to
generate information or
facilitate discussion, including community mapping and
wealth ranking, seasonal calendars

and historical time-lines.
The gender specialist on the team conducted community
discussions, focus group discussions
and life history interviews with women, men and children on
gender-specific topics,
including:
1. Livelihoods: womens work versus mens work;
women traders; gendered
ownership of livestock and property; drought impacts
and gendered coping strategies;
gendered access to resources (credit, food aid, etc.);
aspirations for childrens future
livelihoods.
2. Gender roles: intra-household (decision-making power,
spheres of responsibility); within
the community (conflict resolution, community
committees, savings groups);
consequences of being widowed or divorced; trends
over time (impacts of
modernisation, conflict, etc.).
3. Education and health: gendered access for sons and
daughters; adult literacy; HIV/AIDS.
It must be emphasised that most of the gender work was
conducted in Gashamo, in a
deliberate effort to capture experiences and responses to the
recent drought which had its
epicentre in that district. The views expressed by the women
of Gashamo reflect the
exceptional stress they were facing at the time, and should
not be interpreted as
representative of women in Somali Region as a whole.
22
IDS
RESEA
RCH
REPOR
T 57
Trade and marketing survey

This fieldwork team consisted of a livestock-marketing
specialist and two assistants. In
addition to the mobile team, price monitors were hired and
stationed at four major
markets for three months, collecting prices and volumes
offered and sold on a daily basis.
Among the topics explored in interviews with traders,
marketing agents and bureaucrats, as
well as through direct observation, were the following:
1. Major trade routes: volumes of commodity flows
(livestock, livestock products, cereals, other
food, khat, charcoal, clothes); significance of ports
(Berbera, Bosaso) for export markets.
2. Mapping of main markets: location (geo-referenced using
global positioning system
(GPS) units); commodities traded, market frequency
(daily, weekly); trade volumes (per
market day, monthly or per annum).
3. Evolution of trade routes: effects of border closures,
export bans, droughts, clan
disputes; changes in transportation; emerging markets
and new trade corridors.
4. Constraints to trade and traders: economic (lack of credit
or working capital);
bureaucratic (border controls, export bans, taxation);
entry barriers (cartels, male/clan
monopoly); conflict.
5. Marketing margins: transactions costs (transport, taxes,
agent fees); the roles of
marketing agents; price differentials, seasonality, and
inter-annual variability.
6. Market prices: daily monitoring of livestock and cereal
prices in four major markets
(Jigjiga, Gode, Hartisheik, Tog Wachale) in Somali
Region.
1.3.2 Sampling
Generating robust empirical statistics for Somali Region
requires a statistically representative
sampling frame, which implies relating the sample size to the
regional population, and
applying stratifying rules for disaggregating the regional
population into representative subsamples. Unfortunately, both these prerequisites are

hampered by information and access
constraints. First, there are no credible estimates of the
population of Somali Region, either
in aggregate or at sub-regional (zonal, district, rural and
urban) levels. The most recent
census was conducted in September 1997 (FDRE 1998), after
the data collected in this
region during the 1994 National Population and Housing
Census were rejected as unreliable.
The 1997 re-census was equally controversial, and although
it represents the best
approximation of the population of Somali Region currently
available to researchers and
policy-makers, it provides no basis for constructing a credible
sampling frame.1 Second,
difficulties of accessing several districts within Somali

Region, either because of security
concerns (no-go areas controlled by armed militia) or
because of logistical constraints
(lengthy travelling distances in difficult and sparsely
populated terrain), makes implementing
a random survey problematic and sometimes dangerous.2
It follows that it is not possible to implement a sample survey

of sufficient size and randomness
to be statistically representative of the entire region.
Accordingly, the statistics presented in this
report should be regarded as indicative of conditions
observed in the households surveyed,
1 Re-census statistics for Somali Region remain contested due to the lack
of clarity around regional boundaries
(Somali has several contested borders with neighbouring Oromiya
and Afar regions), the high mobility of the
population, both within the region and across international borders
(with Somalia to the east and Kenya to
the south), security concerns which prevented enumeration of eight
rural kebeles (in Shinile, Fiq, Gode and
Afder Zones), and anecdotal reports that enumerators made no effort
to visit rural areas outside district
capitals. (On a field trip to Guradamole District in May 2004, we
were told that the Census figure of around
3,000 residents was under-enumerated by at least a factor of 10, and
that the actual population as of 2004
might be in excess of 100,000.)
2 Even after fieldwork started for this project, reports of violent clashes
caused us to substitute two districts
included in the original sampling frame, out of concern for the safety
of our fieldwork team.
23
Table 1.2 Sampling frame for the household survey in Somali

Region
Livelihood System Area District Zone Food economy zone
Pastoralism North Shinile Shinile Shinile Pastoral FEZ
North
Gashamo
Degahbur
Lowland/Hawd Pastoral FEZ
Central Shilabo Korahe Lowland/Hawd
Pastoral FEZ
Agro-pastoralism North Kebribayah Jigjiga Jigjiga Agropastoral FEZ
Central Doboweyn Korahe Korahe
Agro-pastoral FEZ
South Cherati Afder Liban/Afder Agropastoral FEZ
Crop farming North Jigjiga (rural) Jigjiga Jigjiga Sedentary
Farming FEZ
Central Kelafo Gode Shabelle Riverine
FEZ
South Dolo Odo Liban Dawa/Ganale
Riverine FEZ
Urban North Jigjiga town Jigjiga n/a
Central Gode town Gode n/a
IDPs North Kebribayah Jigjiga (Hartisheik IDP camp)
rather than being representative of Somali Region as a
whole. For instance, no attempt is made
to generate internationally standardised estimates of poverty
headcounts for the region.
Given these methodological constraints, a sampling strategy
based on livelihood systems was
designed, rather than a stratified random sample based on
population size at zone or district
level. Following the food crisis of 19992000, when lack of
adequate information was
identified as one reason for the delayed humanitarian
response, Save the Children UK (SCUK) and the regional Disaster Prevention and Preparedness
Bureau (DPPB) divided Somali
Region into 15 food economy zones (FEZ).3 Of the 15 FEZ, six
are categorised as pastoralist
zones, six as agro-pastoralist zones, and three as crop farming

zones. To ensure adequate
coverage of these three livelihood systems in our survey,

three pastoralist, three agropastoralist and three farming food economy zones were
purposively selected. A second
criterion was to disperse the survey sites as widely across the
region as possible, so Somali
Region was divided into four geographic areas north,
central, south and east and food
economy zones were selected from each area. Significant
differences within livelihood
systems are also reflected in this selection. Among the
pastoralist sub-samples, Gashamo
and Shilabo are both camel-dominated and Shinile is cattledominated. Among the farming
communities, Kelafo and Dolo Odo are riverine and Jigjiga is
rain-fed.
Apart from rural livelihood systems, two other population
groups were included in the
sampling frame: urban residents and IDPs. About 14 per cent
of Somali Regions approximately
four million inhabitants live in urban areas, and the two
biggest towns in the region Jigjiga
and Gode were purposively selected. In May 2004 Somali
Region had 88,000 people who
had been displaced from their homes, lost their livelihoods,
and were living in camps (79 per
cent drought-induced and 21 per cent conflict-induced).
Instead of a quantitative survey, more
than 20 in-depth life history interviews were conducted with
IDPs in Hartisheik camp in
Kebribayah District. Plans to conduct interviews in other IDP
camps had to be abandoned due
to security problems. In Denan IDP camp, for instance,
violent clashes caused four deaths in
the week before the survey team was due to conduct
interviews in the camp.
The nine selected food economy zones are listed in Table 1.2,
along with the two urban
samples and the IDP camp where interviews were conducted.
This sample covers seven of
Somali Regions nine administrative zones, as well as the two
major urban centres and the
largest IDP camp.
3 A food economy zone (FEZ) is defined as a group or population that
obtains its food and income sources
from a broadly similar combination of means and that have similar
response to shocks (SC-UK 2001a: 6).
24
IDS
RESEA
RCH
REPOR
T 57
1.3.3 Fieldwork
Three teams of researchers conducted the fieldwork: (1) a
household survey team; (2) a trade
and marketing survey team; and (3) a qualitative research
team. Because of the distances and
logistical complexity of conducting fieldwork in Somali
Region, as well as constantly evolving
security concerns in different districts, the household survey
was conducted in four legs over
a period of eight months, as follows:
First leg (North 1): Gashamo (pastoralist);
Kebribayah (agro-pastoralist); Jigjiga town
(urban) (OctoberNovember 2004).
Second leg (Central): Kelafo (farming); Shilabo
(pastoralist); Doboweyn (agro-pastoralist);
(FebruaryMarch 2005).
Third leg (South): Cherati (pastoralist); Dolo Odo
(farming); Gode town (urban) (April
2005).
Fourth leg (North 2): Shinile (pastoralist); Jigjiga
plains (farming); Hartisheik (IDPs) (May
2005).
Within each of the 11 rural and urban sub-samples, 100
households were interviewed from
several randomly selected sites within the district. (The
procedure for identifying households
in pastoralist areas was to divide the district into four blocks,
to set up base camp in the
centre of each block and from there to drive in a different
direction each day until a
settlement or rer was encountered, where five households
would be randomly approached,
until 100 questionnaires were completed. In settled
communities, discussions with
knowledgeable elders and district administrators generated
lists of villages, which were
stratified to reflect the clan composition, population
concentrations and livelihood diversity
within the district.) Given the difficulty of obtaining accurate
estimates of total population
by food economy zone, a sample of 100 is large enough to
make general statements about
conditions in that FEZ or town without, however, claiming

statistical representativeness.
The total sample size for the household survey, therefore, was
1,100, comprising 300
pastoralist households, 300 agro-pastoralist households, 300
farming households and 200
urban households.
The qualitative research team travelled to most of the
communities where the household
survey team was working, to ensure comparability of
findings. However, to deepen our
understanding of the impacts of the 2004 drought on
pastoralist livelihoods and on gender
relations, the qualitative team spent more time in Gashamo
District, the epicentre of the
drought. The trade and marketing survey team did not
accompany the other teams, but
interviewed traders and monitored markets along trade routes
throughout Somali Region,
as well as across the border in Somaliland and Kenya, and in
Addis Ababa.
1.3.4 Consultation processes and workshops

The principal Government of Ethiopia partners for this study
are the Ministry of Federal
Affairs and the Regional Administration of Somali National
Regional State (SNRS). Key
officials in Addis Ababa and Jigjiga were briefed and
consulted about the project before it
started, and were kept fully informed at all stages. Since the
study was undertaken on behalf
of the people of Somali Region, the Gurti and other
representatives of the clans and
districts of the region were also consulted at regular intervals.
In the inception phase
between January and August 2004, many meetings were held
in Jigjiga town and a series
of scoping visits was undertaken throughout the region, to
sensitise local people to the
study and to gather baseline information.
A consultation workshop, titled Food Insecurity and
Livelihood Vulnerability in Somali Region,
was held in the Regional Government Administration Hall in
Jigjiga town, on 12 February
2004, to introduce this project to relevant stakeholders in
Somali Region. This workshop
attracted over 60 participants and was attended by regional
government officials, local and
25
international NGOs, and representatives of all major clans

and zones in Somali Region. The
project was introduced to stakeholders in Addis Ababa in a
public seminar, organised by the
Pastoralist Communication Initiative (PCI) and held at the
Sheraton Hotel on 16 February
2004, titled Food Security Issues in Ethiopia: Comparisons
and Contrasts between Lowland
and Highland Areas (Devereux 2004). This seminar was
covered by the BBC Somali Service
and generated debate on Somali internet chat-rooms. A
follow-up workshop on Food
Security Issues in Ethiopia was held at the Sheraton Hotel on
12 May 2004, which focused
specifically on food security issues, policies and safety nets
in lowland areas.
After the first phase of fieldwork was completed, two reportback workshops were held to
present and discuss the preliminary findings, the first in
Jigjiga (Regional Government
Administration Hall, 19 January 2005), the second in Addis
Ababa (Ghion Hotel,
21 January 2005). Feedback from these workshops was used
to refine the methodology
for the second phase of fieldwork. A four-page Issues Paper
was also produced after the
first phase (published by United Nations Office for the
Coordination of Humanitarian Affairs
(UNOCHA), March 2005), and widely circulated in Addis
Ababa and Somali Region.
Policy issues arising from the research were discussed at a
donor round-table in Addis
Ababa on 25 April 2005. Presentations of findings were also
made at the UNOCHA and
Department for International Development (DFID) Ethiopia
offices in Addis Ababa, and at
the Institute of Development Studies (IDS) Sussex in
Brighton, in July 2005. Following
publication of this Research Report, a range of dissemination
and follow-up activities are
planned during 2006.
2 Somali Region: the

place and the
people
2.1 Population and demography in Somali
Region
According to the (contested) 1997 Census, Somali Region
consists of 9 administrative zones,
44 woredas (administrative districts), and 67 urban settlements.4
The regional population was
enumerated at 3.5 million in 1997, of whom 1.9 million were

male and 1.6 million were
female.5 The region is overwhelmingly rural even Jigjiga town,
the regional capital, had a
population of well under 100,000 in 1998, though the local

administration estimates that
this had risen to 150,000 by 2003. The level of
urbanisation is low, at 14.3 per cent.6
Household size, averaging 6.6 with a range from 5.3 in
urban Jigjiga to 8.6 in rural Korahe
Zone is larger than in highland regions, where nuclear
household units are the norm.
There are approximately 520,000 households in Somali
Region.
The largest zone is Jigjiga, with over 800,000 residents
enumerated in 1997, while the
smallest is Fiq, with 233,000 (see Table 2.1). Somali Region
is linguistically and religiously
homogeneous Ethiopian Somalis accounted for 96 per cent
of the enumerated population
in 1997, and 98.7 per cent of the population is Muslim
(FDRE 1998, Table 2.12, Table 2.17).
The average household size in our Somali survey is 8.1
(n=8,806/1,091). This is somewhat
higher than the Census average of 6.6 (though the Census did
record an average of 8.6
household members in one zone, as noted above). However,
it should also be noted that
people in Somali Region feel strongly that the local
population was systematically under4 Even these numbers are controversial: different maps of Somali Region
show a range of districts, from 44 to
51. The figure currently accepted as most accurate is 48 districts.
5 This sex ratio of 120 males per 100 females is unusual for Africa,
though whether this reflects systematic bias
or gender discrimination against females is unclear.
6 Since the definition of a town in Ethiopia is a settlement of 500+
people, and the smallest enumerated
town in Somali Region had just 590 inhabitants in 1997 (FDRE
1998: 15), even 14.3 per cent might be seen as
an overestimate of the true level of urbanisation.
26
IDS
RESEA
RCH
REPOR
T 57
Table 2.1 Population of Somali Region, 1997 (ranked by

zone population size)
Administrative zones Total Rural Urban
Number % Number % Number %
Jigjiga 813,200 23.6 657,309 80.8 155,891 19.2
Liban 476,881 13.9 432,062 90.6 44,819 9.4
Afder 358,998 10.4 333,251 92.8 25,747 7.2
Shinile 358,703 10.4 305,329 85.1 53,374 14.9
Gode 327,156 9.5 256,657 78.5 70,499 21.5
Warder 324,308 9.4 299,627 92.4 24,681 7.6
Degahbur 304,907 8.9 247,041 81.0 57,866 19.0
Korahe 242,276 7.0 205,050 84.6 37,226 15.4
Fiq 233,431 6.8 210,824 90.3 22,607 9.7
Somali Region 3,439,860 10.4 2,947,150 85.7 492,710 14.3
Source FDRE (1998, Table 2.1).
enumerated in the 1997 Census. If there is any substance to
this allegation, and assuming
our survey is both accurate and representative, then our
survey data would suggest that the
true population is 22.7 per cent higher than the official
Census figure, which would give a
total population for Somali Region in 1997 of 4,221,500.7
Households in our survey are largest in the pure pastoralist
district of Gashamo (with close

to ten members), while the smallest households are found in
the urban centres of Jigjiga
and Gode (7.1) and in Shinile (6.9) (Table 2.2). These
households are larger than is typical for
other regions of Ethiopia, such as the highlands of Amhara

and Tigray. The reasons for these
differentials are mainly socio-cultural: polygamy is practised
more among Muslim lowlanders
than among predominantly Christian highlanders, and
pastoralists tend to live in extended
families while crop farmers in Ethiopia tend to favour nuclear
households. Within Somali
Figure 2.1 Map of Ethiopia, showing Somali Region
7 This is based on a comparison of average household sizes in the Census

and our survey. Our sample is not
comprehensive enough to allow an estimation of the regional
population in 2004/5.
27
Table 2.2 Household size and composition, by district

District Household
Polygamous
size
Males
Females
Sex
Female-
ratio* headed
Pastoralist 8.4 1,321 (52.3%) 1,203 (47.7%) 110.5 5% 17%

Gashamo 9.7 492 (51.1%) 471 (48.9%) 104.5 8% 29%
Shinile 6.9 370 (53.7%) 319 (46.3%) 116.0 5% 5%
Shilabo 8.6 459 (52.5%) 413 (47.4%) 111.1 3% 17%
Agro-pastoral 8.5 1,237 (49.4%) 1,265 (50.6%) 102.0 4%
19%
Kebribayah 8.2 400 (50.5%) 392 (49.5%) 102.0 5% 19%
Doboweyn 8.9 423 (48.7%) 445 (51.3%) 95.1 3% 20%
Cherati 8.4 414 (49.2%) 428 (50.8%) 96.7 3% 19%
Farmer 8.0 1,254 (52.2%) 1,149 (47.8%) 109.1 4% 15%
Kelafo 8.5 429 (50.7%) 417 (49.3%) 102.9 7% 21%
Dolo Odo 8.3 440 (53.3%) 386 (46.7%) 114.0 1% 18%
Jigjiga rural 7.3 385 (52.7%) 346 (47.3%) 111.3 3% 7%
Urban 7.1 700 (50.8%) 677 (49.2%) 103.4 13% 6%
Jigjiga town 6.7 347 (51.6%) 325 (48.4%) 106.8 15% 4%
Gode 7.1 353 (50.1%) 352 (49.9%) 100.3 10% 8%
Total 8.1 4,512 (51.2%) 4,294 (48.8%) 105.1 6% 15%
* Sex ratio = males x 100/ females.
Source: Household survey data (n=1,091).
Region, urban households are typically smaller than rural
households. This pattern is found
the world over fertility rates are generally higher in rural
communities, for a variety of
economic and socio-cultural reasons.
Over 90 per cent of households in our survey are maleheaded, while less than 6 per cent
are female-headed (n=63/1,091 =8 per cent). Female-headed
households are an urban
more than a rural phenomenon in Jigjiga town, one
household in seven is female-headed,
and in Gode town one in ten, while in Dolo Odo the figure is
only one in a hundred.
Conversely, polygamy in Somali Region is more rural than
urban: while almost one
household in six in our sample is polygamous (n=167/1,091
=5.3 per cent), the figure is only
one in 25 in urban Jigjiga, but close to one in three in
Gashamo (29 per cent), and one in
five in Kelafo, Doboweyn, Kebribayah and Cherati (Table

2.2).
Sex ratios in our sample are biased towards males, with 105
males for every 100 females. Male
bias is highest in Shinile (at 114), and lowest in Doboweyn
(95) and Cherati (97), where females
Table 2.3 Males per 100 females in Somali Region, by age
cohort
Cohort Survey Census Cohort Survey Census
04 92.7 113.3 4044 88.8 103.3
59 116.5 123.8 4549 127.9 113.7
1014 129.8 136.3 5054 154.1 131.9
1519 117.1 142.7 5559 172.2 164.4
2024 92.9 127.8 6069 186.0 204.6
2529 60.3 85.2 7079 156.3 213.2
3034 78.2 78.9 80+ 118.2 173.1
3539 100.0 73.6
Source: Census and household survey data.
28
IDS
RESEA
RCH
REPOR
T 57
Table 2.4 Absentee household members, by district*

District Absentees Ethiopia Abroad Males Females
Pastoralist 206 (8.1%) 134 (65%) 72 (35%) 108 (52%) 98
(48%)
Gashamo 90 (9.2%) 42 (47%) 48 (53%) 46 (51%) 44 (49%)
Shinile 44 (6.4%) 28 (64%) 16 (36%) 24 (54%) 20 (46%)
Shilabo 72 (8.2%) 64 (89%) 8 (11%) 38 (53%) 34 (47%)
Agro-pastoral 223 (8.9%) 189 (85%) 34 (15%) 110 (49%)
113 (51%)
Kebribayah 55 (7.0%) 44 (80%) 11 (20%) 28 (51%) 27
(49%)
Doboweyn 87 (10.0%) 76 (87%) 11 (13%) 42 (49%) 45
(51%)
Cherati 81 (9.6%) 69 (85%) 12 (15%) 40 (49%) 41 (51%)
Farmer 205 (6.8%) 173 (84%) 32 (16%) 107 (52%) 98 (48%)
Jigjiga rural 104 (7.4%) 86 (83%) 18 (17%) 55 (53%) 49
(47%)
Kelafo 42 (5.0%) 37 (88%) 5 (12%) 21 (50%) 21 (50%)
Dolo Odo 59 (7.1%) 50 (85%) 9 (15%) 31 (53%) 28 (47%)
Urban 141 (5.8%) 118 (84%) 23 (16%) 72 (51%) 69 (49%)
Jigjiga town 104 (7.4%) 86 (83%) 18 (17%) 54 (52%) 50
(48%)
Gode 37 (5.2%) 32 (86%) 5 (14%) 19 (51%) 18 (49%)
Total 775 (8.1%) 614 (79%) 161 (21%) 397 (51%) 378
(49%)
* Absentees are defined as household members who were not
present in the household
during the week preceding the survey.
marginally outnumber males (Table 2.2). Interestingly, these
two female-biased districts both
have above average household sizes and polygamy rates, and
below average numbers of
female-headed households. Conversely, male-biased
Shinile has the smallest average
household size, the lowest polygamy rate and below average
female-headed households.
Disaggregating the population by age cohort reveals that sex

ratios vary significantly across
the generations. Comparing our sample survey with the
enumerated Census population
(Table 2.3) provides some validation of the demographic data
collected in our household
survey. Except for the youngest cohort (04 years old), both
sets of data display the same
pattern. Males outnumber females among children and
teenagers (519 years old), and then
there is a sharp reversal, with adult females outnumbering
males for the next two decades
(2039 years old). Among middle-aged adults (4059 years
old) there is a second reversal,
with males again dominant. Among older people (60 years
and above), the excess of males
over females becomes very large indeed: in the Census, there
are two men for every
woman over 60 years old. Our survey provides no clear
explanation for this pattern,
although gender bias could be part of the story (see Chapter
12).
Another explanation for this pattern may be differential
mobility between males and
females, with young to middle-aged men in particular being
more likely to travel for
extended periods than women. A total of 671 people were
temporarily absent from these
1,100 households at the time of the survey (Table 2.4). Most
absentees who remained inside
Ethiopia had travelled to other rural areas in Somali Region
(62 per cent), but a significant
number had gone to urban centres such as Jigjiga or Addis
Ababa (22 per cent). Absentees
from farming communities were more likely to remain in
rural Ethiopia, reflecting the
lower mobility of farmers compared to pastoralists. Of those
who were outside Ethiopia,
most (79 per cent) had travelled to neighbouring Somali
countries, while some were in the
Middle East (4 per cent) or in the West Europe or North
America (5 per cent). The main
destination for absentees from urban Jigjiga households was
other urban centres in
Ethiopia, especially Addis Ababa, while some were abroad in
Western countries. Gashamo is
29
Table 2.5 Demographic structures in developing countries

and in Somali Region
Age cohort Developing Somali Age group Developing
Somali
countries
Region
countries Region
04 years 12% 15% Children 35% 46%
59 years 12% 16% (014 years)
1014 years 11% 15% Adults 59% 52%
1519 years 10% 13% (1519 years)
2059 years 49% 39% Older people 7% 3%
60+ years 7% 3% (60+ years)
Male/female ratio 51 / 49 51 / 49 Dependency ratio 0.71 0.95
Source: Save the Children 2004: 25; Household survey data.
the only district where more absentees were in neighbouring
countries than elsewhere in
Ethiopia. Most of these people had been displaced by drought
to Somaliland, where they
had taken some of the familys livestock to negotiate access
to grazing and water. A
significant number of absentees were described as having
married into other households,
which might indicate a rise in early marriages as a droughtcoping strategy (the parents of
the bride receive cash or livestock as a dowry payment).8 A number
of children were also
away at boarding school at the time of our survey.

Interestingly, absentee household members are located
overwhelmingly (99 per cent) in
male-headed households. In 6 of 11 district samples, no
absentees at all come from femaleheaded households. This suggests that women who head
households face severely
constrained mobility, which is consistent with their domestic
reproduction roles.
Somali Region has a very young population, with close to
half (46 per cent) the individuals
enumerated in our survey aged under 15. This proportion is
high even by comparison with
other developing countries, where just over one-third (35
per cent) of the population are
under-15s (Save the Children 2004). Conversely, Somali
Region has a much smaller
proportion of older people (over-60s) than average just 3
per cent compared to 7 per
cent elsewhere. This implies that life expectancy in Somali

Region is abnormally low. It also
means that the crude dependency ratio (children + older
persons per active adult) is close to
one (0.95), higher than the developing countries average
(0.71) (Table 2.5).
Another implication of a very young population is that it
raises questions about the nature
and sustainability of future livelihood opportunities for the
next generation of Somali adults.
Without endorsing a simplistic Malthusian discourse
(whereby population growth leads
inevitably to unsustainable over-exploitation of the fragile
natural resource base), concerns
are already being raised about the adverse synergies between
recurrent (and possibly more
frequent and severe) droughts, and environmental
degradation associated with overgrazing
and charcoal burning. Together with anecdotal evidence
(reported in later chapters) that the
aspirations of Somali youth might lie increasingly outside
pastoralism, we might predict that
levels of urbanisation (within Somali Region) and migration
(to destinations outside the
region) will rise, while the proportion of Ethiopian Somalis
who remain engaged in
livestock-dependent livelihoods will decline.
2.2 The paradox of wealth plus vulnerability

Pastoralism can be a lucrative livelihood. When national
household income and consumption
surveys are conducted in Ethiopia, Somali Region often
emerges as the wealthiest (or least
poor) of all Ethiopias rural regions, with the highest
consumption levels and lowest poverty
incidence. In the 1999/2000 Household Income,
Consumption and Expenditure (HICE)
8 Although we were told that this was happening, sensitivity about
marrying off daughters caused this
question to be removed from the household survey questionnaire
after the pilot test.
30
IDS
RESEA
RCH
REPOR
T 57
Table 2.6 Poverty headcounts and calorie consumption in

Ethiopia, by region
Rank by Region* Poverty headcount (1999/2000) Calorie
intake
poverty
Rural poor Urban poor Total
poor (1995/6)
(%) (%) (%)
1 Somali 44 26 38 2,109
2 Oromiya 40 36 40 2,004
3 Amhara 43 31 42 1,957
4 Gambella 55 38 51 1,917
5 SNNPR 52 40 51 1,800
6 Benshangul-Gumuz 56 29 54 1,767
7 Afar 68 27 56 2,055
8 Tigray 62 61 61 1,902
Total 45% 37% 44% 1,938
* Urban-dominated regions and municipalities such as Addis
Ababa, Dire Dawa and Harar
are excluded from this table.
Source: FDRE (2002).
survey, the poverty headcount in Somali Region was only 38
per cent, lower than the
national rate of 44 per cent and considerably less than the
poorest region Tigray, with 61
per cent (Table 2.6). There is significant variation between
pastoralist areas within Ethiopia,
however. Afar Region, which is also predominantly
pastoralist, is the second poorest region,
with 61 per cent headcount poverty. The difference probably
reflects the higher involvement
of Somalis in livestock marketing and cross-border trade;
people in Afar are much less
engaged with markets.
Although income is a reasonable proxy for well-being, a
more direct indicator is food
consumption. In the 1995/6 HICE survey, Somali Region
again ranked at the top of the list,
with an average consumption per person of 2,109 kilocalories
per day (Table 2.6). Taking 2,100
kilocalories per person per day as a cut-off level for food
security, this suggests that Somali
Region was the only food-secure part of rural Ethiopia in

1996. Interestingly, Afar Region,
though second poorest in terms of household income, was
second most food-secure in terms
of food consumption, only 50 kilocalories behind Somali
Region. These statistics do not mean
that the average Somali household is wealthy in international
terms, but they do imply that
the average Somali pastoralist is better off than the average
crop farmer in highland Ethiopia.
Moving to food security outcome indicators, the picture
changes somewhat. Amhara
Region, which was mid-range in terms of both poverty and
food consumption, has by far
the worst record in a ranking of regions by proportion of
stunted children (65 per cent)
(Table 2.7). Somali Region is no longer first, but third, in this
ranking (with 48 per cent
stunting), behind Gambella and Afar which both have higher
poverty headcounts. Stunting
(the proportion of children under five whose height-for-age
falls below the average for a
reference population) is an anthropometric indicator of
chronic or long-term food deficits. It
is difficult to reconcile a figure of 48 per cent stunted
children in a population whose
average food consumption exceeds 2,100 kilocalories per
day, but it could suggest either
unbalanced diets (too little dietary diversity) or distortions
(e.g. gender or age bias) in the
intra-household allocation of food.
Child wasting (the proportion of children under five whose
weight-for-age falls below the
average for a reference population) is an indicator of acute or
short-term food insecurity. In
the year 2000, levels of wasting were highest in Gambella,
followed by Afar and Somali
the two pastoralist regions. So Somali had the third worst
levels of wasting in the country,
alongside Tigray, despite poverty in Tigray being 23
percentage points higher than Somali.
How can this paradox be explained? Why is Somali Region
apparently so wealthy, and at
the same time so vulnerable?
31
Table 2.7 Anthropometric indicators in Ethiopia, by region

Rank by Region Stunting Rank by Region Wasting
stunting
1999/2000 wasting
1999/2000
(%)
(%)
1 Gambella 40 1 Oromiya 8.8
2 Afar 42 2 SNNPR 9.1
3 Somali 48 3 Amhara 10.9
4 Benshangul-Gumuz 51 4 Benshangul-Gumuz 11.4
5 Oromiya 54 5 Tigray 11.7
6 SNNPR 57 6 Somali 11.7
7Tigray 59 7Afar 11.8
8 Amhara 65 8 Gambella 13.0
Total 57
Total 9.6
Source: FDRE (2002).
This study is a search for an explanation of this paradox of
wealth plus vulnerability among
Ethiopian Somalis. The next section of this report explores
the complex and diverse
livelihood activities and strategies that the people of Somali
Region have evolved to make
their living in a harsh and difficult environment. However, as
will be seen in later chapters,
the risks they face are not just environmental though it is
true that there have been
several droughts in the past few years, with 2000 and 2004
being particularly severe.
Survival in a marginal environment is made more difficult by
a variety of political risks,
which include conflict between and within clans over scarce
natural resources, government
policies that obstruct rather than facilitate trade, weak and
unstable governance in the
region, unpredictable shifts in the external policy
environment (e.g. the ban by Saudi Arabia
on livestock imports from the Horn of Africa), compounded
by the inadequate provision of
essential services (notably education and health), erratic
delivery of social safety nets (mainly
food aid), and a decline in traditional mechanisms of intracommunity support. Given this
extremely uncertain and insecure context, the apparent
contradiction between different
indicators of wealth and well-being in Somali Region

becomes easier to understand.
32
IDS
RESEA
RCH
REPOR
T 57
Section 2 Livelihoods
This section has five chapters. Chapter 3 presents livelihood
profiles for each of the 11
districts of Somali Region where the household survey was
conducted. The next two
chapters examine the dominant livelihood systems in this
region pastoralism (including
livestock marketing) and crop farming (including agropastoralism). Another significant
livelihood system is internally displaced persons living in
camps like Hartisheik, and IDPs are
discussed in Chapter 6. Finally, Chapter 7 analyses levels of
income and inequality among the
1,100 households surveyed.
3 District
profiles
livelihood
Contradicting the romantic vision of pastoralism as nomads

leading a solitary and selfsufficient existence, wandering through the desert and living
off the meat, milk and blood
of their camels, livelihoods in Somali Region are extremely
diversified. Most households
engage in a number of income-generating activities, and most
are active participants in
markets, especially for tradeable commodities. The livestock
off-take system generates
regular income, keeps capital flowing through the economy,
and allows households to

reconstitute their asset base after shocks. On the other hand,
most livelihood activities are
dependent on the natural resource base, and are therefore
vulnerable to covariate shocks
such as a drought, which undermines crop farming, livestock
production, agricultural labour
and herding opportunities simultaneously. Other livelihood
activities that are not derived
directly from the natural resource base depend on incomes of
people who are natural
resource dependent, and these livelihoods are vulnerable to
derived destitution (Sen 1981).
Examples include service providers to pastoralists barbers
and hairdressers, musicians and
poets, as well as traders.
A crucial feature of the Somali Region economy, easily
overlooked by analyses that focus on
pastoralism in isolation, is the interconnected nature of
different livelihood activities. Capital
flows around this system because pastoralists sell animals to
traders and buy food produced by
farmers and agro-pastoralists; relatives with jobs in urban
centres invest in the rural economy;
other relatives living abroad remit cash back to the region.
This dynamic and complex set of
Figure 3.1 Somali Region, with survey sites highlighted
33
economic relationships is a source of strength but also a

source of vulnerability, as any threat
to one set of actors in the system can undermine the
livelihoods of many others.
This chapter presents brief overviews of the livelihood
system in each of the 11 sub-samples
of the household survey. As explained in Chapter 1, the rural
survey was stratified by food
economy zone to ensure adequate representation of all major
livelihood strategies in Somali
Region, with fieldwork being conducted in three districts
dominated by pastoralists
(Gashamo, Shinile and Shilabo), three districts dominated by
agro-pastoralists (Kebribayah,
Doboweyn and Cherati), and three districts dominated by
crop farmers (Kelafo, Dolo Odo
and rural Jigjiga). In addition, two urban centres were
included in the household survey
(Jigjiga town and Gode town). Figure 3.1 shows the dispersal
of these 11 sites throughout
Somali Region, highlighting the districts and towns surveyed.
3.1 Pastoralist districts

Three districts in our survey are predominantly pastoralist:
Gashamo, Shinile and Shilabo.
3.1.1 Gashamo District

Gashamo District, in Degahbur Zone, falls within the
Lowland (Hawd) Pastoral Food
Economy Zone, and has been characterised as camel, shoats,
and Berkad [water reservoirs]
dependent (SC-UK, DPPB and partners 2001a: 1). The
Hawd is largely arid and there are
no rivers or other permanent water sources in Gashamo.
Livelihoods in Gashamo District
are almost exclusively pastoralist, being dominated by
livestock, and are very undiversified (in
our survey, only 1.3 activities were recorded per household).
Almost every household owns
some sheep and goats, while over half of the households
surveyed own camels although
this figure is lower than usual because of the recent droughts.
Camels are especially
important, both as a store of wealth and as a livelihood
resource, because camels are well
adapted to the Hawds harsh and drought-prone environment.
Very few cattle and donkeys
were recorded in our survey. Some cattle were kept in the
grassland plains around

Gashamo town, but almost all these animals died in the
drought of 2004.
After livestock rearing, the next four most common
livelihood activities are all livestock
related trading livestock or livestock products, selling dairy
products (milk, cheese, ghee
clarified butter used in cooking) or hides and skins from
owned animals. There are some
traders who trade foodstuffs, clothes and shoes, but no formal
employment and almost no
service providers, other than a few restaurant workers. But
these are minor activities
compared to the livestock-related occupations. Most business
in Gashamo is conducted in
Somalia Shillings, and most purchased food commodities
(rice, wheat flour, sugar, etc.) are
imported from Somaliland.
Given this reliance on livestock for household income and food
(meat and milk),9 it follows
that the main sources of livelihood risk are those that threaten
livestock production or
marketing. In recent years, the two major livelihood shocks
have been drought and livestock
export bans. First, in September 2000, the Gulf states
imposed an embargo on livestock
imported from Somali Region, which caused a massive loss
of earnings for households from
Gashamo that depend on livestock exports through the
Somaliland port of Berbera to Saudi
Arabia for most of their cash income. Second, although
Gashamo was relatively unaffected
by the drought of 1999/2000 (which had its epicentre further
south), poor Gu rains
(AprilJune long rains) in 2001 were followed by a very
serious Gu rain failure in 2004,
which did affect Gashamo particularly severely, causing
widespread livestock mortality and
the destitution of many pastoralist households.1 0
9 An assessment in 2001 concluded that milk consumption meet s as

much as 50 per cent of annual food
needs in Gashamo households (Save the Children UK et al. 2001a:
2).
10 The humanitarian response to this drought was slow, and in fact a
scoping visit to Gashamo for this research
project by the Pastoralist Communication Initiative (PCI) team in
July 2004 contributed to alerting the
government and donors to the seriousness of the situation.
34
IDS
RESEA
RCH
REPOR
T 57
Paradoxically, given the Ethiopian governments ambition to

sedentarise pastoralists, vulnerability
to drought in Gashamo has been exacerbated by a
proliferation of semi-urban settlements in
recent years. Large villages with permanent buildings are
now found every 1015 kilometres,
especially along the main trading route to Burao and Berbera.
These settlements are financed
mainly from remittances sent to families in Gashamo by
relatives living abroad. The settled
pastoralists of Gashamo keep some livestock near their
houses and some out in the bush.
During the drought of 2004, it was the immobility of this
semi-sedentary lifestyle that caused
cattle, shoats (sheep and goats) and even camels to die in
large numbers. Many of these villages
were partly or wholly abandoned by late 2004, as people left
to escape the drought.
Gashamo is a relatively stable district, being sparsely
populated and dominated by two major Isaq
sub-clans, the Habr Yoonis of the Habr Gerhajis, and the
Habr Jelo. The Isaq also dominate the
Republic of Somaliland, and cross-border relationships are
good. Most tensions within the
region concern disputed access to water (e.g. berkad).
However, relations between the Isaq and
Ogadenis in neighbouring districts are tense, and inter-clan
conflicts occur occasionally.
3.1.2 Shinile District

Shinile is uniquely located in the extreme north of Somali
Region, wedged between Dire
Dawa to the west, Afar Region to the north and Djibouti to
the east. Shinile is the poorest
of our pastoralist districts and one of the poorest parts of
Somali Region. The reasons for
this are not immediately apparent. All households in our
sample own livestock sheep and
goats are most common (found in over 90 per cent of
households surveyed), followed by
donkeys and cattle (over 50 per cent of households) and
rearing livestock is the dominant
livelihood activity. Selling animals and animal by-products is
the main source of income for
the majority of households. This is a sparsely populated area,

with few territorial disputes or
conflicts over natural resources.
The people of Shinile belong to sub-clans of the Issa clan,
which inhabits the territory
stretching from Dire Dawa to the Republic of Djibouti, where
the Issa dominate. The Issa
are united under their constitutional order, the famous Heer
Issa, which promotes peace and
stability throughout Issa territory. However, the Issa have
difficult relations with the Afar,
their northern neighbours. Conflict over the poorly defined
regional border has led to many
deaths on both sides, and villages that are close to the border
and along the tarmac road to
Djibouti are regularly attacked. Some consider the longrunning dispute between the Issa
and Afar to be over access to the Awash river; others consider
it a struggle to control the
contraband trading route from the eastern highlands of
Ethiopia and Dire Dawa to Djibouti.
Most of Shiniles external linkages are with Djibouti rather
than Somalia (or highland
Ethiopia), but Djibouti itself is poor. There is a good and
regular public transport system
linking Djibouti to Addis Ababa and Dire Dawa, via Ayshia
town in Shinile, because Ethiopia
imports most of its oil (and many contraband commodities)
through the port of Djibouti.
Labour migration is one of the main coping strategies in
response to drought and limited
local livelihood options, but employment opportunities in
Djibouti and Dire Dawa are
limited to casual labour and informal services (such as
portering or running a tea-stall). Since
these activities are highly competitive and poorly
remunerated, migration does not generate
significant flows of income and remittances back into Shinile
District.
The people of Shinile have relatively little contact with other
parts of Somali Region. This
restricts their mobility and magnifies the effects of droughts,
as they have few places to
move to with their animals to find water and pasture. The
local climate is harsh, with
droughts being frequent and severe. Importantly, the rainfall
pattern in Shinile is different
from other parts of Somali Region. The main rains fall in the
middle of the year, and are
called Karan. These are equivalent to the long rains called
Kremt in the neighbouring
Ethiopian highlands. Reliance on one major rainy season, and
the long period in between
rains, makes Shinile very susceptible to drought. Most of the

last dozen years are described
as drought years by the local people.
In fact, there is a widespread perception that rain failures
have become more regular, or
even continuous, during the last decade. These droughts are
blamed for increasing
35
poverty, causing herds and flocks to shrink in size, and

undermining the traditional
restocking mechanism, known locally as jiisin. As a result
of the increasing uncertainty
associated with pastoralism, some families with access to
cultivable land are now diversifying
into settled agriculture. In our survey, 11 per cent of
households are cultivating cereal crops.
A related emerging trend is the increasing involvement of
women in providing for their
familys food and other basic needs, by engaging in petty
trading or running tea-stalls.
Another recent phenomenon is a dramatic rise in the numbers
of people collecting
firewood and burning or selling charcoal to supplement the
households income from
livestock rearing. In our sample, two-thirds of households (67
per cent) are selling charcoal
and over half (57 per cent) are selling firewood. These
activities are traditionally pursued by
the poor and by women, but are now being practised by most
households, and even by
many men. This increase might be a temporary response to
livelihood stress, as firewood
and charcoal selling are commonly adopted as a drought
coping strategy, only to be
abandoned again when the rains return and livestock herds
regenerate. Nonetheless, the
combination of rearing livestock and selling charcoal or
firewood is at present the dominant
livelihood strategy in Shinile District.
3.1.3 Shilabo District

Shilabo District is located at the heart of the Ogaden, an area
that can be described as very
hot and dry, with lots of camels. The Ogadenis are
numerically the largest Somali clan in
Ethiopia, and are very active in local politics, holding more
than half the seats in the
regional parliament. Only one quarter of bureau heads in the
regional administration are
Ogadeni, however, due to pressures to allocate these posts to
different clans. Ogadenis also
run many of the biggest businesses in Gode and Jigjiga town.
Shilabo town is an important centre for the district, as it is
well supplied with shops,
markets and public services. There is a telephone facility,
which people use to keep in
contact with relatives abroad. The government has recently

installed a generator in Shilabo
town, which provides 12 hours of electricity each day.
Shilabo town is also a transport hub
for travel to other places within Somali Region, such as
Kelafo and Gode town, and to
Bosaso in Puntland, north-eastern Somalia.
Shilabo is an unstable area. This is a border area that was
badly affected by the
EthiopiaSomalia war in the late 1970s, when Shilabo town
was burnt down once and
relocated twice. Nowadays, the Ogaden National Liberation
Front (ONLF) is very active,
and there are frequent skirmishes between the ONLF and
government forces. There are
also conflicts between sub-clans, usually over livestock
migration routes and access to wells,
which can result in fatalities as the population is heavily
armed. The two major clans are the
Ogaden Makahil, and the Ogaden Bah Geri, which are
usually on good terms. There is
however ongoing friction between two minor clans, the
Hawiye Habr Gedir and the
Marehan. An additional complication has been created by the
extensive natural gas deposits
that are known to exist in the district, ownership and control of which are contested.11
Partly because of its long history of conflict and

displacement, there is a sizeable Ogadeni
diaspora across the world, which remits a large but
unquantifiable volume of income to
Shilabo and neighbouring districts. Nonetheless, livelihoods
in Shilabo are undiversified, with
just under two activities per household recorded in our
survey, but lucrative. Livestock
rearing dominates, especially camels (68 per cent of
households, the highest of any district
surveyed), sheep and goats. Sales of animals and dairy
products generate most of the
household income. No households reported engaging in crop
farming. Six individuals hold a
salaried job. There is limited engagement in craftwork or
service provision, except for five
Koranic teachers and one shoe-repairer. One household in six
has a member who either
runs a tea-shop or sells coffee and cake by the roadside. One
man interviewed works as a
night watchman, guarding the local food aid store. His
employers pay him in kind, by giving
him 2.5 quintals of wheat and 2 gallons of cooking oil
wherever there is a food distribution.
11 More details about the Calub natural gas deposits in Shinile District are
provided in Chapter 11.
36
IDS
RESEA
RCH
REPOR
T 57
Ogadenis define themselves as camel herders (We are good

with camels). They also keep
sheep and goats (shoats) and, to a lesser extent, cattle. To
search for pasture, they move
with their livestock within the extensive Ogaden territories,
both inside Somali Region (to
Kebribayah and other Ogadeni districts) and across the
border into Somalia. Mobility is
facilitated by the fact that we are all from the same family,
and strong reciprocal
arrangements have been practised for generations. Migration
does occasionally lead to
conflicts, especially at times of livelihood stress, but we were
informed that this is rare as
allowing unrestricted reciprocal access to natural resources is
a strict norm within the clan.
There are many deep and shallow wells in Shilabo, which are
used during the hot, dry
season of Jilaal (from January to March). At this time of year
many herders from
neighbouring districts come to Shilabo to water their animals
from these wells.
Importantly, Shilabos wealth derives not only from
pastoralism but also from trading, both
official and contraband, which is facilitated by its
proximity to the border with Somalia
and the port of Bosaso, allowing direct access to the lucrative
import and export markets
of the Gulf states. Several households interviewed (25 per
cent) are engaged in exporting
livestock and importing clothes, shoes, food (cereals, sugar,
coffee) and white goods
(electrical appliances, a major component of the contraband
trade). This lucrative trade is
made possible because of the close relationships and
extensive networks that exist between
Ogadenis in Somali Region and Ogadenis across the border
in Somalia, as well as with
coastal clans like the Hawiye and Herti.
3.2 Agro-pastoralist districts

Three districts in our survey are predominantly agropastoralist: Kebribayah in northern
Somali Region, Doboweyn in the centre of the region, and
Cherati in the south.
3.2.1 Kebribayah District

Kebribayah is located close to the regional capital of Jigjiga,
and it is an agro-pastoral
district. Almost all households keep animals and most are
also farming. By definition, agropastoralism is a more diversified livelihood system than pure
pastoralism, and the average
household pursues three livelihood activities rearing
livestock (especially cattle, sheep and
goats, but also camels, especially in the drier eastern and
southern parts of the district),
crop farming, as well as a minor income-earning activity like
charcoal burning, or collecting
firewood or construction materials for sale. Apart from
growing cereals, some farmers also
cultivate vegetables, khat, or root crops. Growing khat for
sale is a new activity, and is seen
as more reliable than rearing livestock. Kebribayah cannot
compete with Hararghe in
neighbouring Oromiya Region, which has a more conducive
climate and is the centre of
khat production in Ethiopia, in terms of both quantity and
quality. Instead, farmers in
Kebribayah cultivate a lower quality khat that is less labourintensive and reaches a lucrative
market among lower income earners.
Only one person in 100 households surveyed holds a salaried
job, and there are no service
providers, no craft workers and very few traders just two
trading khat and one livestock
trader in our sample. However, Kebribayah is located on
the main trading route into
Somaliland from Jigjiga and the Harar highlands, and trading
is a significant activity, both for
contraband goods (which has boomed in recent years, despite
government efforts to
suppress it) and for official cross-border trade.
Kebribayah is a multi-clan district and is characterised by
intense competition between clans
over access to resources and political representation. These
disputes are generally settled
through negotiation rather than open conflict, but relations
between neighbouring groups are
often tense. Because these settled pastoralists are both
farming and keeping livestock,
competition occurs both for farmland (particularly valleybottom farmland) and for access to
pasture and water. There are two seasonal rivers as well as
pans and dams. People live in stable
and relatively large settlements and farm family plots, while
the surrounding rangeland is
divided into family holdings. During dry seasons and

droughts, herds and flocks are sometimes
moved in search of pasture and water elsewhere, but longdistance mobility is uncommon.
37
Kebribayah is also home to thousands of long-term refugees

and IDPs, most of whom live
in two camps at Hartisheik. Originally, these camps were set
up for refugees from the civil
war in Somalia, but a later wave of refugees came to
Hartisheik after being deported from
Djibouti. More recently, pastoralists who lost their livestock
during the droughts of 1999 and
2004 have come to Hartisheik in the hope of receiving food
aid. However, government and
donor support to the Hartisheik IDP camps has become
increasingly erratic and sporadic,
and child malnutrition and mortality rates reached critical
levels in early 2005 (see Chapter
6). The government suspects that many camp residents are
economic migrants from
Somali or Oromiya Region rather than genuine refugees or
IDPs who are using the
camp as a base for engaging in contraband trading. Many
people believe that the
government is trying to close down the camps in response to
this perception, by
withholding food aid and other essential supplies.
3.2.2 Doboweyn District

Doboweyn is located in Korahe Zone in central Somali
Region, between the pastoralist
district of Shilabo and the riverine farmers of Kelafo. Most
residents live in semi-permanent
settlements, having no need to move because access to water
is more reliable than in many
other parts of the region. Doboweyn has a seasonal river and
a large number of wells, so
water scarcity is rarely a problem, though the water from
many of these wells is salty and
unfit for human consumption. Access to wells is free and
open to pastoralists from other
districts, who often bring their animals through Doboweyn in
the dry season.
Almost all households in our Doboweyn sample rear animals
more households own
donkeys (70 per cent) in Doboweyn than in any other district
surveyed, and the herds or
flocks of cattle, sheep, goats and camels recorded are among
the highest in the region.
Almost half these households are also cultivating cereal crops
(though no pulses, vegetables
or fruits are grown), mainly in the valleys on either side of
the river. In good years, farmers

produce surpluses, which they sell in local markets, as
Doboweyn is not situated on any
major trade route. Perhaps because trade does not feature
strongly in local livelihoods, the
people of Doboweyn are among the poorest in our sample in
terms of cash incomes (along
with the two other agro-pastoral districts of Cherati and
Kebribayah). They are relatively
wealthy in terms of livestock ownership, but do not engage
heavily in livestock marketing.
Apart from selling dairy products (milk, butter, ghee) derived
from their livestock, no other
livelihood activity is pursued by more than one in ten
households surveyed. A few people
sell firewood, construction materials (grass, wooden poles),
or tea and cake in their villages.
The only waged employment reported was a single daily
labourer. There are no craft
workers and hardly any service providers: just one Koranic
teacher and one midwife.
Two Ogadeni clans dominate Doboweyn District: the
majority Bah Geri and the minority
Abdwak. Relations between the two clans, who are heavily
armed, are always tense and
occasionally violent. One recent trigger for a major conflict
was some land that is suitable
for farming and was claimed by both clans (see Chapter 10
for details). Even within each
clan, skirmishes occur frequently between neighbours over
access to farmland, irrigation
rights and borders between family plots, often leading to loss
of life. One explanation for
this apparently disproportionate response to contested
property rights may be that these
people became sedentarised relatively recently.
A time when common access to all the rangeland was
open to all clan members is still
very much in the collective memory. Rapidly changing
circumstances have led to the
settling down of many clan members, who try to pursue
a mixed agricultural
livelihood. Though individual circumstances of
particular incidents differ, this demand to
exclusively hold down choice land for farming is
generally challenged, and defended by
armed confrontations.1 2
12 Abdi Umar, pers. comm.
38
IDS
RESEA
RCH
REPOR
T 57
3.2.3 Cherati District

Cherati is located in Afder Zone, between Gode, the largest
town in central Somali Region,
and Liban Zone in the extreme south of the region. Cherati
has the largest market in Afder,
well placed at the crossing point of two important trade
routes: the northsouth route that
follows the seasonal river Web from its source in the Bale
mountains to the Somalia
border, and the new eastwest road linking Gode to Filtu in
Afder. Cherati District straddles
the River Web, which provides enough water and vegetation
to support agro-pastoralism
on both riverbanks. Beyond this strip of vegetation,
transhumant pastoralism is the main
livelihood system.
The large majority of households interviewed in Cherati
District (over 80 per cent) are agropastoralists who live in settlements along the river, where
they are farming and rearing
animals. More than two-thirds of households own goats,
sheep and donkeys, about half
own cattle and one-third own camels. Farming is dominated
by cereal crops, especially
sorghum and maize. The next most common income-earning
activity is basket- or matmaking (one household in five), with no other activity being
reported by more than one
household in ten. Nobody in these households is employed in
the formal sector, no one is
trading, but there are some informal service providers eight
Koranic teachers, four
circumcisers, one traditional healer and one traditional birth
attendant (TBA). Apart from
selling construction materials, few individuals earn income
from charcoal burning or
gathering natural products such as firewood, precious stones,
wild fruits, incense or gum.
In terms of cash income though not livestock ownership
Cherati is the poorest district in
our sample. Like most rural districts in Somali Region, it is
inadequately provided with basic
services. Only one clinic (operated by Mdecins Sans
Frontires (MSF) Belgium) delivers
health care to an estimated population of 81,000. Water-borne

diseases are common, since
the population depends entirely on surface water the River
Web, which is stagnant for
much of the year, ponds and shallow wells.
3.3 Farming districts

Three districts in our household survey are dominated by
crop farmers: Kelafo and Dolo Odo,
where riverine agriculture is practised, and the Jigjiga plains,
where agriculture is rainfed.
3.3.1 Kelafo District

Kelafo District is home to the so-called Jerer, or Somali
Bantus, who came to Somali
Region many decades ago and settled along the banks of the
Shabelle river. The Jerer have
always been sedentary farmers. Almost all households
surveyed cultivate cereals (maize,
sorghum), as well as vegetables (onions, tomatoes, spinach),
pulses and oil crops (beans,
sesame), and fruits (papaya, bananas). Although farming is
primarily for consumption and
only secondarily for sale, in recent years a new marketing
opportunity has opened up for
farmers who cultivate onions for export to Somalia. Somali
traders come to Kelafo to buy
onions to sell across the border, in Belet Weyne and
Mogadishu. Kelafo town is a good
market for our animals while Belet Weyne is good for onions
and other business. Since
onions are a high-value crop, many farmers are neglecting
subsistence crops to take
advantage of this opportunity. Wealthy Somali businessmen
are also renting land and
recruiting sharecroppers, as well as investing in irrigation
machinery, to cultivate onions for
the export market.
Lack of land is a major problem facing farmers along the
Shabelle river in Kelafo District.
There is no vacant land for farming and all land is allocated
to individuals. The only way to
access land is through inheritance or purchase, but as
population grows so individual land
ownership is falling, as inherited land is sub-divided among
family members. Traditional landsharing institutions are also said to be declining, because land
scarcity reduces the availability
of spare land to share with others. There are also clashes over
access to land between
farmers and rising numbers of pastoralists who have lost their
livestock and are trying to
switch to farming or agro-pastoralism.

39
Despite the dominance of crop farming, livelihoods in our

Kelafo sample are more diversified
than in any other district, with more than four activities
recorded per household. Almost half
of these households rear animals. Cattle are popular, as oxen
are used in farming, and many
families have small flocks of sheep and goats. Donkeys are
also important, as donkey-carts are
used to transport water and farm produce to the market but
donkeys are expensive and
beyond most peoples reach. Petty income is also derived
from making charcoal or selling
firewood. There are a few farm workers and daily labourers,
but no member of these
households has a salaried job, and there are very few traders.
One household in ten engages
in craftwork, such as basket-making, mat-making or making
farm tools. There are various
informal service providers, including traditional healers,
circumcisers, Koranic teachers and
TBAs. Many of these activities are only possible because
these are densely settled
communities so that neighbours can provide services, labour
and craftwork to each other.
There is some resentment and hostility between the Bantu
and Somali peoples in Kelafo
and neighbouring Mustahil District. Even the word Jerer
(hard hair), is a derogatory term
for Bantu people used by Somalis who are Jilee (soft
hair). According to one Bantu
sharecropper: Our area is marginalised by the neighbouring
Ogaden clans who patronise
and discriminate against us. They do not consider us as
Somalis, and treat us as slaves.
Sharecroppers are especially resentful of the treatment they
receive from Somali
landowners. In the early 1990s and early 2000s, there were
outbreaks of armed violence
between the Jilee pastoralists and Jerer farmers. Farmers
houses were burned down and
their fields were grazed by pastoralist flocks and herds.
At the community level, neighbouring Somali clans were
accused of undermining economic
development in Kelafo. A specific complaint relates to the
construction of a dam upstream
from Kelafo, as part of the government-sponsored West Gode
irrigation scheme, to divert
water from the Shabelle river to farms near Gode town. One
farmer explained: If our land

gets water, then we can plant more. We started encountering
this water problem about
eight years ago, when the Ogadenis blocked the river. The
adverse consequence of the
dam at Gode is that the Shabelle no longer floods the
farmland along its banks, which used
to improve soil fertility. Due to lack of floodwater, land
scarcity is intensifying, as farming is
concentrated closer to the riverbank. More generally, the
people of Kelafo argue that their
district has been marginalised by the regional administration
because it is a Bantu area. As
evidence they point to the poor road network, and the fact
that although the area is
densely settled, almost no basic services are provided in these
communities.
3.3.2 Dolo Odo District

Dolo Odo is a district in Liban Zone, and is the most
southerly of our research sites. Four
types of settlement pattern are found in this complex corner
of Somali Region.
Transhumant pastoralists move with their camels and shoats
in the arid northern interior of
the district. Semi-sedentarised agro-pastoralists live near the
river valleys, where their dairy
herds of cattle feed on by-products from their farms. Along
the riverbanks are stable, longterm settled farming communities. Finally, Dolo Odo town is
a large and busy market centre
near the border between Ethiopia, Somalia and Kenya, at the
confluence of the Web,
Dawa and Ganale rivers. The urban residents of Dolo Odo are
traders and business-people,
workers in restaurants and hotels, or truck-drivers.
Fieldwork was conducted among the farming households
who live along the Dawa and
Ganale rivers. Like the communities surveyed in Kelafo
District, the local people are
ethnically Bantu from the Gabaweyn tribe not Somali
(though they are Somalispeaking). Because farming in Dolo Odo takes place along
the river, it is more diversified
than in rain-fed farming areas (a characteristic it shares with
farming along the Shabelle
river in Kelafo District). Most households cultivate cereals
(maize, sorghum), but many also
grow pulses (beans), oil crops (sesame or sunflower),
vegetables (onions, tomatoes,
pumpkins), and fruit (bananas, watermelon).
Most of these farmers keep some animals, including sheep
and goats, bullocks for ploughing
and donkeys for transport. Donkey owners charge high prices

for transporting people and
40
IDS
RESEA
RCH
REPOR
T 57
produce to and from markets, but donkeys are expensive so

only a few people own them.
Crop marketing is problematic. Unlike Kelafo District, there
is no commercial onion farming
in Dolo Odo; however, producing fruit for the market is a
growth sector. Perhaps because
of its location close to the unstable Gedo region of Somalia
few traders come to rural
Dolo Odo. Also, although Dolo Odo town is a bustling hub
for trade routes into Kenya and
Somalia, with regular trucks and public transport across the
border, roads within rural parts
of the district are bad or non-existent. As a consequence,
households in our Dolo Odo
survey are considerably poorer than those in Kelafo. Despite
being located fairly close to the
district capital and the large town of Mandera in northern
Kenya, the riverine farmers have
weak ruralurban linkages, partly because these urban centres
are dominated by Somalis.
Persistently low rainfall since the 1990s has affected both
crops and livestock production in
Dolo Odo. The river provides some water for both farming
and animals, but since there are
no irrigation facilities crop failures do occur. When the rains
are good, adequate pasture is
available for the livestock, but in bad years animals have to
be moved long distances to
Oromiya Region or into Somalia and northern Kenya in
search of grazing and water.
Cattle and sheep have been worst affected by recent droughts,
while camels and goats
offer more resistance. However, camel herds have been
afflicted by diseases (known locally
as Goudaan, Mattaq and Shibnir), and there are no veterinary
services in the district.
Formal employment is almost non-existent in these
communities. There are no traders and
very few service providers just one water-carrier and two
healers in 100 households
interviewed. An important secondary livelihood activity is
making mats and baskets for sale,
using grasses and reeds collected from the riverbank. Onethird of households in our Dolo
Odo survey engage in basket- or mat-making, especially
during the dry season, as the grass

must be left in the sun to dry for several days. These reeds
and grasses are also used as
roofing materials, so along with firewood and fodder they are
gathered mainly by women
and children for both domestic use and sale.
Conflict within the riverine farming communities of Dolo
Odo is almost unknown. The
ethnic Bantu are already marginalised within Somali Region;
they are under-represented in
regional political institutions, and they do not engage in clan
politics. Having settled along
the river many decades ago, households generally own their
own land, acquired through
inheritance, so rights to land are well established and are not
contested. However, some
former (Somali) pastoralists have settled fairly recently and
become agro-pastoralists and
farmers. Also some entrepreneurs in Dolo Odo are financing
a type of sharecropping in the
riverine communities, which exacerbates inequality between
the ethnic groups and is a
potential source of tension and resentment.
3.3.3 Jigjiga rural

Jigjiga Zone is the most heavily settled of Somali Regions
nine zones, with almost a quarter
of the regions population, 80 per cent of whom are rural.
Over 650,000 people are agropastoralists or farmers in this zone. Our household survey
team interviewed 100 farming
households in the plains outside Jigjiga town.
In the absence of permanent or seasonal rivers, people of the
Jigjiga plains depend on rain
for both domestic water consumption and crop farming. This
part of Somali Region receives
three rainy seasons rather than two: Gu (April to May),
Hagaa (June to July) and Karan
(August to September). The Karan rains are unique to the
northern zones of Jigjiga and
Shinile, and allow the cultivation of long-maturing rain-fed
crops, which is not feasible in
other parts of the region. Crops cultivated on the Jigjiga
plains include cereals (mainly
wheat and barley, also maize and oats), pulses (chickpeas,
lentils) and oil seeds (flax). Smaller
amounts of Irish and sweet potato, peas, onions and garlic are
also grown, mainly for sale.
The main harvest occurs towards the end of the year. Farmers
consume from their
granaries from December until food stocks are exhausted,
which varies according to factors
such as landholding and rainfall. For the remaining months of
the year until the next

harvest, these families depend on market purchases of food,
which they finance by selling
livestock or milk, and searching for daily labour. Poor
households are forced into selling
41
some food staple at harvest to meet their cash needs, and to

buy food back later at higher
prices, which is a source of livelihood vulnerability. Since
these households depend on
cropping for much of their food and their cash income, they
are extremely vulnerable to
drought or erratic rains.
Most farming households keep some cattle: cows are kept for
their milk, while oxen are
used in ploughing. Having a pair of oxen is a determinant or
relative wealth: the household
who has one ox has breakfast, while those who have two
oxen are able to have breakfast
and lunch (SC-UK, DPPB and partners 2001b: 12). Some
sheep and goats are also kept by
most families, while donkeys are used as pack animals.
Livestock are fed on crop residue
(stalks, cobs and husks) as well as pasture. Local farmers
have access to several large markets
for selling farm produce or livestock and buying food,
including Jigjiga town, Hartisheik,
Chinahsan, Babile and Dire Dawa.
Despite their location close to the regional capital, the people
of rural Jigjiga are
inadequately provided with basic services. There are very
few schools and health clinics,
although since the decentralisation process started some
primary schools have been
constructed. There is no potable water for human
consumption in these rural communities,
and people depend on unclean surface water (e.g. ponds),
which contributes to the high
prevalence of preventable illnesses.
3.4 Urban centres

Two surveys were conducted in the largest urban centres of
Somali Region: Jigjiga and Gode.
3.4.1 Jigjiga town

As the capital of Somali National Regional State, Jigjiga is a
rapidly growing town with a
large number of professionals working for the government,
private sector and NGOs, in
addition to a variety of informal sector activities. Many of the
government and NGO
workers are strangers to Somali Region, having been posted
to Jigjiga from Addis Ababa or
abroad. However, many other urban residents have strong
linkages with rural communities,

having migrated from rural areas themselves, and the nature
and strength of ruralurban
linkages can be an important determinant of livelihood
security or vulnerability.
Urban centres offer a very different set of livelihood
opportunities to rural areas. Two
striking features of livelihoods in Jigjiga town are the variety
of ways that people earn
income many of which are not available to people living in
rural areas and the wide
range of returns to each livelihood activity.
Salaried employment tends to offer the highest and most
reliable income. In one
household, the 25-year-old eldest son is a junior civil servant
in the Regional Administration,
and earns a salary of 1,200 Birr per month. In another Jigjiga
household the 34-year-old
household head is also employed as a civil servant, earning
1,058 Birr per month.
Interestingly, some of the best-paid jobs in Somali Region are
working for NGOs. One 48year-old man in Jigjiga town works for an international NGO
and earns a salary of 3,500
Birr per month. He also gets a transport allowance of 600
Birr per month and a medical
allowance of up to 3,000 Birr per month. With salaries
offered by NGOs up to three
times higher than those paid by the Government of Ethiopia,
it is hardly surprising that
many of the most talented professionals in Somali Region
seek employment outside the
public sector. A further deterrent is that government
employment in Ethiopia is not as
reliable as in civil services elsewhere, which are often seen as
providing job-for-life
sheltered employment. During 2004, for instance, the
Regional Government terminated
the employment of 900 officials, alleging that their education
documents were forged. This
was the second incident of mass employment termination by
the Somali Regional
Administration within the same year.
In the informal sector, incomes are even more irregular, and
often unpredictable. One
traditional healer in Jigjiga, asked to estimate his monthly
income, replied: It varies.
42
IDS
RESEA
RCH
REPOR
T 57
Sometimes I can make 450 Birr in a month, but its not every
day that I get income. Many
informal livelihood activities in urban areas are illicit or
illegal. It is reported that there has
recently been a rise in the number of young women in Jigjiga
turning to prostitution to
support themselves and their families. Informal activities also
tend to provide lower returns.
One man who makes his living by begging in Jigjiga market
estimated his monthly earnings
at 150 Birr. At the other end of the income scale, one
prosperous trader in Jigjiga town
claimed to earn approximately 30,000 Birr each month from
selling gemstones.
Trading can be extremely lucrative, but informal cross-border
trade has been declared illegal
by the Ethiopian government. Many of the wealthiest people
in Jigjiga acquired their
wealth through smuggling or selling contraband goods. In the
last few years, the
government has clamped down heavily on the contraband
trade, and there have been
several violent confrontations between traders and the state.
In one incident, Ethiopian
military personnel and Customs officials intercepted a
convoy of trucks bringing contraband
commodities to Jigjiga market. The smugglers were armed
and gunfire was exchanged,
with loss of life on both sides. Traders interviewed in Jigjiga
complain that the government
claims to support a liberalised free market economy but
instead imposes restrictions on
trade that are preventing many traders from trading, as well
as disrupting supplies and
causing prices of basic commodities to rise. In this way, even
urban residents enjoying the
highest incomes in Somali Region are vulnerable to
unpredictable shocks though mainly
through the market rather than drought or conflict.
3.4.2 Gode town

Gode is the former capital of Somali Region, and as such it is
a large and busy urban
settlement with well-developed infrastructure. The town
centre has electricity, piped water,

a hospital, clinics and pharmacies, primary and secondary
schools, and an agricultural
college. Flights to and from Jigjiga and Addis Ababa leave
daily from Gode airport. The
Shabelle river runs past the town, providing water for a largescale irrigated farming
scheme. On the other hand, Gode is also home to thousands
of former pastoralists who
live in informal settlements on the outskirts of town, having
migrated to Gode after losing
all their livestock during the drought of 1999/2000.
Almost all livelihood activities recorded in our household
survey (n=55/64, or 86 per cent)
are practised by the 100 households interviewed in Gode
town, but no single activity or
cluster of activities is dominant. This partly reflects the
greater range of livelihood options
available in urban areas, in contrast to rural areas. Farming,
however, is not practised by any
members of this sample, and only a few households own any
animals.
On the other hand, one household in five interviewed has a
member with a salaried job a
higher proportion even than the larger town of Jigjiga.
Several people are employed as
construction workers, carpenters, restaurant workers and
housemaids. Four young men are
in military service. As is common in towns, the informal
sector is an important source of
employment and income. There are artisans making baskets
and mats, tailors, potters and a
woman who makes jewellery. Many families earn income
from renting out rooms. Various
services are provided in Gode town, including henna
decorators (women who paint henna
patterns on other womens hands and feet), barbers,
hairdressers, tailors, musicians,
marriage counsellors and traditional healers. Three
individuals admitted to begging for a
living no beggars were found in rural areas.
Many wealthy businessmen and businesswomen live in
Gode, running wholesale or retail
stores, and trading locally or long distance. A wide range of
commodities are traded by
members of our surveyed households, including clothes and
shoes, livestock and livestock
products, food crops and other food stuffs (sugar, flour, etc.),
contraband (such as
electronic goods smuggled from the Middle East) and khat.
In recent years, the Government
of Ethiopia has clamped down on contraband trading (see
Chapter 4), which has affected
businesspeople in Gode. Nonetheless, average incomes in our

Gode sample are significantly
higher than incomes in any rural district, and second only to
incomes in Jigjiga town.
43
4 Livestock rearing and

marketing
Livelihoods in rural Somali Region are dominated by
livestock, which are the main assets
owned by the majority of households, are a major source of
food consumption, and are
traded both live animals and animal by-products to
generate cash income for
pastoralists, agro-pastoralists, traders and a range of livestock
marketing agents. This chapter
provides an overview of livestock ownership from our
household survey and the problems
faced by livestock rearers, describes the livestock-marketing
system within Somali Region,
and highlights the importance and risks of the cross-border
trade through Somalia and
Kenya into the Gulf states.
4.1 Livestock ownership in Somali Region

With the exception of households interviewed in the two
urban centres of Jigjiga and
Gode town, most households in our survey own livestock of
various kinds. Different kinds of
animals are reared in different parts of the region, and
different mixes of animals are kept
by pastoralists, agro-pastoralists, and farmers. Goats and
sheep (shoats) are most common,
and are found throughout the region, though they are
favoured most by pastoralists and
least by farmers. In the pastoralist districts of Gashamo and
Shinile, almost all households
keep sheep and goats (averaging about 20 head of each), but
in the riverine farming
communities of Kelafo only one in four households keeps
shoats (averaging just four head of
each) (Table 4.1 and Table 4.2).
Perhaps the defining image of pastoralism in the Horn of
Africa is the camel herder, and
camels are indeed found throughout Somali Region,
especially in the drier eastern parts of
the Hawd (Warder Zone) where water and grazing are scarce.
In our survey, camels are
owned by more than half the households in Gashamo,
Shinile, Kebribayah and Shilabo (Table
4.2). Shilabo has the highest proportion of camel owners
(two-thirds of households), though

it was less severely affected by the drought of 2004 than
Gashamo and Kebribayah, when
Table 4.1 Households owning livestock, by district
District Camels (%) Cattle (%) Goats (%) Sheep (%)
Donkeys (%) Number of
households
Pastoralist 59 27 83 82 29 301
Gashamo 55 6 91 82 10 100
Shinile 53 53 94 93 66 100
Shilabo 68 21 65 70 12 101
Agro-pastoral 43 72 65 65 57 298
Kebribayah 57 92 64 64 32 100
Doboweyn 38 65 61 64 70 98
Cherati 34 57 71 68 70 100
Farmer 4 55 45 43 44 301
Kelafo 2 44 28 24 20 101
Dolo Odo 3 42 51 44 54 100
Jigjiga rural 8 79 56 61 60 100
Urban 1 3 2 2 1 200
Jigjiga town 2 4 0 0 0 100
Gode 0 2 3 4 2 100
Total 29 % 42 % 53 % 52 % 36 % 1,100
44
IDS
RESEA
RCH
REPOR
T 57
Table 4.2 Livestock ownership, by district (all households)

District Camels Cattle Goats Sheep Donkeys TLUs
Pastoralist 8.2 3.9 22.0 18.4 1.3 20.4
Gashamo 5.0 0.3 19.5 19.6 0.0 12.2
Shinile 4.9 7.7 24.5 19.4 2.2 18.4
Shilabo 14.7 3.8 22.0 16.3 1.4 30.5
Agro-pastoral 5.7 11.4 12.3 11.8 1.9 20.3
Kebribayah 4.1 8.6 4.3 6.5 1.2 14.0
Doboweyn 7.7 15.2 16.5 15.4 2.6 27.1
Cherati 5.4 10.6 16.2 13.6 2.1 19.9
Farmer 0.4 8.0 7.2 6.4 1.5 8.2
Kelafo 0.3 7.5 4.0 3.6 0.8 6.8
Dolo Odo 0.7 4.8 9.8 7.2 1.7 6.8
Jigjiga rural 0.3 11.7 8.0 8.6 2.1 11.1
Urban 0.2 0.7 0.4 0.3 0.1 0.9
Jigjiga town 0.4 1.0 0.0 0.0 0.0 1.3
Gode 0.0 0.4 0.9 0.6 0.1 0.5
Average 3.9 6.5 11.4 10.1 1.3 13.5
many households lost all their animals. Pastoralists in Shilabo
also had the largest reported
herds of camels at the time of our survey (15 head each, or 22
head excluding households
with no camels),1 3 followed by the agro-pastoralist district of
Doboweyn (eight head each),
which borders Shilabo and was also relatively unaffected by

the 2004 drought. Almost no
farming or urban household owns any camels.
Cattle are favoured by agro-pastoralists, with herds averaging
11 head being owned by
almost all households in Kebribayah, two-thirds of
households in Doboweyn, and over half
the households in Cherati. Similar sized herds of cattle are
owned by almost 80 per cent of
farmers in the Jigjiga plains, and (in smaller numbers) by just
under half the farmers in
Kelafo and Dolo Odo. Cattle are less well adapted to
pastoralist conditions, so are not
widespread in Shilabo or Gashamo where many cattle died

during the 2004 drought
but they are popular in the northern Somali district of Shinile,
which has a different rainfall
system that provides more water and grazing.
Finally, donkeys are popular throughout rural Somali Region
as pack animals, either
transporting water and firewood for domestic use, or carrying
commodities charcoal,
firewood, farm produce, milk to market. Since they have
limited mobility relative to
camels, they tend not to accompany camel caravans on longdistance trading trips, and they
are found in relatively few pastoralist households except for
Shinile, which practises a form
of pastoralism based on cattle and less migration with
animals than in the arid Hawd.
Donkeys are most common in the agro-pastoralist districts of
Doboweyn and Cherati,
followed by pastoralist Shinile, and they are also favoured by
farmers in the Jigjiga plains and
in riverine communities of Dolo Odo. On the other hand, the
number of donkeys owned
tends to be lower than other types of livestock usually one
donkey or a pair for domestic
use, sometimes two or three more for renting out (Table 4.2).
13 Like income, livestock ownership is susceptible to under-reporting in
household surveys. Animals that are
being reared in other households might not be reported as owned
by either household, for instance. While
efforts were taken to avoid this, it is possible that the numbers
recorded here understate actual ownership
levels, even if the trends across districts are accurate.
45
Finally, Table 4.2 shows the average tropical livestock units

(TLUs) per household. TLUs are a
method for converting different types of livestock into a
single index, so that total livestock
owned can be compared across different herd and flock
compositions.1 4 This calculation
reveals that livestock ownership is almost equal (at 20 TLUs)

in pastoralist and agropastoralist households. Pastoralists own more camels (which
have the highest TLU value) and
shoats, but agro-pastoralists own more cattle and donkeys.
The highest average TLU is
recorded for Shilabo (at just over 30), followed by
Doboweyn (27 TLUs), Cherati (20) and
Shinile (18). Farming communities have much lower values
(averaging 8 TLUs), while urban
households have a negligible amount (0.9 TLU).
Although most households in our rural sample reported that
they own some livestock,
almost all of them claim that the numbers of animals they
own today is fewer than in the
past. Table 4.3 records the reasons given by survey
respondents to the question: If the
number of livestock owned by your household has decreased
during the last ten years (since
1995), what are the reasons? As can be seen, died in
drought is by far the most common
reason, being mentioned by more than 100 per cent of people
who currently own camels
and cattle, and 98 per cent of owners of shoats. (This means
that many people who owned
camels and cattle ten years ago have since lost all these
animals during the recent droughts,
and own none today.) Disease is the second most common
reason for declining livestock
ownership. Although less visible and dramatic than drought
events, livestock diseases
emerge from this survey as a major killer of camels, cattle,
sheep and goats. This suggests a
large and urgent unmet need for veterinary services in Somali
Region, especially
immunisation of livestock.
After drought and disease, which are largely beyond the
control of livestock owners to
prevent, the next most common reasons for declining
livestock numbers reflect decisions
taken for nutritional, socio-cultural, or economic reasons.
Livestock off-take is a standard
livestock management strategy. Nutritional reasons include

selling animals to buy other
types of food (mainly cereals) or consuming the animals at
home. Socio-cultural reasons
include contributing animals as zakaat, sacrificing or
donating animals for religious or
charitable reasons, and making in-kind dowry or
compensation payments. Economic reasons
include selling animals to meet non-food needs for cash, and
lending or renting some
animals out to others, which keeps herds and flocks
manageable and diversifies risk, as well
as strengthening social relationships. These factors all
affected camels, cattle and shoats to a
significant extent, while donkeys were affected much less.
Finally, a number of misfortunes can reduce herds and flocks,
the most prevalent of which is
been taken and eaten by wild animals such as hyenas, which
was mentioned by over onethird of camel and cattle owners and half of sheep and goat
owners. Other misfortunes
include animals being lost, stolen, poisoned, raided or killed
in conflicts. Each of these
affected relatively few households, though the impacts on
their livelihoods might well have
been devastating, especially where camels and cattle were
involved.
Although livestock rearing is the most popular livelihood
activity recorded in our survey,
being practised by 70 per cent of all households and 86 per
cent of rural households
interviewed, it is clear that this is a highly risky way of
making a living. Herds and flocks
fluctuate quite dramatically according to drought episodes or
disease outbreaks, and income
earned from livestock is unpredictable and irregular, being
determined by a range of factors
such as market conditions and seasonality, as well as interannual variations in rainfall. Asked
to estimate his households monthly income from rearing and
selling livestock during 2004,
one Gashamo pastoralist replied: It depends on the prices
and the season we are in. A
second Gashamo pastoralist who lost most of his camels and
shoats to the drought
reported a 65 per cent decline in income from livestock sales,
while a third told how he
withdrew from livestock marketing after losing most of his
animals.
14 Tropical livestock units are calculated by multiplying animals owned
by an index value obtained from the FAO.
The technical definition is One Tropical Livestock Unit (TLU) is
equivalent to an animal of 250 kg liveweight

(FAO 1998). The index TLU values for east Africa are: camels = 1.6;
cattle = 0.7; goats and sheep = 0.1;
donkeys = 0.4.
46
IDS
RESEA
RCH
REPOR
T 57
Table 4.3 Reasons for decreasing livestock ownership, 1995

2005
Reason Camels Cattle Shoats* Donkeys
Died in drought 363 (114%) 477 (103%) 570 (98%) 67
(17%)
Disease 247 (77%) 265 (57%) 346 (59%) 23 (6%)
Contribution (zakaat) 143 (45%) 165 (36%) 262 (45%) 1
(0%)
Sold for food 124 (39%) 213 (46%) 257 (44%) 12 (3%)
Eaten by wild animals 117 (37%) 170 (37%) 287 (49%) 13
(3%)
Consumed at home 103 (32%) 175 (38%) 391 (67%) 5
(1%)
Dowry payment 96 (30%) 116 (25%) 79 (14%) 0 (0%)
Offering (Allah bari) 73 (23%) 114 (25%) 329 (56%) 1
(0%)
Charity (qaadhaan) 68 (21%) 97 (21%) 211 (36%) 4 (1%)
Compensation payment 62 (19%) 44 (10%) 25 (4%) 0 (0%)
Sold (not for food) 53 (17%) 95 (21%) 117 (20%) 3 (1%)
Lent or rented out 30 (9%) 29 (6%) 51 (9%) 2 (1%)
Lost 24 (8%) 31 (7%) 50 (9%) 4 (1%)
Stolen 16 (5%) 14 (3%) 20 (3%) 2 (1%)
Conflict 8 (3%) 1 (0%) 1 (0%) 0 (0%)
Poisoned 6 (2%) 20 (4%) 26 (4%) 2 (1%)
Raiding 5 (2%) 3 (1%) 2 (0%) 0 (0%)
Total livestock owners 320 (100%) 462 (100%) 583
(100%) 396 (100%)
* Responses for sheep and goats were virtually identical, so
are merged as shoats.
I used to sell animals every six months and I used to get
about 1 million Somali
Shillings or more each time. Now because of the
drought I only have some goats left
to sell and last time I earned 350,000 Somali Shillings.
I used to sell about 20 animals every year camels,
cattle, sheep and goats. For the
last three years I earned about 20 million Somali
Shillings from livestock sales. Now
almost all the livestock I owned are dead due to the
drought. I have a few animals left,

but I am not selling them. I only slaughter a sheep or a
goat when we have nothing
else to eat.
Pastoralists perceive themselves as having been wealthy in
the past, but impoverished by
recent livelihood shocks. Thirty years ago, we pastoralists
were rich. Now we are nearly all
poor.15 The empirical evidence does not entirely support this bleak
assertion. A comparison
of household incomes across districts (see Table 7.1 in

Chapter 7) reveals that the pastoralist
districts of Gashamo and Shilabo are the wealthiest of nine
rural districts in our survey, in
terms of monthly cash income, and that farmers and
pastoralists earned almost equal levels
of income during 2005. However, pastoralist Shinile is one of
the poorest districts, ahead
only of Cherati, one of the three agro-pastoralist districts,
which occupy the lowest
positions in the income ranking. Agro-pastoralists are by far
the poorest communities in
Somali Region, suggesting that the income and food
generated through farming does not
offset the risks of keeping livestock in a drought-prone
region, probably because agropastoralists are sedentary and therefore less mobile than
pastoralists.
On the other hand, these income figures exclude households
reporting zero cash income in
2004. When these are included (i.e. calculating average
income across all households),
Gashamo drops from first to fifth wealthiest district, behind
all three farming districts. This
highlights the devastating impact of the 2004 drought on
household incomes in Gashamo,
and confirms that the source of pastoralist wealth is also the
source of its vulnerability. Most
15 Female focus group participant, rural Gashamo.
47
Table 4.4 Gross daily revenue from livestock sales in four

Somali markets, 2005 (Birr)
Animals Tog
Gode Hartisheik Jigjiga Total (Birr)
Average daily
Wachale
number sold
Bulls 606,305 197,863 77,029 14,187 895,384 569
Cows 36,991 9,848 37,730 22,398 106,967 124
Goats 10,468 35,626 36,177 7,540 89,811 496
Sheep 18,844 25,641 44,810 14,894 104,189 550
Camels 10,117 13,444 64,440 22,478 110,479 66
Total 682,725 282,422 260,186 81,497 1,306,830 1,805
Source: Baulch and Umar (2005: 12, 35).
households in Gashamo suffered a dramatic loss of wealth
and income during 2004, and
many lost all of their primary source of wealth their
animals. One Gashamo pastoralist
said: We lost all our livestock during the drought. We do not
have anything to eat and
nothing to live on. Some families dropped out of pastoralism
altogether in 2004, either
temporarily or permanently, being unable to sustain a
livelihood without animals. The same
trend was observed during the 1999/2000 drought in central
Somali Region, when many
families were forced to abandon livestock rearing for their
living, and migrated to IDP
camps or informal settlements on the edge of Gode and other
towns, in search of food aid
or petty income from selling firewood and charcoal to urban
residents (IDS 2002).
Finally, livestock ownership is significantly gendered, with
men owning almost all large stock
(camels and cattle) and women owning mainly small stock
(goats and sheep). According to
some women: In Somali culture, women are not really
supposed to own animals; We are
pastoralists by name only: it is the men who are the actual
pastoralists since they own the
animals, not the women. Other women contradicted these
generalisations, arguing that
Somali women can and do own animals. Women have
always been able to own animals
when they are unmarried or divorced. If a woman owns

animals before marriage, they
remain hers after marriage; Things are changing, because
younger women are starting to
own animals.16 Nonetheless, womens perception that Somali men
are married to their
animals explains much of the ambivalence towards

pastoralism (discussed later in this
report) that many Somali women expressed in discussions
during this research.
4.2 Livestock marketing in Somali Region

The Ethiopia Customs Authority reported that a total of
41,966 live animals were officially
exported from all of Ethiopia in 2003/4, but an estimate made
for this project puts the
number of animals exported annually from two markets in
Somali Region (Hartisheik and
Tog Wachale) at between 133,641 and 272,787 animals 3.2
to 6.5 times higher than the
official national export numbers for one year earlier.1 7 This
discrepancy partly reflects the
high volume of informal cross-border trade out of the region.

Nonetheless, pastoralists and
traders are unable to sell as many animals as they would like.
During three months of daily
market monitoring in 2005, more animals were offered than
purchased every day, in all
four markets monitored (Baulch and Umar 2005: 11). This
pattern of persistent excess supply
is confirmed by the Bureau of Agriculture and Rural
Development: in Jigjiga livestock
market, the average number of animals offered for sale daily
during 2001/2 was 803, while
the number sold was 182. In 2003/4, the number of animals
offered each day had risen to
1,153, but only 249 were sold.
16 Female focus group participants, rural Gashamo.

17 Baulch and Umar (2005: 8), whose estimate is based on three months
of monitoring the four major
livestock markets of Jigjiga, Hartisheik, Gode and Tog Wachale,
between April and June 2005.
48
IDS
RESEA
RCH
REPOR
T 57
Table 4.5 Time-line for livestock traders in Somali Region,

19922005
Years Assessment
199293 Good years unrestricted trade, supply and demand
both high and reliable.
199495 Malable (Honeyed years): Dhirta malab ka
kacay (Honey poured out of the
bushes); it rained from the Gu until the Jilaal,
and money was plentiful.
1996 Qeyla Weyn (Big shout out): there was no water, so
everyone was shouting
out for external assistance a very bad year for
livestock and people.
1997 Biyo Badan (Plentiful waters): heavy El Nio rains
caused flooding in some
areas, but had only a minor negative impact on
livestock trading.
1998 Rift Valley Fever: livestock died, prices collapsed,
Saudi Arabia banned imports
from the Horn, exports through Somaliland
collapsed a terrible year for
traders.
19992000 Sima (Equaliser): severe drought around Gode
caused an increase in distress
sales by pastoralists, benefiting traders who
bought up livestock at low prices.
2001 Second outbreak of Rift Valley Fever and another
import ban by the Gulf
States; Government of Ethiopia authorities
started intervening against
contraband trade (border closures, confiscation
of vehicles, livestock and
commodities) these interferences with trade
caused great hardship to
pastoralists and traders.
20022005 Some Gulf states lifted the livestock import ban
but Saudi Arabia upheld it;
trade recovered partially as livestock were
smuggled into Saudi Arabia through
Yemen.
2002 Government of Ethiopia imposed a strict border
closure policy with Somalia
and Somaliland (February) and banned all

foreign-registered vehicles, to
eradicate contraband trade; this border closure
was relaxed later in the year
(August).
2003 Good year: sheep production was good and sales were
good; the exchange
rate between Somali Shillings and the US dollar
was also favourable for
traders.
2004 Tuur ku Qaat (Carry on the back): severe drought in
northern Somali Region
caused pack camels to die; many pastoralists
lost all their animals; livestock
traders were forced to abandon trading; many
businesses collapsed.
2005 Government of Ethiopia continued its war on
contraband, impounding
vehicles,
confiscating
livestock
and
commodities, ransacking markets, and
banning use of all foreign currencies inside
Ethiopia traders lost heavily,
businesses closed.
Source: Trade and marketing survey data.
Table 4.4 summarises the average daily revenue taken and
animals sold in four major livestock
markets in Somali Region, during a three-month monitoring
period in 2005. The scale of this
trade is impressive. Over 1,800 animals are sold every market
day, which equates to over
600,000 animals traded every year. Over 1.3 million Birr
changes hands daily in these four
markets alone, equivalent to over US$140,000, or US$50 million
annually.18 Given the
18 These estimates should be taken as crude approximations only. They
are extrapolated from a continuous
three-month monitoring period of just four markets, and are not
adjusted for seasonality or disruptions to
market activity. Total trade volumes and revenues are probably
substantially higher than our estimates, but are
also likely to display high inter-annual variability.
49
evidence of over-supply of animals, the potential market is

perhaps as much as four times
greater than at present.
On the other hand, Somali livestock producers pastoralists
and agro-pastoralists receive
only a proportion of this income. Because supply generally
exceeds demand and markets do
not clear, producers and traders are price-takers. Also, as
explained below, there are large
numbers of market intermediaries requiring payments for
their services, which erodes the
returns to those who actually rear these animals. In a sense,
therefore, there is no livestock
marketing in Somali Region. Rather, livestock owners (and
traders) respond reactively to
signals of demand from Gulf states and other export markets.
Similarly, riverine farmers in
Kelafo are responding to a demand for vegetables from
Somalia, by producing and selling as
many onions as they can. But there are no innovations by
producers in livestock or crop
marketing, no flexibility in trade routes, no search for new
markets. Prices are accepted
rather than negotiated, and producers have little power in the
market to influence either
prices or consumer demand. These characteristics of the
marketing system reduce rural
household incomes and introduce additional sources of
vulnerability to the livelihoods of
pastoralists and farmers in Somali Region.
The unpredictable and fluctuating returns to livestock trading
in Somali Region are shown in
the following time-line, constructed from interviews with
traders (Table 4.5), which suggests
that there have been equal numbers of good and bad years
since the early 1990s. Table 4.5
also highlights the multiple sources of risk that traders have
to face, which are discussed
later in this chapter.
4.3 The marketing system in Somali Region

This section describes the main commodities that are traded
into and out of Somali Region;
the main trading routes through and out of the region; and the
main actors in the
marketing system, including the role of women traders.
4.3.1 Commodities traded in Somali Region
Although most discussions of trade in Somali Region focus

on exports of live animals out of
the region to neighbouring countries and the Gulf states,
trading is vibrant and complex,
involving many more commodities than livestock, imports as
well as exports, and several
distinct trade routes. Our survey of trade and marketing
found that the most significant
commodity movements can be disaggregated into four
clusters.
1 Exports from Somali Region to neighbouring countries:
Livestock (1) especially cattle and sheep, through the
Somalia ports of Berbera,
Bosaso and Mogadishu to the Middle East.
Livestock (2) mainly goats and cattle, through
border towns to markets in Kenya.
Perishable food items animal products (milk, butter,
ghee), local cereals (millet,
sorghum), and vegetables (onions, tomatoes, potatoes),
mainly to Somalia.
2 Exports from Somali Region to other regions of Ethiopia:
Livestock (3) mainly goats, to export abattoirs based
in and around Addis Ababa.
3 Reexports from elsewhere in Ethiopia to neighbouring
countries:
Khat grown in the highlands of eastern Haraghe, to
the towns and trading centres of
northern Somali Region, and to the neighbouring
countries of Somalia and Djibouti.
Vegetables from Oromiya through Somali Region to
markets in Somalia.
4 Imports into Somali Region through the ports of Somalia
and Djibouti:
Non-perishable food items rice, wheat flour, pasta,
cooking oil, biscuits, sugar, tea.
Clothes new and used (second-hand).
Household groceries soap, kerosene, batteries.
Consumer durables torches, lamps, thermos flasks,
pots and pans, radios.
50
IDS
RESEA
RCH
REPOR
T 57
By far the most important export from Ethiopia by value is

khat, though most of the profits
from khat exports accrue to traders from Haraghe and
Somalia. Informal exports of live
animals, sold to traders by Somali pastoralists, generate the
largest income for Somali
Region itself. These animals serve the demand for meat in
Saudi Arabia and other Gulf
states, and the trade is highly seasonal, peaking around the
annual Haj. On the other hand,
there is limited movement of livestock from lowland Somali
Region to highland Ethiopia,
not even to large urban markets like Addis Ababa. The main
reason is differences in tastes:
Somali meat is valued by consumers in the Gulf states, but
not by Ethiopian highlanders,
who prefer highland breeds. Traders from Kebribayah told
us: Since there is no market in
Ethiopia we send our sheep and goats to Somalia.
Another large market exists for high-value vegetable crops
like onions and tomatoes in
Somalia, which at present is being partly met by farmers
from Oromiya whose fresh
vegetables are trucked across Somali Region to towns inside
Somalia, where wholesalers
and retailers purchase them for resale in towns like Galkayo
and Mogadishu. (Ironically, some
of this produce is re-imported into Ethiopian districts such as
Bokh from Galkayo, where it
is more expensive than just across the border!) Recently,
farmers in irrigated and riverine
areas of Somali Region, such as Gode and Kelafo, have
responded to this market
opportunity by cultivating more vegetables for cross-border
sale. More generally, these
lucrative markets outside Ethiopia offer essential incomeearning opportunities for
pastoralists, farmers and traders from Somali Region, since
local markets are much smaller,
and local purchasing power is lower. It follows that any
restrictions on cross-border trade
impose serious constraints on livelihoods in Somali Region.
While most exports from Somali Region are live animals and
perishable produce, most
imports are non-perishable food items and durable goods
(clothes, utensils, electronic

goods). New and second-hand clothes are sold on open-air
stalls and in permanent markets
like Taiwan so called because many cheap clothes on sale
are made in China in Jigjiga
town and Harar. Almost all of these imports are informal and
considered as contraband by
the Government of Ethiopia. As discussed below, the
governments hostility to cross-border
trade, both into and out of Somali Region, is a major source
of livelihood vulnerability.
4.3.2 Trade routes from Somali Region

There are several distinct and long-established trade routes
through and beyond Somali
Region that connect landlocked Ethiopia with the outside
world through several ports in
Djibouti, Somalia (including Somaliland and Puntland)1 9
Kenya. The orientation of these
and
trade routes is related to marketsheds that are defined by

clan territorial boundaries,
geographical features like access to water points and grazing
along the route, and distances
between purchasing centres and export markets or ports.
Economic considerations include
the price and availability of public or private transport,
various transactions costs incurred in
moving livestock and commodities across long distances, and
the prices offered by traders.
An additional factor that can modify these trade routes is
conflict or aabsi fear and
tension caused by insecurity and latent conflict between
groups. The corollary of this
observation is that certain clans control and dominate each
route, and relations between
clans along the route, or between members of the same clan
on either side of the national
border, determine how smoothly traffic flows and how secure
the route is.
There are three trade corridors out of northern Somali
Region, known as the Issa, Berbera
and Bosaso corridors. A fourth trade route, the riverine
trading area, links central Somali
Region with Somalia and northern Kenya, and southern
Somali Region is connected to
northern Kenya through the Zone Liban.
19 Somalia has effectively been a state without a government since the
civil war ended with Siyad Barres
overthrow in 1989. The Republic of Somaliland is an unrecognised
self-proclaimed state in northern Somalia
that declared its independence in May 1991. Its capital is Hargeisa
and the main port is Berbera. Puntland, in
north-eastern Somalia, declared it self an autonomous self-governing
state within Somalia in 1998. Unlike

Somaliland, it is not seeking independence from Somalia but favours
a decentralised federal system. It s capital
is Garoowe and the main port is Bosaso.
51
1 Issa corridor. This trade route traverses the Shinile Zone in

northern Somali Region,
and links Addis Ababa with Djibouti. Ethiopias oil is
imported through the port of
Djibouti and transported through Shinile to Addis
Ababa and other highland towns.
People from Shinile do not appear to profit as much as
might be expected from this
good road and well-developed public transport system,
even though the Issa corridor is
known to be a major route for the contraband trade into
Ethiopia.
2 Berbera corridor. Commodity movements between Somali
Region and the Somaliland
port of Berbera follow two routes: along the Hargeisa
road to Jigjiga and Harar, and
along the Burao road into the grazing areas of eastern
Jigjiga and Degahbur zones.
2a Jigjiga route. The largest volume of trade between Somali
Region and neighbouring
countries occurs on the route linking Harar and
Jigjiga to Berbera and Somaliland
towns like Hargeisa and Burao, via Hartisheik and
the border town of Tog Wachale.
Buses and trucks carry khat, fresh vegetables, grain
and livestock (especially cattle) to
Somaliland, and return bringing staple foods, clothes,
and various contraband goods
such as electronics.
2b Hawd route. This route serves rural zones and small towns
in north-central Somali
Region, such as Aware, Gashamo, Degahbur and Fiq,
and is oriented towards rural
livestock producers and their consumption needs.
Major exports include small stock,
charcoal, gums and resins, while imports are
dominated by staple foods, groceries
and household items.
3 Bosaso corridor. Areas served by this route include the
border districts of Boh, Galadi and
Warder, in Warder Zone. These areas have a
commercialised pastoralist population,
which holds large livestock herds. Economically and
socio-culturally the zone is oriented
towards Puntland, especially the port of Bosaso and the
important desert town of
Galkayo. People are from the same clan on both sides of
the border, there are daily bus
services between Warder and Puntland, and the Somali

Shilling is the common currency.
4 Riverine trading area. There are several less well-defined
trade routes out of central
Somali Region, which connect Gode, Fiq, Korahe and
northern Afder Zone with
markets in Somalia and Kenya, or link up with the
major Berbera and Bosaso trade
corridors. Gode town the former capital of Somali
Region is the main trading
centre in this area, attracting imports of consumer goods
and foodstuffs for its urban
residents, while farmers along the Shabelle river export
vegetables and fruit to
Somaliland and the city of Mogadishu. Cattle from
these zones are trekked north to
Berbera, while sheep and goats are trucked to Bosaso.
5 Zone Liban. Ethiopian Somalis call the entire southern
part of Somali Region Zone
Liban, including Liban Zone itself and neighbouring
districts such as Cherati in Afder
Zone, which share similar clan backgrounds and
dialects. Zone Liban is oriented
toward northern Kenya as well as southern Somalia.
The people of southern Somali
Region engage in cross-border trade with the Kenyan
towns of Mandera and Moyale,
as well as the city of Mogadishu, but not much with the
port of Kismayu in recent
years, due to civil strife in southern Somalia.
Despite this evidence of a dynamic and complex network of
trade routes and commodity
flows into and out of Somali Region, some important features
of this trading system need
to be noted. Firstly, the domination of specific clans and
market operators along each trade
corridor means that they are not substitutable. If the route
used by a pastoralist or trader
becomes inaccessible (e.g. due to conflict or insecurity) or
the market collapses (e.g. during a
drought, or because of a government clampdown on
contraband trade), there is often no
alternative. This reliance on a single marketing channel
makes everyone who depends on
this channel extremely vulnerable to a sudden change in
market conditions.
During the drought of 2004, for instance, pastoralists in
Gashamo complained that traders
no longer came to purchase their livestock, so they were
unable to offload their animals
52
IDS
RESEA
RCH
REPOR
T 57
even at low prices (distress sales), and livestock mortality

was higher than it would
otherwise have been. This would appear to contradict
evidence from famines elsewhere,
when traders often stand accused of exploiting livestock
owners by buying up animals
cheaply and making excess profits. The explanation that
traders in Somali Region gave us
explained the paradox in terms of market rigidities. When
livestock conditions deteriorated
in Gashamo and neighbouring districts during 2004, traders
found it increasingly difficult to
sell these animals to export buyers, who favoured better
quality animals from parts of the
region that were unaffected by the drought. Not only were
drought-affected livestock
owners unable to find a market for their animals, livestock
traders were unable to move
into other markets, since these markets were already served
by other traders.
This symbiotic relationship between pastoralists and traders
is beneficial to both parties in
good times, but both suffered a drastic collapse in incomes
during 2004. As a consequence,
not only were many pastoralists forced to abandon
pastoralism because of the drought, but
many traders were also forced to abandon trading altogether.
4.3.3 Market actors

Marketing in Somali Region is much more complicated than
the neoclassical model of a
producer selling to a consumer at a negotiated marketclearing price, perhaps with a
wholesaler or retailer as market intermediaries. Partly
because live animals are often involved,
partly because much of the trade is informal even illegal
and crosses national boundaries,
and partly because of the complex interrelationship between
trade routes and clan territories,
there are a large number of market actors between primary
producers and final consumers.
The result is a marketing system that is far from anonymous
and impersonal, but instead is a
network of personal and clan-based relationships, with each
actor dependent on the others

in a way that both protects and constrains their options and
opportunities. Although every
actor plays a vital role in linking producers with consumers,
the multiplicity of intermediaries
and the personalised nature of these relationships certainly
raise transactions costs, and can
increase the vulnerability of the many livelihoods that depend
on each marketing chain. The
most important of these intermediaries are described here.
Shirkad (company): six large companies operate
out of Hargeisa and Bosaso, where
they own large warehouses as well as employing
wholesalers and purchasing agents who
are active in the ports, the Middle East and Somali
Region. Often the shirkad provides
loans to traders to purchase livestock in Somali Region
and bring the animals to the
ports, for export to Saudi Arabia and other Gulf states.
They also import shiploads of
food, groceries and electronic goods, which they sell on
to traders and shopkeepers who
supply markets and towns inside Ethiopia. In addition,
the shirkad often has other business
interests, such as money transfers, telecommunications
and construction.
Ganasade (trader): just below the shirkad level are
large-scale traders known as
ganasade. There are currently several ganasade
operating as livestock traders in Jigjiga
town, and many more based in large livestock assembly
markets such as Tog Wachale
and Hartisheik. Importantly, most of this trade is
formalised these traders usually have
a modified Letters of Credit arrangement, which allows
them to export livestock legally
out of Somali Region.
Jeble (pocket): the name jeble reflects their function
they are the pockets that
provide the capital with which livestock are purchased,
other market actors (e.g. dilala)
are hired, and transporters are paid. Often their capital
has been loaned to them by a
shirkad to fill a specific purchasing order. Other jeble
live in rural areas and are known as
urursade (collectors), because they buy small
numbers of animals in various local
(bush) markets and collect them until they have
sufficient numbers to transport
them to secondary or assembly markets. Jeble also
facilitate access to markets for
pastoralists, sometimes providing clan members with
loans in return for guaranteed

future sales of animals, or sending vehicles into rural
areas to collect sheep and goats
and transport them to market, for a percentage of the
selling price.
53
Dilal: these middlemen are found in marketplaces

throughout the Horn of Africa,
facilitating business deals and transactions, especially
livestock sales. Each clan has their
own dilal in every major market who works exclusively
for their clan, which means that
there is a strong basis of trust between livestock sellers
and their dilal. It also means
that there is little reason for competitiveness between
dilala, since the volume of
business each handles is independent of the volume
handled by other dilala serving
different clans. Dilala are especially respected for their
knowledge of animals: they can
estimate the weight and value of live animals by sight,
and they follow trends in market
demand and supply closely. They also negotiate prices
with traders, and handle the
money once the deal is concluded. According to one
livestock trader:
You cannot go into a market to buy animals without
the help of a local dilal. How
can you guarantee that the animal that you have
bought is not a stolen animal
whose owners will come after you the next day? The
dilal will look at the bulls that
have come into the market their size, condition,
colour. He will then make an
assessment of their worth, and then arrives at the
average price that the animals
should be quoted at.2 0
Dulsar (put on top): these are small-scale traders

who buy animals and resell them for
a small profit, often on the same day and in the same
market. Alternatively, they might
purchase animals from pastoralists who need to travel
immediately back to their village,
and keep the animals for a few days until prices rise.
Gesgara (punch each side): these are brokers who
step in when negotiations
between two dilala are deadlocked. The gesgara
shuttles between the two dilala until a
compromise is reached hence the nickname punch
each side when each dilal pays
the same amount to the gesgara as a facilitation fee.
Alamadiye (branders): the alama (sign) is the
owners unique mark, found on every
animal. Alamadiye are present in every major market,
branding animals immediately
after purchase.
Cawse (grass-workers): these people keep stocks of
grass and hay at the salesyards,
which they sell to pastoralists and traders to feed the
animals that are being held in
pens before being sold.
Garaa (beaters): they operate in cattle and camel
markets, ensuring that bulls do not
fight one another or injure themselves while penned,
being loaded into lorries, or in
slaughterhouses.
Raii (followers): these are the trekkers who move
animals on foot between
salesyards, which can be a distance of a hundred
kilometres or more and a journey of
many days.
4.3.4 Women traders

Women have always been actively engaged in trading in
Somali Region, and their numbers
are rising. Itinerant women traders have acquired mildly
derogatory labels khararaf and
asha kushi because of their recent proliferation and their
willingness to accept tiny profit
margins. Nearly all retailers of khat are women, especially in
urban areas. The wealthiest
business person in Jigjiga town is a woman who exports khat
to Somaliland and abroad. In
the livestock sector, most of the sheep and goats traded in the
region are bought and sold
by women, who are also entering the cattle trade, though
mainly in local markets rather
than long-distance trading. There appear to be few cultural
impediments to Somali women
engaging in trade (despite this being a Muslim society,
women are not prohibited from
working outside the home), except those related to the
gendered division of household
reproductive roles such as inability to travel far from home,
because of child-care duties
and other domestic responsibilities.
20 Livestock trader, Tog Wachale market.
54
IDS
RESEA
RCH
REPOR
T 57
Box 4.1 Women livestock traders in Degahbur

There is a group of 14 women who live in Degahbur
and have been bringing large
stock from their home area to Jigjiga market for 56
years. They purchase animals in
Degahbur and Fiq Zone. It takes 710 days to collect a
marketable herd. On every trip
each woman brings in approximately 20 animals. Two
raii trek the animals to Jigjiga,
charging 40 Birr per head, where another 10 Birr is
paid per head for a dilal. Return
fares to Garbo for the traders are 100 Birr. They also
spend about 10 Birr per day on
living expenses. Total taxes and market fees paid
average 10 Birr per head. After
handing over the animals to a khidmad (trustee) in
Jigjiga, who facilitates their sale at
a charge of 10 Birr per animal, the women return to
Degahbur. The fare back to
Degahbur is 30 Birr. Large bulls are sold for 2,300 to
2,500 Birr, while the smaller
bulls fetch 800900 Birr. Profit margins are tight, and
the maximum profit made
per animal is 200 Birr. However, losses can be
incurred in the same range up to
200 Birr per animal depending on market conditions.
Source: Interview with a woman trader, Hartisheik
market.
Most of the other constraints that women traders face are
practical difficulties in raising
working capital, or risks to their personal safety if they travel
unaccompanied. One solution
to the working capital constraint is through traditional
savings and credit groups. Many
women participate in these groups, paying in small amounts
of cash and withdrawing lump
sums at occasional intervals, which can be used to purchase
commodities or livestock for
trading activities. Women are said to save more than men,
because they chew khat less and
do not expend their resources on patronage politics. This last
point gives women a unique
advantage over men in trading. Since women tend to be only
peripherally involved in interclan conflicts, they can work with other women across clans
to build solidarity and resolve
conflicts behind the scenes. Women cross clan boundaries
more easily than men. This is
an important attribute in trading, since women who enter
rival clan territories are seen as
less threatening than men. Box 4.1 shows how women work
together to scale up their
trading activities to a viable level.
Women are also becoming active as various categories of
market agents. Of about 70 jeble
currently active in Tog Wachale market, 25 are women, and
there are equal numbers of
male and female dilala. In some markets, most of the dulsar
(who buy and resell animals the
same day for a small profit margin) are women. Women are
also heavily engaged in the
marketing of animal by-products milk, butter, ghee, hides
and meat. In the past 10 to 15
years, women have come to dominate the abattoirs and
butcheries in Jigjiga and Hargeisa,
especially for sheep and goats.
4.4 Livestock prices

The main determinant of pastoralist and trader incomes is
livestock prices, but prices of animals
vary greatly between markets and over time. Some factors
that influence price determination
are related to the condition of each animal, such as its breed
(e.g. the dominant and preferred
breed of sheep in Somali Region is the Berbera Blackhead),
age, sex, weight and preferred
traits such as colouring and body shape, which affect
breeding potential.
A second set of factors that affect livestock prices is market
conditions. Dominant among
these are temporary rises or falls in demand or supply (e.g. a
collapse of supply due to
animal disease or drought-related mortality, or a collapse in
demand due to the Saudi ban
on Somali livestock). Spatial variations in livestock prices are
also observed across the region,
due either to differences in transport and transactions costs on
different trade routes, or to
market inefficiencies such as traders taking higher profit
margins in monopolistic markets.
Another complex factor influencing prices is seasonality,
which affects the condition of
animals and is associated with particular patterns of market
demand. In general, prices are
lower during the dry seasons, when animals lose weight, and
higher during the rainy
seasons, when the animals are fatter. Specific trends, season

by season, are as follows.
55
Jilaal: during the hot dry season (January to March),

livestock are cheap because
pasture and water are scarce, animals are susceptible to
diseases, and pastoralists are in
great need for purchased food, so they sell their animals
at almost any price.
Gu: during the first rainy period (April to June),
livestock prices rise as pasture and
water are available. Animals are calving and are in milk,
and pastoralists are under less
pressure to purchase food, so the volumes offered for
sale are low. Instead of selling,
many pastoralists are buying animals for fattening,
pushing up prices.
Hagaa: during the dry season (July to September),
prices fall as animals lose weight
and exports are limited by shipping constraints, some
caused by strong winds blowing
in the Red Sea.
Deyr: during this second rainy season (October to
December), prices start to rise again.
From the pastoralist perspective, livestock marketing seasons
are either good or bad,
according to supply and demand conditions rather than
rainfall though the two are
related. Good seasons are Iga ibi (literally, please sell to
me!), when demand for livestock is
so high that it becomes a sellers market, and traders offer
high prices to induce pastoralists
to sell as much as 250 Birr for a sheep. The second season
is Iga ibso (please buy!), when
demand for livestock is low, and pastoralists beg traders to
buy their animals. During
droughts, the price of a sheep can fall as low as 60 Birr.
Pastoralists who are forced to sell
at these distress prices often say: Wan tuurey (I threw
away the animal).
Apart from climatic seasonality, another powerful seasonal
feature of livestock marketing is
religious festivals and holidays. Most significantly, demand
for sheep and goats peaks
between the Islamic months of Ramadan and the Idd Arafa
the exact timing of which
varies from year to year when prices also peak. During the
Idd ul Haj, pilgrims to Mecca
and Moslems who remain at home are required to slaughter a
sheep. Large numbers of live
sheep are brought into Saudi Arabia, mostly imported from

greater Somalia (including
Somali Region) and Sudan.
If all the above factors are known, market analysts can
predict sales volumes and price
trends by extrapolating from trends in previous years,
allowing for inflation and exchange
rate fluctuations. However, a final set of factors is political
interference, which complicates
the analysis and introduces distortions that are not found in
well-functioning markets.
Because of its unpredictability, political intervention raises
marketing risks for all actors in
the marketing chain. In Somali Region, political factors
include border closures and
government clampdowns on contraband trade, including
confiscating commodities. Some of
these political factors are examined in the following section.
4.5 Marketing constraints in Somali Region

Livestock marketing generates substantial revenues and
provides employment for large
numbers of people in Somali Region. However, returns to
livestock marketing are variable
across markets and seasons, and extremely unpredictable at
all times. This reflects the
paradox of wealth plus vulnerability that is identified in this
research as a defining feature of
pastoralist livelihoods in Somali Region: trading can yield
high incomes for all involved, but it
is also highly risky. Some of these risks such as drought
and disease are endemic to
livestock production and marketing in arid and semi-arid
environments everywhere. Other
risks are a function of the complex nature of mobile, clanbased societies and their
relationship to key productive natural resources (livestock,
water, pasture). A third set of risks
relates to policies and politics, which are generally disruptive
and undermining, as
bureaucrats and politicians attempt to control and formalise
informal trade.
The pastoralist production system, developed over literally
hundreds of years, has found ways
of reducing some of these risks. Construction of wells and
berkad provides access to water
during dry seasons and droughts. Migration with animals in
search of pasture maximises
utilisation of the rangelands. Veterinary services and drugs
though scandalously
56
IDS
RESEA
RCH
REPOR
T 57
undersupplied in Somali Region reduce livestock losses to

disease. Disputes between clans
over access to resources do occur, but are controlled through
conflict resolution
mechanisms and informal institutions such as compensation
payments.
Similarly, the pastoral marketing system has evolved
mechanisms for mitigating some of the
risks associated with livestock trading. Informal trade
operates with a complete lack of
formal finance and credit, and in the absence of legally
enforceable contracts. As noted
above, trade routes are controlled by clans based on territorial
boundaries, and key
marketing agents such as dilala tend to come from the same
clan as the seller, which
introduces a high degree of trust into market transactions.
Because of the long distances
and security hazards on many routes, control over
commodities is often passed on to many
different agents along the route, with trekkers and traders
specialising in working in familiar
territory which they understand well. Ownership of the same
livestock can change hands
several times between the primary producer and the final
consumer. For example, a bull
reared in rural east Hararghe might be bought and sold five or
six times within a few weeks
before being slaughtered in Saudi Arabia: first it is taken to
Babile market where a trader
buys and resells it in Jigjiga market, where another trader
buys it and pays a trekker to walk
it to the border, then a third trader buys the bull and sells it to
a Somaliland trader in Tog
Wachale, who takes it to Berbera port where it is shipped to
Yemen, bought by a Yemeni
trader who transports it to Saudi Arabia, where it is sold to a
Saudi trader who sells it to a
butcher or supermarket chain. This division of labour
serves the intended purpose of
spreading the costs and risks and benefits of trade among
many market actors, but it
also raises transactions costs and reduces the returns to
producers.
The impact of drought risks on trade is examined elsewhere
in this report. Four main

sources of risk in informal marketing are examined here: (1)
conflict and insecurity; (2) Saudi
import bans; (3) Government of Ethiopia policies; (4)
exchange rate fluctuations and
currency bans.
4.5.1 Conflict and civil insecurity

Traders interviewed in eastern parts of Somali Region were
unanimous that 1991 and 1992
were the most difficult years for trading in their lifetime,
because there was no functioning
government in either Somalia or Ethiopia at that time.
Somalias civil war was coming to a
head, and the Derg regime had just been overthrown in
Ethiopia. During this period of
lawlessness and escalating violence, there was widespread
looting of property, especially of
vehicles and livestock. Many traders lost all their property to
armed bandits or militia. Some
beneficial effects from the resultant dispersal of Somalis
throughout the world were
mentioned, however. Many individuals who fled to the
Middle East and the West found new
livelihoods and started supporting relatives back home. Many
sent cash remittances, some
sent pick-up trucks and lorries to support trading activities in
the pastoral areas, while others
provided start-up capital or working capital. Before the
Somali civil war there were only ten
vehicles in Bokh District, but by 2004 every trading centre in
Bokh had five or six locally
owned vehicles. In Jigjiga town, members of the Somali
diaspora have financed several
hotels, residential and commercial buildings, and businesses,
especially trading enterprises.
4.5.2 Saudi import bans

The biggest and most lucrative market for Somali livestock
producers and traders is the
Middle East, especially the Kingdom of Saudi Arabia, which
imports literally millions of
sheep, goats and cattle every year, especially during
Ramadan and the Haj. Within the last
ten years, however, this market has twice been severely
disrupted by the imposition of a
ban on livestock imports from the Horn of Africa. The first
ban was imposed by Saudi
Arabia, for just over a year between February 1998 and April
1999. The reason for this ban
was an outbreak of Rift Valley Fever in Kenya, Somalia and
Ethiopia, which caused
widespread livestock mortality, especially of sheep. The
effects on trade were more dramatic

than the mortality itself. Livestock prices fell throughout East
Africa, and exports through
the port of Berbera in Somaliland about half of which
originate from Somali Region
57
Box 4.2 Impact of the 1998 Saudi livestock ban on

Somali traders
I have been a trader for 12 years, mainly in importing
goods like food and clothes,
and exporting livestock. I used to export livestock to
Saudi Arabia, but now I only take
animals to Hargeisa and Hartisheik. The worst year for
me was 1998, when I bought
260 animals at a cost of 180 Birr each. The Saudi ban
caught me before I exported
the animals, and I had to hold them for one year at an
additional expense of 70 Birr
per animal. After losing some of the animals to
disease, and realising that the ban
may take longer to be lifted, I trekked the animals to
Djibouti hoping that I could
export them from there. Eventually I had to sell them
at a loss in Djibouti.
Source: Interview with a trader in Hartisheik.
immediately halved, from over three million head in 1997 to
1.5 million head in 1998,
resulting in estimated revenue losses of US$100 million
(FSAU 2004). One traders
experience of the 1998 livestock ban is presented in Box 4.2.
The fact that the livestock export market did not completely
collapse in 1998 was due to
the fact that other Gulf countries continued to import East
African livestock, and many
animals were in fact smuggled into Saudi Arabia through
Yemen. Although the embargo
was partially lifted in 1999, a second outbreak of Rift Valley
Fever actually in southern
Saudi Arabia, but imported animals were blamed provoked
a second and more
comprehensive ban in September 2000. Six Gulf states
(Saudi Arabia, Bahrain, Oman, Qatar,
the United Arab Emirates and Yemen) banned all livestock
imports from eight East African
countries, including Ethiopia. During 2001, most of the Gulf
states independently
announced that they were lifting the ban, but Saudi Arabia
by far the largest market for
live animals from East Africa had yet to do so by late 2005.
The consequences of both livestock bans can be clearly seen
in Figure 4.1, which shows
dramatic collapses of livestock exports through Somaliland in

1998 and in 2001. The first
shock followed three years of exports exceeding three million
head; the second followed
two years when exports had exceeded two million head. The
figures also reveal that
recovery of trade after each shock is only partial export
volumes in the two years after
each embargo was lifted remain substantially lower than in
the two years before each ban
was imposed. One explanation is that other suppliers (e.g.
from Australia and New Zealand)
seized the opportunity to enter the lucrative Saudi Arabian
market, especially for live sheep,
thereby displacing east African suppliers.
A second reason for the incomplete recovery of livestock
exports is that many traders and
agents working in East African markets were driven out of
business by losses incurred when
the bans were imposed. This was exacerbated by the market
rigidities discussed above,
which means that traders are locked into particular trade
corridors and unable to switch to
alternative routes or markets. This consequence of the ban
was confirmed by a large trader
interviewed in mid-2005.
The biggest threat to our business as livestock traders is
the tremendous effect of the
governments in the Gulf when they exercise their power
by locking us out of their
markets. So far nothing equals the ban placed by Saudi
Arabia on importing of
livestock from the Horn of Africa. Two-thirds of the
livestock traders quit the business
after the ban. Many were extinguished by heav y losses
Today the situation is
improving, as Yemen and the Arab Emirates take our
animals. But the kind of livestock
export volumes that were realised before the ban have not yet been realised since.2 1
A related and ongoing problem faced by traders in the Horn

is that marketing costs have
risen, particularly for those engaged in the risky (but highly
competitive) business of
smuggling animals into Saudi Arabia through Yemen. This
circuitous route involves hiring
extra market agents, payment of bribes to two or three sets of
border guards, and higher
21 Interview with owner of Farole Livestock Enterprises.
58
IDS
RESEA
RCH
REPOR
T 57
Figure 4.1 Annual livestock exports from Berbera and Bosaso

ports, 19942004
3500
Camels
Cattle
3000
Shoats
2500
2000
1500
1000
500
0
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Source: FSAU (2004: 11).

risks of animals not being sold in terminal markets.
Sometimes boats are sunk in the Red
Sea, with the loss of all animals on board. These additional
costs and risks have eroded
profit margins to such an extent that many traders can no
longer make a living.
A final observation about the livestock ban is that Somali
pastoralists and traders received
no support from the Government of Ethiopia, in negotiating
with governments in the Gulf
to lift the ban. The Government of Kenya, by contrast,
worked hard to re-open its livestock
export markets, and achieved some success by approaching
the Red Sea Livestock
Commission and several Gulf states directly. It might be
assumed that the Ethiopian
government has no interest in getting this ban lifted, given
their hostility to informal exports
and its many attempts to regulate or eradicate this trade, as
discussed below.
4.5.3 Government policies with respect to traders

The Government of Ethiopia has persistently attempted to
regulate the marketing of
livestock and commodities into and out of Somali Region. It

has decreed informal crossborder trade to be illegal both exports of livestock and
imports of contraband goods.
The official reason for these policies is that Ethiopia is
losing large numbers of livestock
and the Ministry of Finance is losing substantial customs
revenue from export taxes and
import duties. Ethiopias Customs Proclamation (Article 74)
states that anybody found using,
holding, hiding, selling or distributing goods which have not
been subject to customs
procedures shall be fined the equivalent of the appropriate
duty, the goods will be
confiscated and individuals could face up to five years
imprisonment.22 Traders believe that
this proclamation effectively criminalises trade, which they

believe should be liberalised. We
are tired of being treated like criminals, when we are just trying to
make a living.2 3 Since
the behaviour of border officials is often unpredictable and

arbitrary, and to avoid paying
taxes, traders often avoid crossing at official frontier points.
Incidents of harassment of
traders by Ethiopian officials the military, the police and
the Customs Authority have
been reported from all parts of Somali Region, including
raids on urban markets as well as
along rural trading routes, even extending to foreign traders
bringing commodities into
Ethiopia.
22 The Ethiopian Customs Authority includes as one component of its

Mission Statement: It controls and
prevents illegal trades that are detrimental to social well-being and
economic development
(www.ethiomarket.com/ecua/index.htm, accessed 28 December
2005).
23 Male trader, Hartisheik market.
59
Box 4.3 Government harassment of livestock traders,

20045
In March 2004, two businessmen from Jigjiga zone
bought 200 bulls from Gode,
which they started trekking to Hartisheik market. On
the road, however, these bulls
were confiscated by government officials, ostensibly
for security reasons.
In May 2005, 300 bulls heading towards Bosaso
belonging to Bokh-based traders
were seized by security forces near Shilabo. As this
was during a period when all
overland transport along the Bosaso corridor was
halted, panic spread along the
route.
Three traders in northern Kenya decided to bring about
300 sheep and goats to
Moyale market, where demand for livestock from
buying agents of the El-Fora
trading company was high. Having obtained the
necessary livestock movement
permits in Kenya, they drove their flock to an official
frontier crossing-point. They
were stopped at the border by the Ethiopian Customs
Authority, arrested as livestock
smugglers, and locked up for three weeks. Their
animals were confiscated and
auctioned off in Moyale by the Ethiopian officials.
Source: Interviews with traders in Somali Region.
In February 2002, for instance, the Ethiopian authorities
closed the border posts with
Somalia and Somaliland, and deployed security forces along
the length of the border. Entire
lorry-loads of livestock intended for export, as well as
commodities being imported into
Somali Region, were confiscated without compensation.
Because of the dependence of
many pastoralists and urban residents on imported food, the
banning order on food imports
was soon lifted, but for other imports and for livestock
exports the border remained closed
until August 2002, causing great damage to the regions
economy. At the same time, all
vehicles that were not registered in Ethiopia were banned
from operating inside Somali
Region. This has resulted in large numbers of animals
destined for export being trekked long

distances, instead of being trucked in vehicles from
Somaliland, as before.
A more recent action against contraband trade was launched
by the Government of Ethiopia
in mid-March 2005, when more than 50 trucks bringing food
from Bosaso in Puntland were
seized and impounded at police stations and military
compounds in Warder, Korahe and
Gode zones. In Warder, over 30 trucks were seized, each
loaded with more than 20 tons of
food items wheat, rice, pasta, cooking oil and sugar. This
immediately removed at least 600
tons of food from markets in Somali Region. At the same
time, many trucks loaded with up
to 360380 sheep and goats each, for export through Bosaso,
were also impounded. All
movement of trucks, commodities and livestock along the
Bosaso trade corridor virtually
ceased. Since Bosaso is the main port for importing staple
food items into parts of eastern
Somali Region that do not cultivate crops, and since exports
of livestock finance the imports
of these foods, this disruption to market supplies and the
livestock trade was extremely
serious for household food security in these zones. By early
April, the 80+ trucks were still
impounded, food stocks in many markets were running low,
and food prices started to rise.
Then on 9 April, a combined force of Ethiopian military,
police and local security officers
ransacked the Taiwan markets of Jigjiga and Harar towns,
confiscating large stocks of
imported clothes and other items classified as contraband
goods. Panic gripped the region,
and prices of basic commodities spiralled as people stocked
up in anticipation of further
disruptions to market supplies. Instead, on 14 April the food
trucks were finally released, and
prices started to fall as supplies were delivered to grocery
stores and rural markets.
Traders in Hartisheik were among the worst affected by this
action. As one trader said: The
little light we had in this Hartisheik market has been just been
turned out by the Ethiopian
government, in its new war on contraband2 4 Some more recent
incidents (from many
collected in our fieldwork) of harassment of traders and

confiscation of their property are
summarised in Box 4.3.
24 Woman trader, Hartisheik market.
60
IDS
RESEA
RCH
REPOR
T 57
Box 4.4 Case study: contraband trader in Jigjiga

I trade in bulk food, which I bring into Jigjiga at
reasonable prices. Yes, my business
can be said to be contraband, which is extremely risky,
as you know. What can I do,
when trying to bring in anything legally is made so
impossible? I do not see the
Finance as anything except legal bandits, as they
search for us day and night, and will
use whatever weapons they have to loot us of our
property. If they stop us, they will
take everything we have, including the vehicles that
we are using. If they do not
manage to stop us, they will shoot at us. Many of our
people have died in the bullets
that they rain on us mercilessly, there are many injured
and maimed people living in
Jigjiga who have had these experiences. The regional
government is developing a
system of licensing, whereby we are allowed to import
food items worth up to US$
10,000. But this system is not recognised by the
Customs Authorities in Jigjiga who
report directly to Addis Ababa. Whatever licence the
regional authorities give us, we
are hunted down along the routes to the border, and
force is used to take away our
property. After a number of months, all the confiscated
property is sold into the local
and highland markets by government, at lower prices
than market rates. All the
money realised then is declared by government to be
state income the original
owner is never compensated at all!
Source: Interview with a trader in Jigjiga town.
Apart from the direct losses that seizure of property and
disruption to markets causes
traders, the uncertainty created by the unpredictability of
government policies and
interventions raises the risks that traders face and discourages
some from continuing.
Many businesses have collapsed under the weight of
worry as their property runs the

risk of confiscation by the Finance authorities, and the
gambling nature of the trade is
beyond the capacity of many. One can only sustain a
small number of total losses of
ones property before one has to exit the business.2 5
The weight of worry and the threat of total losses of ones

property have real impacts on
profit margins, which needs to be taken into consideration,
given the complaints often levelled
against traders that they exploit pastoralists and farmers by
extracting excessive margins. A
graphic exposition of the physical dangers as well as
economic risks that contraband traders
face in Somali Region is provided by a food trader in Jigjiga
market (Box 4.4).
4.5.4 Exchange rate variability and currency bans

Several currencies are used in Somali Region, and exchange
rates between them are
variable and unpredictable. The interconnectedness of the
economies of Somali Region and
Somalia (including Somaliland and Puntland) is revealed by
the fact that the common
currency in much of rural Somali Region is the Somalia
Shilling, while the Somaliland
Shilling is also used in some markets near the Somalia
border. The United States Dollar is
used for large deals, and is the basis for calculating exchange
rates.2 6 Most of the food and
goods sold in Somali towns are priced in US dollars, which

has effectively dollarised the
local economies. The fall of the dollar in 2004 had a dramatic
impact on both Somali
currencies, devaluing them and resulting in loss of wealth for
those holding shillings. Further
devaluations occur whenever warlords in Somalia print new
banknotes. The effects of these
devaluations are transmitted into Somali Region because
livestock exports to Bosaso and
Mogadishu are priced in Somali currencies. The vulnerability
that currency fluctuations
introduce to trading activities is reflected in one traders story
(Box 4.5).
25 Male trader, Hartisheik market.

26 In late 2004, the following exchange rates prevailed in rural Somali
Region: US$ 100 = Somaliland Shillings
584,800 = Somali Shillings 1,500,000 = Ethiopian Birr 890;
Ethiopian Birr 100 = Somaliland Shillings 68,000
= Somali Shillings 150,000.
61
Box 4.5 Case study: impacts of currency fluctuations

on Somali traders
I used to work for the Ogaden Welfare Society, which
was the largest local NGO
working in the Somali Region. In 2002, the OWS was
closed down by the
government for security reasons, and senior officials
landed in prison. Along with
many junior workers, I lost my job. When I found
myself jobless, I started trading,
using my small savings. I used to buy cattle from the
southern parts of Somali
Region, which I would transport to the border towns of
Moyale and Mandera in
Kenya. I would carry back small baggage materials
on my way back. However, I
encountered great difficulties due to the fluctuating
value of the currencies and their
rising or lowering exchange rate. The sheer amounts
by which these changes would
happen, and the fact that I had no idea why and when
one particular currency would
go up or down made my work similar to wild
gambling! The margins that I was
operating at were quite narrow and sharp. After a
successful sale in Kenya, I would
have large amounts of cash in local currency in my
possession. Items that I would
purchase for resale in Ethiopia by then may have
become more expensive in Kenya!
Somehow, I had to transfer my money back into
Ethiopian currency without losing
the value of what I had.
Source: Interview with a trader in Moyale.
In May 2005, the Government of Ethiopia banned the use of
any currency other than the
Birr for economic transactions inside Ethiopias borders. The
intention was to stamp out
trade conducted with Somaliland Shillings or Somalia
Shillings in Somali Region. Since the
policy was introduced overnight, Ethiopians holding these
currencies incurred heavy losses,
and traders working with the Somalia currencies stopped
coming to areas where the policy
was enforced effectively. In parts of central Somali Region
such as Korahe Zone, the

livestock trade declined dramatically and many households
and businesses were severely
affected. Shilabo District was particularly badly hit. Before
the currency ban and the war on
contraband, approximately 40 truckloads of sheep and goats
were exported from Shilabo
town every month, but this trade ceased altogether. Many
businesses in the town closed
down. A physical count of businesses operating in Shilabo
town before and after the
governments trade and currency restrictions found that more
than half the wholesalers (32
of 46), retail stores (57 of 98), tea-shops (49 of 82) and
restaurants (15 of 22) had closed by
May 2005. This is another example of unpredictable policy
changes having drastic negative
impacts on trading, and livelihoods that depend on trading
income, in Somali Region.
5 Farming and agropastoralism

Although lowland areas are often perceived as arid regions
with no potential for agriculture,
this perception is almost always inaccurate. In many
pastoralist areas, livestock rearers coexist with farmers, and are often cultivating crops
themselves. It is not surprising, therefore,
that Somali Region supports a diversity of livestock-based,
crop-based, and mixed (agropastoral) livelihood systems. The region has many seasonal
rivers that allow riverine farming,
there are irrigation schemes, and in some districts rainfall is
adequate to support rain-fed
farming. There are, of course, extensive parts of the region
that are too dry for farming,
notably in the eastern Hawd zones of Warder as well as parts
of Degahbur and Korahe.
This chapter presents information on agricultural activities
pursued by households in Somali
Region, drawing on data from our household survey as well
as qualitative research. The
chapter begins with a brief description of the different types
of crop farming practised in
the region, and then considers land ownership and access to
land. The next section
summarises crop harvests in our survey households, and the
chapter concludes by discussing
trends in crop production over time.
62
IDS
RESEA
RCH
REPOR
T 57
5.1 Farming systems in Somali Region

There are two distinct rainfall patterns in Somali Region,
which define two dominant
farming systems. Riverine farming is practised along the
banks of the perennial rivers, the
Shabelle and Dawa/Ganale in central and southern Somali
Region. In these low-lying areas,
the long rains (Gu) fall from April to June, and this is also
the planting season. The short
rains (Deyr) fall between October and December, but are
considered unreliable by farmers.
Since total rainfall is inadequate for rain-fed agriculture,
farming occurs along riverbanks and
in river valleys, and is dominated by Bantu Somalis. Lowinput methods are used, including
human labour, hand-hoes and traditional irrigation methods
such as furrows and channels,
though mechanical pumps are used to divert river water to
some fields. After the harvest,
during the hot dry Jilaal season (January to March), there is
little agricultural activity and
farmers consume their harvests. Once their granaries are
exhausted, they turn to the
market for food, using income earned from sales of cash
crops (e.g. onions, common in
Kelafo) or income-generating activities (e.g. making baskets
and mats, common in Dolo
Odo). Following the recent series of droughts, increasing
numbers of people rely on food
aid or assistance from relatives and friends.
In the northern Somali zones of Shinile and Jigjiga, a
different rainfall pattern prevails. A third
rainy season Karan, between August and September
provides sufficient moisture and a
growing season that is lengthy enough to allow the
cultivation of crops, even long-maturing
varieties. Rainfed agriculture is practised by Somalis on the
Jigjiga plans. Farming methods are
similar to those in the Ethiopian highlands smallholders use
family labour and cultivate with
oxen and ploughs. A third farming system found in Somali
Region could be described as
opportunistic agriculture, which exploits niches within
pastoralist areas, such as along seasonal
river beds or in valley bottoms that retain moisture after the
rains. Furrows and channels are

used to harvest rainwater. Much of this farming is practised
by pastoralists as a secondary
activity that yields modest harvests in most years and
significant harvests only occasionally.
Staple cereals are cultivated by almost every farmer and
agro-pastoralist in Somali Region. In
riverine farming communities, maize and sorghum are the
preferred staples. Agropastoralists throughout the region also favour maize and
sorghum. On the Jigjiga plains,
however, farmers plant wheat and barley as well as maize,
which they sell in Jigjiga town.
However, since the influx of food aid (mostly in the form of
American and European wheat)
after the famine of 2000, prices for local wheat have
collapsed in the face of competition
from sales of food aid, and many Jigjiga farmers have
stopped cultivating wheat altogether.
The plains farmers of northern Somali Region also farm
pulses (chickpeas, lentils) and root
crops (potatoes), and small amounts of vegetables, for both
consumption and sale. Riverine
farming in southern and central Somali Region is more
diversified, with commercial volumes
of vegetables (especially onions), fruit and oil seeds being
produced for domestic markets
and export to neighbouring countries.
Apart from large-scale irrigated agricultural schemes,
farming in Somali Region can be
described as low input, low output agriculture. Ox-ploughs
are used, but there is no
irrigation technology on family farms. Farmers dig furrows
(called mangat or moos) to
channel water from rivers or ponds to their plots. In Dolo
Odo, erratic rainfall is described
as the main problem that farmers face and the need for
irrigation pumps is widely
recognised, but this machinery is beyond the purchasing
power of local farmers, and no
credit or extension services are available. Very few farmers
apply chemical fertiliser only
one in a hundred (n=6/489 =1.2 per cent) in our survey, all in
Kelafo District. Moreover,
despite high levels of livestock ownership throughout the
region, very few farmers or agropastoralists apply animal manure just one in 20 in our
survey (n=23/489 =4.7 per cent),
mostly farmers on the Jigjiga plains.
5.2 Land ownership and access

Constitutionally, land in Ethiopia is jointly owned by the
state and the people. In practice,
this means that individual citizens cannot buy and sell land.
Instead, land rights are allocated.
Almost half the households we surveyed own rights to land,
mostly for farming
63
Table 5.1 Farming and access to land, by district

District Land-owning Farmed last If farmed last season:
households season Farmed Sharecropped
Rented
ownland in
land in land
Pastoralist 7% 2% 100% 0% 0%
Gashamo 0% 0% n/a n/a n/a
Shinile 20% 6% 100% 0% 0%
Shilabo 1% 0% n/a n/a n/a
Agro-pastoral 78% 75% 100% 0% 0%
Kebribayah 97% 97% 100% 0% 0%
Doboweyn 53% 45% 100% 0% 0%
Cherati 83% 83% 99% 1% 0%
Farmer 89% 86% 97% 3% 1%
Kelafo 95% 93% 98% 1% 4%
Dolo Odo 85% 86% 94% 6% 0%
Jigjiga rural 86% 78% 100% 1% 0%
Urban 2% 1% 100% 0% 0%
Jigjiga town 1% 0% n/a n/a n/a
Gode 2% 1% 100% 0% 0%
Total 523 (47.5%) 489 (44.5%) 483 (98.9%) 9 (1.1%) 4
(0.5%)
(n=523/1,091 =47.5 per cent). However, this average
conceals a reality of two extremes,
with ownership of land being close to universal in some
districts but close to zero in
others. In Kelafo (95 per cent) and Kebribayah (97 per cent)
almost all households own land,
while in Cherati, Dolo Odo and rural Jigjiga over 80 per cent
own land. In our Gashamo,
Shilabo and both urban samples, on the other hand, almost no
household owns any land.
Only in agro-pastoralist Doboweyn (53 per cent) is the
sample divided roughly equally
between households that own some land and those that do not
(Table 5.1).
Slightly fewer households farmed in the previous season than
own land (n=489/1,091 =44.5
per cent). Farming closely follows land ownership, with more
than 80 per cent of

households in Cherati and Dolo Odo, and over 90 per cent in
Kelafo and Kebribayah,
engaged in farming, but no sampled households farming in
the pastoralist districts of Shinile
and Shilabo, and only one farmer in either urban sample (in
Gode). In Doboweyn, where
just over half of the 100 households owns any land (n=53),
just under half are actively
farming (n=45).
Almost all households who farmed last season farmed their
own land in the last farming
season (n=483/489 =98.9 per cent). Nine households
sharecropped in some land, and four
rented in some land (Table 5.1). No farming households
reported having access to free land
from someone else. No households reported sharecropping
out, renting out or giving free
access to any of their own land. In general, land rental
markets appear to be less developed
in Somali Region than in highland areas, where severe land
scarcities exist.
Access to arable land is constrained in some farming
communities, but not all. Along
riverbanks, most arable land is allocated and under more or
less permanent cultivation.
Families divide their plots among their relatives and most
land is inherited. Most farmers are
satisfied with their landholdings, perhaps because the land
they can farm is constrained by
the availability of family labour, rather than the availability
of land. In our survey, 78 per cent
of farmers and agro-pastoralists (n=379/484) reported that
they have enough land for their
farming. The highest proportion of households claiming to
have inadequate farmland was in
64
IDS
RESEA
RCH
REPOR
T 57
Table 5.2 Most recent harvests, by district

District Farmers Harvest (kg) Consumed (kg) Sold (kg)
Given away (kg)
Kelafo 93 1,147 503 (44%) 505 (44%) 139 (12%)
Dolo Odo 86 443 275 (62%) 114 (26%) 54 (12%)
Jigjiga rural 78 447 318 (71%) 75 (17%) 54 (12%)
Kebribayah 97 237 196 (83%) 17 (7%) 24 (10%)
Doboweyn 45 533 216 (41%) 233 (44%) 83 (16%)
Cherati 83 332 262 (79%) 24 (7%) 46 (14%)
Average 440 527 311 (59%) 152 (29%) 64 (12%)
Source: Household survey data (n=484) (farming and agropastoralists households only).
agro-pastoralist Kebribayah (45 per cent), where average
farm sizes are 11 qodi, or 1.4
hectares larger than average farms in highland Ethiopia.
As in highland Ethiopia, though, landlessness in farming
communities is closely associated
with poverty. Because arable land is scarce in lowland Somali
Region, landlessness is
increasing. Among the riverine farmers of Kelafo District, the
landless are considered to be
poorer and more vulnerable than those with land.
There are many people these days without land. They
survive by iskutab
[sharecropping]. Others just work on farms to get their
daily bread. The situation is
caused by an increase in population without an
expansion in land size. The landless are
sometimes assisted by their relatives, who give them a
piece of land to farm for some
time.27
5.3 Harvests
Farmers in our survey were asked about their most recent
harvest. Table 5.2 summarises the
total production of major crops cultivated (cereals, pulses and
vegetables), and how this
harvest was used (consumed, sold or given away). Although
every farming household grows
staple cereals to meet household subsistence needs, it is

significant that most farmers and
many agro-pastoralists are selling some proportion of their
produce. By far the largest
harvests were recorded in Kelafo District. Most of these
riverine farmers are producing
onions for sale, and almost half their harvest (44 per cent) is
sold. In other districts, farming
for consumption dominates, and smaller proportions of the
harvest are marketed as low
as 7 per cent in agro-pastoralist Kebribayah and Cherati
Districts, from smaller than average
harvests. Also significant is the proportion of the harvest that
is given away rather than
consumed or sold at least 10 per cent in every district, even
where harvests are small.
Crop donations are made either for zakaat or other religious
and social obligations, or to
provide assistance to struggling relatives and neighbours.
In general, agro-pastoralists cultivate food crops to minimise
the number of livestock they
have to sell to buy cereals. One agro-pastoralist household
from Doboweyn that farms 10
qodi of land (11/4 hectares) harvested six 50-kg sacks of maize and
two sacks of sorghum in
2004, a total of 400kg. All of this grain was consumed by the

household none of it was
sold. Another household in Cherati farms maize for food, and
sells charcoal, firewood and
grass for cash income. Despite owning several head of cattle,
sheep and goats, they rarely
sell their livestock for cash. (We only sell our animals when
there is no food at all in the
house. When its a must I will sell one animal.) Many agropastoralists and farmers pursue
this mixed strategy of growing maize or sorghum for food,
selling firewood or charcoal for
cash, and selling livestock only if and when a consumption
gap remains, for instance in
27 Elder, Kelafo District.
65
Box 5.1 Case study: riverine farmer

Like most residents of Afdub Kebelle, Abdullahi
belongs to the Rer Bare clan, who
are descended from Bantu people who came to Kelafo
District many decades ago.
Abdullahi has two wives and 11 children, who live in
two households with their
mothers Abdullahi divides his time between the two
mother-centred homes. The
family farms sorghum and maize on three hectares of
land on the banks of the
Shabelle river. Abdullahi inherited this land from his
grandfather. Abdullahis oldest
son is 32 years old. He has married and is now living
in his own household, farming a
plot allocated to him from the familys land. All the
other children work on the family
farm and help with tilling, planting, weeding and
harvesting. Most of the harvest is
consumed by the family, with any surplus production
sold on the local market for
cash to buy basic necessities like sugar and cooking
oil. Abdullahi owns a small flock
of sheep and goats, some of which are sold to buy food
in years when sorghum and
maize production is insufficient. Abdullahi also owns
two donkeys, one for each of his
households to transport water for drinking and cooking
from the river, and to
transport produce to the market.
Source: Qualitative survey interview, Kelafo District.
drought years (Box 5.1). As another agro-pastoralist
explained: I dont get my income from
selling animals. When the drought came is when I sold my
sheep to buy food for my family.
Farmers and agro-pastoralists agree that those doing well are
those who farm vegetable
crops (onions, tomatoes) or khat, as well as staple cereals
such as maize and sorghum.
Vegetables and khat are high-value crops that generate
substantial cash income. They also
provide insurance against bad rainfall years, when vegetable
plots that are watered by hand
are less prone to crop failure. A good case study of this
insurance effect of cash crops is
provided in Box 5.2.
5.4 Trends in crop production

Household survey respondents were asked to compare their
most recent harvest with
harvests in past years one, two, three, four, five and ten
years ago. Results are
summarised in Figure 5.1. There are possible signs of a
nostalgia effect in the perception
that harvests ten years ago were considerably better than
harvests in more recent years,
though this finding might also have some basis in reality. The
droughts of 2000 and 2004
are highlighted in Figure 5.1 as years when the highest
proportion of farming households
suffered below average harvests, with 2000 confirmed as the
worst year for farming in
recent times.
Farming households were also asked why recent harvests
have generally been perceived as
worse than harvests of the past. Not surprisingly, lack of
rainfall which can be interpreted
Box 5.2 Case study: agro-pastoralist household
One agro-pastoralist household with 14 qodi planted 7
qodi to maize and 3 qodi to
sorghum in 2004, but harvested no maize or sorghum
at all. We did not get
anything this year, due to shortage of rainfall and the
problem of too much sunshine.
Before the drought this household grew maize
commercially, harvesting 20 50kg
sacks in 2003 and selling them for 500 Birr. Apart
from maize and sorghum, this
household also planted 11/2 qodi to khat, 11/2 qodi to
tomatoes, and 1 qodi to onions.
Although their khat crop also failed because of the

drought, they harvested ten sacks
of tomatoes, and were about to harvest their onion at
the time of our survey. This
was only possible because they bought water and
watered the tomatoes and onions
regularly through the growing season.
Source: Qualitative survey interview, Kebribayah
District.
66
IDS
RESEA
RCH
REPOR
T 57
Figure 5.1 Perceived trends in harvests, 19942004

100%
2%
8%
Crop Failure
24%
31%
39%
80%
49
%
35%
59%
Poor Harvest
60%
51% 42%
40%
Average
40%
Harvest
41
%
22%
55%
20%
Good
23%
15% 19% 25%
Harvest
9%
4% 2% 2% 1%
1%
0%
10 years ago 2000 2001 2002 2003 2004
as including erratic rainfall as well as droughts is
overwhelmingly the most frequent
response. Around half of all farmers also mentioned lack of
fertiliser and lack of tools or
technology, including irrigation machinery. These are
agricultural inputs that can be
purchased or provided to farmers by government agencies or
NGOs. The next three factors
are all natural crop pests, natural disasters, destruction of
crops by animals or birds. The
final explanation was mentioned by relatively few
respondents, but is extremely important,
as it relates to the chronic lack of agricultural extension
services in Somali Region: Lack of

agricultural experts to show the community how to use the
farm.
Table 5.3 Reasons why recent harvests are worse than in past
years
Rank Reasons given by farmers Number % of
farmers
1 Lack of rainfall
476 98.3
2 Lack of fertiliser 256 52.9
3 Lack of tools or technology 238 49.2
4 Crop pests
181 37.4
5 Natural disasters 88 18.2
6 Animals/birds destroyed crops 55 11.4
7 Poor soils
25 5.2
8 Lack of agricultural training 16 3.3
67
6 Internally displaced
persons
The United Nations High Commission for Refugees
(UNHCR) defines internally displaced
persons as:
Persons or groups of persons who have been forced or
obliged to flee or to leave
their homes or places of habitual residence in particular
as a result of or in order to
avoid the effects of armed conflict, situations of
generalised violence, violations of
human rights or natural or human-made disasters, and
who have not crossed an
internationally recognised State border.
The most common cause of displacement in Somali Region is
drought, which caused
widespread destitution of pastoralists in 1999 and 2004,
followed by conflict or insecurity
political struggles, clan conflicts and border disputes. In
2003, several thousand Somalis
were deported from Djibouti, and many came to Ethiopia.
According to UNOCHA, 69,095
(79 per cent) of the 87,995 IDPs in Somali Region in May
2004 were drought induced, and
18,900 (21 per cent) were conflict induced (UNOCHA
2004). The failure of the Gu rains in
2004 and the new IDPs related to a referendum in
communities bordering Oromiya, have
created two new waves of internally displaced people whose
status is ambiguous the
Ethiopian government does not recognise some of these
IDPs.
The vulnerability of IDPs in Somali Region was exacerbated
in early 2005, after the
government suspended food aid deliveries for several months
(for reasons explained below).
Together with ongoing problems of inadequate housing, lack
of sanitation and poor health
conditions, this resulted in critical malnutrition and mortality
rates in several IDP camps. This
chapter presents qualitative evidence on livelihoods and
living conditions among IDPs in
Somali Region, drawing mainly on life histories of several
IDPs in Hartisheik camp, and
interviews conducted with local and regional officials.2 8
Box 6.1 Abdi Koranes story

I am an Ogaden Somali from Degahbur. I have been
living here in Hartisheik Camp 1
for five years, since 1999. The main reason I came
here is due to drought. The
drought killed all my livestock and since I only reared
animals, it was a really major
crisis. We also got information that there were IDPs
staying in Hartisheik and getting
assistance. The only option I had was to move to
Hartisheik after getting this
information. I left my house and most of our
belongings behind. We only gathered
what we could carry and travelled here with our young
children, together with other
families from Degahbur. It was not easy.
We were 14 in number but now we are seven. Since
coming to Hartisheik I lost my
brother and six of my 11 children. I am left with five
children and my wife. My
children died due to malnutrition, malaria, cholera,
dysentery and tuberculosis, so did
other people from other families. There are no health
services in the camp. If there
were health services the death rate wont be so high.
The only treatment you can
get is from the pharmacy, that is if you have money.
Before coming we heard that Hartisheik has good job
opportunities, being a business
centre. But never believe what you hear. There is no
work for me or my wife. We send
our children to work, herding livestock, washing
clothes and polishing shoes. Thats the
only source of income we get. We get food aid from
the Regional Government, but
only two or three times in a year. When the food aid
comes, each family gets about
10kg of wheat to feed the whole family, which is not
enough for even one month.
Children are dying from hunger and diseases, so are
the old and the adults.
Source: Interview, Hartisheik camp (April 2005).
28 All the quotations in Chapter 6 come from interviews conducted with
IDPs in Hartisheik camp in April 2005.
68
IDS
RESEA
RCH
REPOR
T 57
Box 6.2 Feinuzs story

I came from Garbo because of the war there. We had
many animals before the war,
and were living okay. Then, when the war came, like
all the men, my husband was
involved in the fighting. We women werent asked
about whether they should fight
or anything about the war: in Somali culture, women
arent involved with such
matters. They simply come and tell us what they have
decided and we have to accept
it. We dont really agree with it, but we have no
choice. The war was between two
sub-clans of the same clan. The cause was just simple,
even stupid really. Two small
boys fought; one boy used an iron bar and injured the
other. Immediately the war
started and continued for two-and-a-half years.
With the war, many animals were killed, usually they
were killed on purpose by the
opposing clan. Most people on both sides lost animals.
It was just stupid really. We
were all making ourselves and each other poor: brother
was killing brother, brother
was making brother poor. Some animals were also sold
by the men to get money to
buy guns and weapons. Women would never think of
selling animals, our wealth, to
buy guns to fight our brothers. But men think
differently when it comes to fighting.
My husband was killed about six months before I left
Garbo. By then, everything was
difficult. There was nobody to help me. Before, the
community would help you if you
lost your husband. But, then, everybody was in a bad
situation and there were many
widows. So, when some people were leaving, I
decided to leave also because I had
no animals and no husband there wasnt anything
there for me.
6.1 Livelihoods in Hartisheik camp

Hartisheik is located 75km south-east of Jigjiga town, in
Kebribayah District of Jigjiga Zone.
During the civil war in Somalia, it was the site of the worlds
largest refugee camp. The
refugee camp was closed in late 2004 and UNHCR withdrew
from the region. However,
two IDP camps remained. As of mid-2005, approximately
5,000 IDPs were living in Camp
A and Camp B. Most IDPs in Hartisheik are Somalis, but
around 10 per cent are Oromos.
Box 6.1 and Box 6.2 present case studies of two residents of
Hartisheik, one displaced there
by drought and one by conflict.
Hartisheik residents survive mainly by providing various
services to the local population.
Women work as domestic servants, cleaners, tea-shop
waitresses and clothes washers in
Hartisheik town. Men sell khat, collect firewood and look for
daily labour. Child labour is
common. Young girls help their mothers with washing
clothes; boys work as shepherds,
porters or shoe-shine boys. Since the governments recent
clampdown on contraband
trading, however, the level of economic activity around
Hartisheik has declined, and IDPs
find it harder now to earn a living.
People from Hartisheik travel to towns like Tog Wachale (a
Somaliland border town) or
Bosaso (a port in Puntland) in search of work. Many men
have abandoned their families.
Several mothers admitted resorting to more drastic survival
strategies, including raising
money by selling their childrens labour.
Because of hunger, some mothers have even been
selling their children. Even me, Ive
got three children: two are still small and the third is a
5-year-old girl. I took her to
town and sold her for 100 Birr to another woman who
took her to look after her
goats. She was supposed to give me more money every
month, but never did. I saw
her once afterwards, and then not again. I dont think
Ill see my child again. But at
least she has food where she is. Even if I fetched her
now, what would I feed her? I
cant just watch my child die. Its bad enough thinking
how Im going to feed the other
two children. If they were older and able to work,
maybe I would have sold them too.
69
Box 6.3 Fatimas story

First, Hartisheik was a busy trading place. So, when
we first got here, many people
were able to do some trading and other things to earn
money. We werent sitting
and waiting for NGOs to help us like now. We were
independent and trying to help
ourselves. But the government started to ban what they
called the contraband
business in Hartisheik, so many people left and the
town started to get quiet.
Of course, we are still trying to survive, especially
among the women in the camp.
Myself, I have three sons and sent them to work for
different people to earn money.
I used to do washing for people to earn money. My
husband did nothing - what can
a pastoralist do in a town? Also, men arent like
women: they are too proud to do
something small, even if they are starving. A woman
will do anything to earn an
income. A womans pride only goes as far as her
hungry children.
I had an agreement with the people I sold my daughter
to. They were supposed to pay
me 100 Birr every month. Then they changed the
agreement and gave me 50 Birr
after three months. Finally they said: We have nothing,
so take your child. But
because I had nothing, I left her with them. Even if we
go back to live as pastoralists
again, I think Ill leave her there.
I had five children. I sold my 5-year-old daughter to be
someones housemaid and my
10-year old son to the same family, to be a goat-herd.
They were supposed to pay me
100 Birr every month. They gave me the first 200 Birr
for both, then didnt pay me
again. They live far from here. But even if they were
near here, I wouldnt fetch them
because I have no food or anything to give them. I
would be fetching them to starve
with all of us here in the camp.
6.2 Conditions in Hartisheik camp

Life as an IDP in Somali Region has always been difficult,
but conditions in Hartisheik are
particularly harsh. A study of eight IDP sites across Ethiopia
in 2002 found that: Conditions
at Hartisheik site appear to be the worst. IDPs at Hartisheik
are completely destitute and
wait for food aid. Starvation is a major concern (Miz-Hazab
Research Centre 2002). A
survey in 2001 found that food supplies to IDP camps in
Somali Region were irregular and
inadequate, sources of water were unsafe, and sanitary
facilities were non-existent, as was
access to education. None of these factors has improved since
2001. Health services are
limited to a single clinic run by a local NGO, and people
have to pay for treatment and
drugs, which most cannot afford. The combination of chronic
hunger, overcrowding, poor
housing, unsafe water and lack of sanitation has created a
lethal public health environment
where the risks of disease and premature death are
persistently and unacceptably high. My
wife and son died. Death is common here; One of my
children died aged five, due to lack
of food; Three of my children died, and my first husband.
My husband died of tuberculosis
and my children of malnutrition and diarrhoea; Every day
we bury three or four babies and
old people. It is a sin that people are living and dying like
this.
Two surveys in Hartisheik IDP camp since 2004 found rates
of global acute malnutrition
(GAM)2 9 that indicate a serious situation (>20 per cent), severe
acute malnutrition (SAM)
that indicate an emergency (>5 per cent), and under-five

mortality rates that indicate an
emergency out of control (>4/10,000/day) according to
international guidelines (World
Food Programme 2000). In November 2003, 502 children in
Hartisheik under five years old
29 Global acute malnutrition (GAM) is defined as weight-for-height
below 2 Z-scores and/or oedema, or less
than 80 per cent of the median weight-for-height and/or oedema.
Children with severe acute malnutrition
(SAM) fall below 3 Z-scores or 70 per cent of the median weightfor-height.
70
IDS
RESEA
RCH
REPOR
T 57
were measured. A GAM of 28.5 per cent and a SAM rate of

4.4 per cent were recorded. In
March 2005, Save the Children UK conducted a nutrition
survey in Hartisheik, at the
request of the regional DPPB (Save the Children UK 2005).
All camp residents aged 659
months (855 children) were measured, and retrospective
mortality data was collected from
all households for the previous three months. Preliminary
results found a GAM of 24.2 per
cent and a SAM of 5.1 per cent. The under-five mortality rate
(U5MR) was 4.7/10,000/day.
According to Ethiopias emergency nutrition assessment
guidelines, these figures classify the
situation in Hartisheik camp as critical.
6.3 Food aid in Hartisheik camp

In October 2004, IDPs in Hartisheik reported that food aid
was delivered by Regional
Government (DPPB) or the World Food Programme (WFP)
only two or three times a year,
and in small quantities 10kg or 15kg of wheat per
household, irrespective of household
size, sometimes supplemented with corn soya blend (CSB)
and beans. OWDA (the Ogaden
Welfare and Development Association) also provided small
amounts of food aid during the
month of Ramadan 6kg of porridge, 11kg of sorghum, 1/4 cup of
oil, and 1/4 cup of dates.
We do receive food aid but very rarely, and its inadequate.

10kg is not enough for even
one week, let alone three or four months. We receive only
dry wheat, 10kg one time
every three months, which is not enough. They distribute the
same amount of food to all
families, not according to the family size. This is not fair. By
April 2005, IDPs reported a
much more serious situation.
We havent received food for five months. We have no
idea why it has stopped. All
we know is people are dying of hunger in the camp and
nobody is coming to tell us
when we will receive anything.
Yesterday, we had three burials. There are more deaths

these days now that food
hasnt been delivered. Before we had one burial a week.
There were never as many as
this even a few months ago.
Yesterday my baby had a cup of sugar water and a small
piece of bread. Hes so thin
and sick, he cant take anything more. Im scared
because so many babies have died in
the last few months.
6.4 Government policy towards the IDP camps

In March 2005 the government issued a directive that food
aid trucks could not move to
Degahbur, Fiq, Korahe, Gode and Warder zones without an
armed escort. This was a
response to attacks by rebel groups on over 40 trucks
carrying food aid during the past year,
when some food was taken as a tax or all the food was
looted by the rebels. A further
argument for the need for military escorts were reports of
food aid diversion (selling and
misuse) either during transportation or at the warehouses.
Whatever the merits of these
arguments, the result has been that food had not been
delivered throughout the region,
including the IDP camp in Hartisheik. A senior DPPB official
in Jigjiga reported in early April:
There is a consignment of food sitting in Dire Dawa,
although we know the
beneficiaries are waiting. The transporters and trucks
are ready and have been waiting
for 7 days to get the necessary military authorisation
and escort. Another convoy
which went without a military escort has been stopped
in Harar. DPPB has resolved
to write a letter to the Regional government stating that
food is now critical and that
any deaths that might result will not be our fault, but
that of the Regional
government. The policy should be that where the
military are ready to escort, well
welcome them, but if they cause delays, we need to go
without them.
Key actors in food aid distribution provided other possible
contributory explanations for the
current delay in food delivery. First, the governments new
Safety Net Programme in the
region will affect the way in which beneficiaries are
categorised and food is distributed.
71
Second, the IDPs repatriation and/or their length of stay in

Hartisheik might also be
affecting decisions regarding food aid. For instance, a DPPB
official reported that it was
now government policy that if IDPs were in a camp for four
years or more, they would no
longer be eligible for food aid as IDPs, but should be
integrated and included with the rest
of the resident population. Third, poor communication
between the DPPB and the Disaster
Prevention and Preparedness Commission (DPPC) is
exacerbated by the asymmetrical
power relationship between the DPPB, DPPC, the regional
government and the federal
government. Fourth, there is always a tension between
humanitarian and political priorities
and considerations in the region.
Whatever, the combination and validity of factors and actors
that explain the current
situation, in early April 2005, IDPs were clearly confused and
angry about the non-delivery
of food aid to the camp. But they did display an astute
understanding of the actors involved.
We send people all the time to the Administration
offices, but get no help from
them. I think they are the administration for the people
of Hartisheik, and not for us
in the camp.
The government is telling the NGOs we dont want to
go back. Two things make
them want to keep us here. First, they steal half of the
food that we get from the
NGOs. As long as were here, they will have this food.
The second thing is that our
children have become their servants, doing all their
dirty work for them. So, thats
another reason that they dont want us to leave here.
The government distributes food received from WFP.
We dont know why we havent
got food now. Weve been going to the government
offices in Hartisheik asking, but
they give us no answers. So, I cant tell you where the
problem is: it can be with WFP,
or DPPB or the government sitting in Jigjiga or even
Addis.
Every day we go to the Administration office to ask
about our food, but they dont
even want to speak to us anymore. They just send us

home and tell us to go and wait
for the result. We are asking you to tell the government
that: we need to be
recognised as real refugees; or they must give us
transport to go back.
Their representation of the local Administration officers was
borne out by the remarks of
one of the senior officers, who said in early April:
The IDPs are highly problematic to work with: they are
always fighting (e.g. when we
are trying to deliver food and water). They just want to
continue receiving food here in
the camp. The only way to get them to go is to stop
feeding them in the camp, move
them to their homes and give them something once they
are there. Otherwise, theyll
simply stay here forever.
However, he did confirm that food had not been delivered for
the previous four months.
He said that they (i.e. the local administration) had written to
the DPPB about it, and that
the Mayor of Hartisheik was also looking into it.
6.5 The way forward?

There are a number of contested perspectives regarding the
future of IDPs when they leave
the camp. Emphasising the difficulties associated with a
future in pastoralism, women
articulated a desire for alternative livelihoods, particularly for
their children.
We have all seen on television how people live: in
houses with water in taps and enough
food to buy in the shops. We want to live like that
instead of the hard life we lead. If we
had a choice, of course we want that life for our
children also. We are tired of the hard
life of pastoralism. Now our children also know about
these alternatives, like living in
72
IDS
RESEA
RCH
REPOR
T 57
cities and being educated. When they see that, of course

they prefer that easier life. They
arent interested in being pastoralists any more.
Honestly, if they had a choice and we
had a choice, we would educate them all and forget
about animals. I think many men
even agree with this. They and their fathers before them
have been pastoralists but they
know that things are worse than ever before and will not
get better.
I dont want my children to be pastoralists. It isnt a life
anymore. The situation has
changed too much. You can be rich one year with many
animals, and then you suffer
two or three harsh droughts, and everything is gone.
Because the droughts are so
severe these days, it isnt possible ever to recover your
losses.
I think that men like having livestock and think that is
what a Somali man is supposed
to do. They believe that is what makes a Somali man a
man: without animals, they
dont feel like Somali men. But the young boys are
changing now. They dont want to
be pastoralists. Even before the last drought, we were
starting to hear boys and young
men saying they want to go to the towns and earn
money and dont want anything to
do with animals.
Somali like their cattle, but we Somali women know
that its not possible anymore to
make a living like that. Men know this, but many are
too proud to admit it. But if you
enter the camp now, you will find many men who are
even admitting that if they
found an alternative, they would leave pastoralism.
Men generally expressed a preference for returning to
pastoralism after they leave the camp.
On the other hand, they also often acknowledged the
difficulties.
We would like to continue having animals, but it is very
difficult. There are too many
animals, too many people, too many bad droughts, too
little grazing. We are Somalis,

we are pastoralists. But the weather is changing and its
not possible to carry on
anymore.
We are Somalis and only know pastoralism. But we
dont think that we can survive on
it anymore. I dont think our sons are going to be able to
do it. So, its better if we
send our sons and daughters to school.
My future plan is to go back to my original homeland,
despite the problems we faced
there. Living in this camp has made life impossible. No
jobs, no food, no health
services, no education. Going back home is 100 per cent
better than staying here. If
we could get support from NGOs like first helping us
either by restocking us with a
few livestock or helping us with farming equipment. In
that way we would work and
support our families. We would also send our children
to school if they could provide
free education.
However, in the short term, women and men share the desire
to be repatriated to their
areas of origin:
Even if we get transport tonight, well go immediately
even without food or
animals. We dont have a life here.
Even if we get transport and no restocking, we would
go today. We are tired of
sitting here and starving and dying of illness. We are
totally dependent on their food.
When it doesnt come, we die. Its as simple as that.
We all want to go back as soon as possible. I dont think
so many of us would have
been dying if we were in our own areas. There would
have been others to help. Here,
everyone hasnt received food for four months, so there
isnt anyone to turn to for
help. We just have to stand by and watch our children
die.
73
7 Incomes
inequality
and
This chapter presents several types of data, mostly from our

household questionnaire survey,
that provide estimates or proxies of income and inequality in
Somali Region. Part 7.1
summarises and compares average cash earnings from over
60 income-earning activities
pursued by 1,100 households surveyed, while part 7.2
comments on income inequality
between households. Part 7.3 identifies the most lucrative
and least lucrative livelihood
activities, ranks the most common activities and finds that the
most popular livelihood
activities are clustered among the low-return activities. It also
discusses some secondary
livelihoods not discussed elsewhere in this report gathering
and selling natural products
(firewood, charcoal), craftwork (mat-makers) and artisans
(shoe-repairers, tailors). Part 7.4
introduces a different indicator of well-being or poverty
food consumption and
calculates dietary diversity scores. Another proxy indicator
purchases of clothes and basic
grocery items is explored in Part 7.5. The chapter concludes
with a discussion of the main
problems that respondents reported in making a living, and
finds strong evidence of
occupational immobility among the households surveyed.
7.1 Household incomes

The household survey questionnaire asked respondents to
report on all livelihood activities
in which they had engaged during the past year,30 and to estimate
the average income
earned doing each activity.31 Table 7.1 summarises monthly cash
incomes earned by sampled
households, by district and livelihood category, both for

households reporting positive
income and for those who recorded no cash income at all.
Approximately one in five
households interviewed (n=227/1,098 =20.7 per cent)
reported undertaking no incomegenerating activities in the 12 months preceding the survey,
and a further 119 households
(10.8 per cent) could not (or would not) estimate the income
they earned from the activities
they do undertake.3 2 Most households recording zero income are
agro-pastoralists
(n=95/227 =42 per cent) or farmers (n=79/227 =35 per cent).

Only four urban households
(2 per cent) reported zero income, reflecting the higher
degree of monetisation of urban
livelihoods.
Many incomes in rural communities were reported in Somali
Shillings or Somaliland Shillings
(including most incomes in pastoralist Gashamo), reflecting
the fact that rural livelihoods in
Somali Region are highly integrated with the economy of
neighbouring Somalia and
Somaliland.3 3 Conversely, almost all urban respondents reported
their incomes in Birr,
reflecting the fact that urban residents tend to be more closely

connected with the
mainstream Ethiopian economy.
Table 7.1 reveals that average household incomes in urban
centres of Somali Region are, not
unexpectedly, substantially higher than in rural areas. On the
other hand, costs of living are
30 A list of over 60 different livelihood activities was read out to each
respondent, who stated whether (or not)
at least one member of his or her household had been engaged in that
activity during the past year. The list
of activities was compiled in training and during pilot testing of the
questionnaire.
31 Incomes are notoriously difficult to estimate through direct questions
in a household survey, and for this
reason household income is usually proxied by consumption or
expenditure data. Our purpose in this survey
was to understand household livelihood portfolios and to assess
relative returns to different livelihood
activities. The income statistics reported here should therefore be
regarded as indicative and relative. They
almost certainly underestimate total income, as they exclude the
value of remittances and other transfers.
They should not be taken as measurements of absolute income from
which poverty headcounts could be
derived.
32 Households reporting no livelihood activities could be surviving on a
range of alternative sources of food and
income that are not recorded in this table, including food crop
farming and livestock rearing (consuming own
crops and animal products), barter exchange, illicit livelihood
activities (e.g. contraband trade), remittances,
food aid and other transfers.
33 These incomes were converted into Ethiopian Birr at the exchange
rates prevailing at the time of the sur vey:
1 Birr = 1,500 Somali Shillings; 1 Birr = 685 Somaliland Shillings.
74
IDS
RESEA
RCH
REPOR
T 57
Table 7.1 Average household cash income, by district (Birr

per month)
Excluding
zero
income
households
Including zero income households*
District Mean (Standard Gini Mean (Standard Gini
income deviation) coefficient income
deviation) coefficient
Pastoralist 340 (471) 0.56 217 (410) 0.72
Gashamo 516 (811) 0.65 196 (556) 0.87
Shinile 159 (100) 0.34 132 (109) 0.45
Shilabo 460 (419) 0.47 319 (408) 0.63
Agro-pastoral 199 (217) 0.52 97 (181) 0.77
Kebribayah 230 (228) 0.48 97 (186) 0.78
Doboweyn 238 (263) 0.53 121 (222) 0.76
Cherati 137 (133) 0.51 73 (118) 0.74
Farmer 345 (460) 0.56 254 (423) 0.67
Jigjiga rural 337 (332) 0.44 253 (322) 0.58
Kelafo 431 (640) 0.63 303 (571) 0.74
Dolo Odo 272 (346) 0.55 206 (323) 0.66
Urban 1,103 (2,263) 0.52 1,081 (2,245) 0.53
Jigjiga town 1,225 (3,120) 0.63 1,225 (3,121) 0.63
Gode 979 (656) 0.35 940 (671) 0.38
Total 511 (1,252) 0.61 351 (1,064) 0.74
By household head
Male head 517 (1,288) 0.62 354 (1,093) 0.74
Female head 424 (449) 0.52 297 (423) 0.66
* 346 households reported zero cash income from any source
(n=752).
higher in urban areas for rental, electricity, transport, food
that must be purchased (rather
than produced or farmed or consumed from livestock) and
other expenses that are lower or
negligible for rural residents so in real terms the difference
is less than the figures suggest.
Nonetheless, living standards are generally higher in towns,
and livelihood vulnerability is
lower.
Cash incomes in Jigjiga town, for example, are fully 17 times
higher than in Cherati and 12
times higher than in Kebribayah. Both of these are agropastoral districts, and the three
agro-pastoralist communities surveyed occupy the lowest
positions in a ranking of cash
incomes by district (for all households, including those with
zero reported income).
Interestingly, farmers are, on average, the wealthiest rural households in our survey,3 4
although the wealthiest rural district is Shilabo, a pastoralist

area. Excluding zero-income
households, Gashamo has the highest average income, which
suggests that but for the
effects of the 2004 drought, pastoralist-based livelihoods may
be the most lucrative way of
making a living in rural Somali Region but also the most
vulnerable.
34 It must be reiterated that only cash income is considered in this

analysis. This includes income earned from
crop sales, but it does not include the imputed value of domestically
consumed agricultural production. While
this information is available for crop producers from the household
survey, comparable data on milk and meat
consumption an equally important source of food consumption
among livestock producers is not.
75
Figure 7.1 Lorenz curves of household incomes, Somali Region3 6
(Excluding
(Including income=0)
income=0)
7.2 Income inequality

Inequality between male- and female-headed households is in
the predictable direction.
Table 7.1 reveals that households headed by men are 20 per
cent wealthier than households
headed by women. If anything, the extent of inequality is
rather less than expected. One
factor explaining this is that female-headed households are
disproportionately resident in
towns rather than in rural areas, where many of these women
are wealthy independent
traders. Within rural communities, the income differential
between male- and femaleheaded households is considerably wider. On the other hand,
the fact that female-headed
households have average incomes that are higher than
average incomes in most rural
districts suggests that there is no basis for classifying
households headed by women as a
vulnerable group requiring targeted safety net
interventions.
Table 7.1 also presents Gini coefficients for household cash
incomes, by district and livelihood

system.35 Even excluding zero-income households, inequality in
terms of cash income earned
is rather high (Gini=0.61). Across the full sample, including

those with zero income,
inequality is extremely high (Gini=0.74). Figure 7.1 plots the
distribution of household
incomes in the form of Lorenz curves, which illustrates how
unequally distributed cash
incomes are, especially across the full sample. Interestingly,
income inequality is substantially
lower in urban areas (Gini=0.53), where incomes are higher,
than in rural areas, where
average incomes are lower but the range of incomes is wider.
It is intriguing that the
highest Gini coefficient (0.87) of all is registered for
Gashamo District. Although a sample
size of 100 is too small to draw meaningful conclusions for a
large and complex district, this
could reflect the polarising effects of the recent drought.
7.3 Returns to livelihood activities

Many of the most lucrative sources of income are
preferentially or exclusively accessible to urban
residents, such as most categories of salaried employment,
construction work or carpentry, and
renting out property. Conversely, many livelihood activities
that are exclusively rural, in that they
derive from natural products not found in towns, generate
very low returns. These include
selling firewood, wild fruit and charcoal, making baskets and
mats, and beekeeping (Table 7.2).
Table 7.3 ranks the 20 most common livelihood activities
pursued by the 1,100 households in
our survey. It is striking how many of these activities are
derived directly from the natural
resource base: 11 of the 20 are agricultural activities (farming
or livestock-based), and a further
three involve gathering and selling natural products
(firewood, charcoal, building materials).
35 The Gini coefficient is a measure of income or wealth inequality
between households in a given population,
where complete equality is signified by a Gini value of 0, and
complete inequality (all income or assets accrue
to a single household) by a Gini value of 1.
36 The x-axis in these diagrams shows the cumulative proportion of the
sample population; the y-axis shows the
cumulative proportion of total income earned. The flatter the cur ve,
the more equal is the distribution of
income across households.
76
IDS
RESEA
RCH
REPOR
T 57
Table 7.2 Average income by livelihood category, and by

highest and lowest returns
Income category Birr/month Most lucrative activities
Birr/month
Trading 615 1. Contraband trader 1,607
Rents
502 2. Construction worker 1,307
Employment or labour 447 3. Carpenter or metal-worker 873
Services
300
4.
Khat
trader
868
Food & drink processing 244 5. Selling meat (from own
livestock) 853
Livestock production 216 Least lucrative activities
Crop farming 210 60. Charcoal seller 100
Crafts & small industry 182 61. Firewood collector 88
Begging 123 62. Basket-maker or mat-maker 88
Sale of natural products 117 63. Selling eggs (from own
chickens) 79
Average income by activity 397 64. Beekeeper (selling
honey or wax) 77
Of the remaining six, craftwork (making baskets or mats) and
livestock trading also rely
directly on natural resources. Only services (religious
teachers, traditional healers, tea-stalls)
and employment (salaried job, daily labourer) are largely
independent of the physical
environment, though of course the demand for services
depends on the incomes of local
households, and some casual labour is agricultural or
livestock herding. Table 7.3 therefore
reveals a principal source of livelihood vulnerability in rural
Somali Region, namely, that a
dangerously high proportion of household income derives
from an environmental base that is
limited and fragile, and varies according to unpredictable fluctuations in the weather.3 7
Table 7.3 Most prevalent livelihood activities in Somali

Region
Livelihood activity
Total Average
Total
hou
seh
olds
(%)
inco
me
(Birr)
1 Livestock: rearing animals (camels, cattle, shoats) 763 69.9
258
2 Farming: cereal crops (sorghum, maize, wheat) 474 43.4
142
3
Natural
products:
firewood
186 17.0 88
4
Natural
products:
charcoal
160 14.7 100
5 Livestock: dairy products (milk, butter, ghee, cheese) 89
8.2 125
6 Craftwork: basket-making, mat-making 69 6.3 88
7 Farming: pulses (beans, cow-pea, chick-pea) 63 5.8 103
8 Services: selling tea, coffee, cake, bread 63 5.8 199
9 Farming: oil crops (sesame, sunflower) 50 4.6 176
10 Trading: livestock (cattle, sheep, goats, camels) 41 3.8 562
11 Farming: vegetables (onion, tomato, pumpkin) 41 3.8 397
12 Natural products: construction materials (grass, poles) 36
3.3 166
13
Employment:
salaried
job
35 3.2 584
14 Services: religious teacher (Koranic) 34 3.1 231
15
Services:
traditional
healer
26 2.4 360
16
Employment:
daily
labourer
26 2.4 233
17 Farming: root crops (Irish potato, sweet potato) 26 2.4 141
18 Farming: fruits (mango, papaya, banana, orange) 26 2.4
282
19 Farming: khat
25 2.3
586
20 Livestock: rearing chickens (selling eggs) 25 2.3 79
37 As Sen (1981) demonstrated, when agricultural incomes collapse so
does the demand for informal services in
affected communities, so rural service providers are equally
vulnerable to natural disasters, through a process
that Sen characterised as derived destitution.
77
Box 7.1 Case study: charcoal sellers

Retailer: I have been involved in the charcoal
business for the past three years,
because most of our animals perished because of the
droughts. I buy charcoal directly
from the men in the bush who prepare the charcoal.
Then I sell it in small bags to
passing traffic on the road. We buy it for 7 Birr and
sell it for 10 Birr. If the market
is good and active, I can sell 30 bags in a month, but
when the market is bad, I only
sell 10 to 15 bags. Different things affect the traffic: if
people have money, they travel
more, so there is more passing traffic. In the drought,
people dont have money to
travel, so I sell less charcoal. More women are selling
charcoal now, since they face
similar circumstances.
Wholesaler: Khadija started in the charcoal business
in 1997. She buys charcoal in bulk
from the trucks that transport charcoal from the
Degahbur area to Jigjiga market.
The purchase price from the producer is 10 Birr per
sack, and she resells each sack at
25 Birr. Transport costs are 6 Birr per sack and she
pays tax in Kebribayah (100 Birr
per truck) and Jigjiga (150 Birr per truck). When the
trade is good, we bring in
seven big trucks every month. Sometimes the
charcoal is treated as contraband
vehicles are impounded by the authorities and the
charcoal is confiscated. During
these periods, we use camels to transport our charcoal
into the city. Transport by
camel is more expensive, at 20 Birr per sack, and it is
slower. Profits for traders
declines, as volumes are affected.
Source: Life history interviews with traders.
It is also significant that none of the five most lucrative
livelihood activities listed in Table
7.2 appear in this list of the 20 most popular livelihood
activities in our survey. Conversely,
four of the five least lucrative activities (all except
beekeeping) do appear on this list.
Clearly, there are entry barriers to the most attractive
occupations in Somali Region such

as working capital constraints, skills requirements or
educational qualifications that exclude
the majority of local people from pursuing these activities.
Instead, they are mostly confined
to pastoralism, farming, gathering natural products and
craftwork. Since pastoralism and
crop farming have been discussed elsewhere in this report,
below we discuss some of the
other most common livelihood activities.
7.3.1 Gathering and selling natural products

Gathering and selling natural products firewood, charcoal,
poles and grass for housing,
fodder, wild fruits and incense where these are available is
an important source of
secondary income for many residents of rural Somali Region.
Selling charcoal and firewood
are, in fact, the most common livelihood activities recorded
in rural communities, after
livestock rearing and crop farming (Table 7.3). However,
these ways of generating income
should not be seen as chosen or preferred, but instead as last
resort options adopted by
people who are poor and desperate for any income at all. The
work is arduous and timeconsuming, and the returns are tiny. Very often, these
activities are adopted after normal
sources of income have failed; many people told us they
resorted to selling firewood only
after drought had brought them close to destitution.
I started selling charcoal because of the drought. It is an
activity that poor people do,
because you have to go some distance to find and buy
charcoal. The income is very
small.
It wasnt like this before. There werent so many
droughts. Before the last two
droughts, everyone owned animals and was basically
self-sufficient. Since then,
women first started selling firewood and charcoal for the first time.3 8
38 Two women in rural Gashamo.
78
IDS
RESEA
RCH
REPOR
T 57
Box 7.2 Case study: building materials trader

Amina, a 26-year-old woman from an agro-pastoralist
family in Borale village, started
selling khat after she was divorced. During the rains,
however, khat grows rapidly, there
is over-supply and the market collapses, so Amina
changed her business to selling
building materials instead. She buys from local
suppliers in Borale, where there is an
abundance of good quality wood, and sells in Jigjiga
market. At first she suffered heavy
losses, but now she has built relationships with certain
suppliers, and has established
herself in the market despite being a woman with no
relatives to help her get started.
But it is difficult, because competition between the
traders in Jigjiga is fierce.
The building materials trade is wholly within the
Somali Region, so is not affected by
border closures or the erratic access to the Arab
markets. The main problem the
traders face is taxation. In Gorey, 20 Birr tax is
charged per trip. In Dagahle, 150 Birr
is charged per trip. At Karamarda, 70 Birr is charged
per trip. Finally at the Jigjiga
market centre, 10 Birr is charged per month. This
multiple taxation is a burden!
Amina is still one of the few women involved in
trading building materials in Somali
Region. As a woman, it is better than farming or
rearing livestock, because you make
your own profits or losses depending on your hard
work and skills. In business, it
doesnt matter very much if you male or female;
everybody is in it for their profits,
and everybody is competing against each other. There
are none of those stiff
traditions that exclude women, unlike in pastoralism
and politics.
Source: Case study interviews with traders.
Charcoal burning is often criticised as causing environmental
damage, and some attempts
have been made to regulate it, but these regulations are

weakly enforced. Local people
themselves recognise this problem, but argue that those who
burn charcoal are forced into
this by poverty. Recently, commercial charcoal associations
have been formed that are
exporting large quantities of charcoal to Somaliland. This
commercialisation of a
traditional informal livelihood activity is increasing the
incentives to over-exploit the
natural resource base, and according to some people is even
undermining Somali culture.
One key informant interviewed in Harshin elaborated on the
negative interactions between
drought, charcoal burning and khat consumption:
Before the 2004 drought, charcoaling was happening on
a very limited scale. But as
peoples livelihood of pastoralism is being threatened
by the droughts, they are being
forced to rely more and more on charcoal burning and
selling. The land is being turned
into a desert right before our very eyes, because the
community is burning too many
trees for charcoal. In Harshin, there is a large-scale
export charcoal business charcoal
is exported to Somaliland and elsewhere. If they were
only producing charcoal for
themselves and our community here, it wouldnt be so
serious. Pastoralists in the bush
are working for the big charcoal associations. They are
paid to fence the land in the
bush, prepare the charcoal and sell it to the association
in town. They have a
connection to Somaliland.
Before the drought, it was only very, very poor people
who were involved in charcoaling.
Now, its a big business involving people that werent
that poor: we even have some
people exporting to the Gulf. There is a fine of 100 Birr
for cutting trees. But people
dont listen. Im surprised they dont listen to us
according to Somali culture and
tradition, people are supposed to listen to their elders
and leaders. It seems the culture is
changing. People are more concerned with making
money than anything else. Before,
men were not chewing as much khat as now. Now, even
young boys are chewing khat.
It shows that something is seriously wrong in our
community. The Association is also
enticing young boys to burn charcoal and gives them money or khat as payment.39
39 Key informant, Harshin.
79
Box 7.3 Case study: mat-maker

Dherey is a 39-year-old woman who was born and still
lives in Ukta, in Dolo Odo.
Dherey is married to Suleiman, and they have ten
children. Ukta is a close-knit
community most people are Bantu, from the same
sub-clan. Dhereys family farms
maize along the river. They generally harvest only
threefour bags of maize, which is
not enough to feed all 12 of them. To supplement their
crop production, Dherey
collects grass and makes mats that she sells in Dolo
town, about 16 km away, while
Suleiman collects firewood and construction materials
(grass and wooden poles for
house-building) and sells these in Dolo town. The
couple earns about 30 Birr each
week from these activities, most of which they spend
on buying food for the family,
as well as sugar, tea and other essentials.
All mat-making in the village is done by women,
though children and some men help
with collecting the materials. One problem associated
with mat-making is market
fluctuations sometimes Dherey returns from Dolo
having sold no mats at all.
Another problem is the distance involved in travelling
to the market, which is
daunting and expensive. She rents a donkey to
transport her mats to Dolo market,
paying the donkey owner 6 Birr per trip. If I could
afford a donkey, the cost would
be less and I would make more profit. Nonetheless,
Dherey is also teaching her
daughters how to make mats, so that they can earn an
independent income for
themselves when they grow up.
Source: Qualitative survey interview, Dolo Odo
District.
Firewood for cooking is a significant expense in most rural
Somali households. One bundle
of firewood (called a hidmo) generally costs 2 Birr and lasts for 1
11/2 days. Firewood is
collected from the bush by women and donkey owners, who

tend to favour deadwood
dry bush, broken branches and fallen trees as long as it is

available, but also cut green
wood when there is no alternative. About eightten donkeyloads of firewood are brought
into Jigjiga market every day. Each donkey can carry ten
bundles of wood. Many women
also walk to Jigjiga market carrying firewood for sale;
usually four or five bundles each. If
they sell everything they earn just 8 or 10 Birr for many
hours of physical labour gathering
wood, walking with a heavy load (often several kilometres)
to market, sitting at the market
all day and walking back home.
Wood is also used as a building material: to reinforce the
walls of pastoralists moveable
homes, as rafters to support roofing in permanent structures,
and to be made into doors
and window-frames. Hardwood suitable as construction
materials is collected from certain
districts where the appropriate trees either grow abundantly
in the wild or are cultivated
specifically for this purpose, mainly by agro-pastoralists.
Jigjiga market is supplied with
hardwood from three districts to the west of Jigjiga town and
south of the JigjigaHarar
road.
Building wood is also sold in bundles like firewood, but sells
for twice or three times the
price: 57 Birr per bundle. Traders rent lorries from Jigjiga
town to transport building
wood, up to 800 bundles each load. The trip from one
supplying district Borale takes as
long as 36 hours. Traders buy wood from villagers, usually
paying 2 Birr per bundle. Loaders
are also hired to load and unload the wood; they charge 0.1
Birr per bundle. Traders also
have to pay several taxes on the road and in the market itself,
which reduces their profit
margins (see Box 7.2).
7.3.2 Craft workers and artisans

In parts of rural Somali Region where the raw materials are
freely available, craftwork based
on natural products is fairly common. This includes making
mats and baskets from grasses
and reeds, which is practised extensively by communities
living alongside rivers like the
Shabelle and the Dawa/Ganale (Box 7.3).
80
IDS
RESEA
RCH
REPOR
T 57
Box 7.4 Case study: shoe-repairer

Abdi is a 54-year-old man who is married with four
children. At the time of being
interviewed, he owned 15 camels and 30 shoats. Two
of his older sons (aged 21 and 16)
look after the livestock. Abdi has been a shoe- repairer
since 1981, when he started his
one-man business in Somalia. He sees this trade as
giving him a reliable source of
disposable income to buy essential commodities like
wheat, oil, sugar and salt. It is
difficult to rely only on livestock for sustenance. They
are prone to drought and dry
seasons and produce minimal milk that is not even
enough for household consumption.
Abdi said that most Somalis look down upon him, as
they consider shoe-repairing to be
a slave job that is only for people with limited
means. I am the only one doing this
work in this place. Somalis are pastoralists by nature
and they tend to patronise anyone
who is outside it. However, Abdi claims that he is
better off than most pastoralists, as
he makes on average about 25 Birr per day, enough to
meet his households needs and
also make some savings. His business is connected
with pastoralists as most of his
clients are pastoralists. He gets his tools nails, polish
and glue from Bosaso in
Somalia, by placing orders with traders who travel
back and forth to Somalia.
Abdi does not intend to pass on his skills to his
children as he claims that they, like
other Somalis, look down on shoe-repairing and
despise it as a backward occupation.
Source: Qualitative survey interview, Shilabo District.
Some people living in rural communities have artisanal skills
such as shoe-repairing or
tailoring, though this is uncommon. In one case of a shoerepairer from Shilabo (Box 7.4), he
had acquired his skills and experience working as a young
man in a town in Somalia, which
he brought with him to Ethiopia. Although Abdi sees his

work as providing a supplementary
source of income to compensate for the unreliability of
livestock, his clients are mostly
pastoralists, which means this occupation is vulnerable to
derived destitution (Sen 1981)
the possibility that demand for his services will collapse
following a shock to pastoralist
livelihoods.
Some artisans are generally rural-based, such as potters,
weavers or blacksmiths. Often
these specialised skills are passed on from father to son, or
mother to daughter. In Ethiopia
these groups often belong to minority castes, and are socially
excluded within their villages.
Other artisans such as tailors are more usually based in
towns, where the demand for their
services is higher and more reliable. Even in urban areas,
however, artisans are despised,
and one tailor interviewed in Gode town plans to educate his
children in the hope that they
find alternative employment in adult life (Box 7.5).
7.4 Food consumption

Contradicting the image of pastoralists as surviving mainly
on meat, milk and blood, the
analysis of food consumption in Somali Region reveals a
great deal of dietary diversity across
districts and livelihood groups, and a high dependence on
cereals as well as on animal
products. Table 7.4 shows that pastoralist diets are dominated
by milk and meat, but also by
maize and wheat. Farmers also consume milk but have less
access to meat, though riverine
farmers eat some fish, and their staple cereals are maize and
sorghum. Farmers have better
access to vegetables and beans than do agro-pastoralists and
pastoralists.
Agro-pastoralists in this survey consume maize, wheat and/or
sorghum almost equally. Most
consume milk and about half eat meat, but otherwise they
have a very undiversified diet.
Urban diets are the most diversified more than half the
urban residents surveyed regularly
consume rice, pasta, wheat and bread, most also consume
milk and meat, and more than
two-thirds of urban households eat vegetables.
The number of different food items consumed by a household
is a simple but robust
indicator of food security. It reflects the level of dietary
diversity and therefore the quality of
peoples diets. A diverse diet is also associated with higher
food consumption: a 1 percent
81
Box 7.5 Case study: tailor, Gode town

Ahmed is a 30-year-old tailor who has lived his whole
life in Gode town. He married
in September 2003 and the couple has two young
children: a son and a daughter,
both under two years old. Since his fathers death his
widowed mother is also living
with them. Tailoring is his only job. Although his
father was a farmer, Ahmed is a
tailor, having started his one-man tailoring business in
2000. Most of his customers
are women and children. He makes between 15 and 20
Birr a day from making and
repairing clothes. He rents the sewing machine he uses
from a friend, for a rental of
4 Birr a day. Ahmeds wife is not engaged in any
business apart from looking after
the children. Although they are on the food aid
register, they receive very little food
(4 kg of wheat per person) and only very occasionally
once or twice a year.
Due to their limited means, Ahmed does not take his
children to hospital when they
are ill, instead relying on what he calls Gods mercy
to heal them. If all goes well for
them, he plans for his children to go to school and
continue up to university so that
they may improve themselves and support their
parents. But he also intends teach
them basic tailoring skills, in case they cannot make it
in school or other trades. He
values education highly, as he has seen a number of
people whom he knew who are
in good jobs and earning good money. Ahmed says
that Somalis hold negative
attitudes towards tailors, as this is considered as a
profession only for the poor
people with small means.
Source: Qualitative survey interview, Gode town.
increase in dietary diversity is associated with a 1 percent
increase in per capita consumption.
the magnitude of the association between dietary diversity
and caloric availability at the
household level increases with the mean level of caloric
availability (Hoddinott and
Yohannes 2002: iii). A recent categorisation of dietary

diversity identified three distinct
groups of households, based on their food consumption
patterns (WFP 2005: 28):
Low dietary diversity: these households consume less
than four food items per day
belonging to one or two food groups, i.e. cereals and fat.
Borderline dietary diversity: these households
consume five to six food items per day
belonging to only three food groups, i.e. cereals, fat and
vegetable proteins.
High dietary diversity: these households consume
more than six food items per day
belonging to several food groups, i.e. cereals, fat,
vegetable proteins, tubers,
vegetables, fruits and animal proteins.
These patterns of food consumption were found to be well
correlated with other indicators
of food insecurity, including meals per day, average food
gap or food surplus (the
distance between actual consumption and a minimum food
consumption threshold), and
experiences of hunger. Applying the same cut-off thresholds
to our survey households,
agro-pastoralists have the most inadequate diets in terms of
diversity. Farmers diets also
score in the low diversity range, but are slightly more
diversified than agro-pastoralist
diets. Pastoralist diets are borderline in terms of dietary
diversity. Only urban households
have adequate diets characterised by high diversity.
These findings provide some confirmation of the data on
relative cash incomes provided
earlier in this chapter. Urban households are wealthier on all
indicators than rural
households in Somali Region, and they enjoy more
diversified and healthier diets,
consuming more vegetables and fruit even than the farmers
who produce these crops.
Agro-pastoralists are the poorest households in our survey,
and have the most limited diets.
Farmers have more restricted diets than pastoralists, despite
earning marginally more cash
income, on average (Table 7.1). Two explanations can be
offered for this apparent anomaly.
One is to point out that average pastoralist incomes are pulled
down by Shinile, an atypical
part of Somali Region and one of the poorest districts in our
survey. (Average incomes in
82
IDS
RESEA
RCH
REPOR
T 57
Table 7.4 Consumption of different food types, by livelihood

category*
Food type Pastoralist Agro-pastoralist Farmer Urban
Total
Staple cereals
Maize 72% 61% 90% 48% 70.4%
Wheat 61% 61% 39% 79% 57.2%
Sorghum 49% 58% 64% 14% 50.7%
Rice 32% 7% 10% 85% 26.1%
Pasta 18% 1% 4% 73% 17.0%
Bread 9% 1% 2% 54% 10.9%
Barley 0% 1% 4% 3% 1.6%
Meat, fish and dairy
Milk 93% 91% 83% 85% 88.3%
Meat 76% 55% 46% 72% 60.9%
Eggs 6% 0% 11% 16% 7.0%
Fish 1% 0% 16% 7% 5.8%
Vegetables and fruit
Beans 8% 9% 15% 43% 15.2%
Vegetables 8% 1% 8% 70% 14.9%
Fruit 2% 0% 2% 34% 6.1%
Dietary diversity score 4.3 3.4 3.9 6.8 4.3
* This question was not asked in the first phase of the
fieldwork, so no data are available for
Gashamo, Kebribayah and urban Jigjiga.
Source: Household survey data (n=789).
Gashamo and Shilabo are higher than in any farming district
surveyed.) The second point is
that pastoralists are more market-dependent than farmers for
their food needs, so they
tend to purchase a wider range of local and imported food
items than do farmers, who are
more dependent on their own food production.
7.5 Household purchases

Another good proxy indicator of well-being is household
purchases of basic items such as
clothes and groceries. Our questionnaire survey asked
respondents how many times clothes

have been purchased within the last three years by the
household, for both adults and
children. (Both new and second-hand clothes purchases were
recorded; the pattern of
responses is almost identical.) Clustering households by
livelihood activity, urban households
purchase new clothes twice as frequently as rural households
(Table 7.5). Farmers buy new
(and second-hand) clothes slightly more often than
pastoralists and agro-pastoralists, who
again emerge as the poorest category by this single
indicator. Among rural districts,
farmers in the Jigjiga plains purchase clothes most frequently.
This could reflect their close
proximity to the large clothing markets in Jigjiga town, as
access to markets is an important
determinant of consumer behaviour, not just levels of
income.
Survey respondents were also asked if four common grocery
items salt, sugar, tea-leaves
and kerosene were present in the home on the day of the
interview. These items were
identified in pilot testing as basic essentials for Somali
households. A familiar pattern is
found as with other proxy indicators of poverty. Urban
households are most likely to have
some or all of these items in their homes three-quarters of
our urban households had all
four items present when they were interviewed (Table 7.6).
Agro-pastoralist households are
least likely to have these items at home just one in 10 had
sugar and one in 25 had
kerosene when they were interviewed. As with other
indicators of well-being discussed in
this chapter, there is no significant difference between
farmers and pastoralists.
83
Table 7.5 Purchases of new clothes, by livelihood category

Clothes for adults (% households) Clothes
for children (% households)
Livelihood 0 1 2 3 4+ Mean 0 1 2 3 4+ Mean
Pastoral 35 24 36 4 1 1.1 33 18 32 15 2 1.4
Agro-pastoral 29 49 20 1 0 0.9 26 43 27 2 0 1.1
Farmer 16 43 37 3 0 1.3 18 30 38 13 1 1.5
Urban 8 6 34 44 8 2.4 9 2 29 46 14 2.7
Average 23% 34% 31% 9% 2% 1.3 22% 25% 32% 17%
3% 1.6
Once again, it is agro-pastoralist households who emerge as
the worst-off group in our
survey. As one woman from Kebribayah told us: I am so
poor these days that I can only
afford black tea; if I had extra money, I would have milk in my
tea!4 0 This is an important
finding, given the popular view that livelihood diversification

is a rational response to risk,
and that spreading risk among multiple occupations reduces
income fluctuations and offers
more stable income and consumption over time. This does
not appear to be borne out by
our survey data for Somali Region. Even during a drought
that affected pastoralists most
severely, pastoralists and farmers are better-off on every
indicator reported here.
Paradoxically, it seems that agro-pastoralism, by combining
crop farming with livestock
rearing, results in the worst of both worlds.
7.6 Livelihood problems and changes

Households interviewed in our survey reported very different
livelihood problems,
depending on their location and the livelihood strategy they
are pursuing. Among
pastoralists, the drought in Gashamo clearly affected the
ranking of problems reported.
Food shortages ranked much lower in Shinile and Shilabo
than in Gashamo, for instance. In
Shilabo, which has been badly affected by economic shocks
and conflict in recent years, lack
of democracy ranked second, but was not mentioned at all in
either Gashamo or Shinile.
Similarly, shortages of food and water ranked first in
Kebribayah, which was also seriously

affected by the recent drought in 2004, but ranked much
lower in the other agropastoralist districts of Cherati and Doboweyn, which are
further south and escaped the
worst of the drought.
Farmers also saw lack of access to health services as their
most pressing problem, followed
by lack of access to education. This can be interpreted as
reflecting the marginalisation of
the ethnic Bantu of Somali Region, who dominate the
riverine farming communities of
Dolo Odo and Kelafo. After all, one of the governments
arguments for sedentarisation of
pastoralists is that it is easier and more cost-effective to
provide settled communities with
public services such as schools and clinics, yet farmers in
Somali Region who have been
settled for decades see little evidence of schools and clinics
in their villages.
Table 7.6 Basic items in the home, by livelihood category (%
of households)
Livelihood system Salt Sugar Tea Kerosene Mean
Pastoralist 27 21 23 16 21.8
Agro-pastoral 20 10 18 4 13.0
Farming 28 15 24 13 22.3
Urban 89 78 82 75 81.0
Average 36.4 26.6 32.9 22.3 26.8
40 A woman interviewed in Kebribayah.
84
IDS
RESEA
RCH
REPOR
T 57
Table 7.7 Livelihood problems ranked, by livelihood

category
Rank Pastoralists Agro-pastoralists Farmers Urban
residents
1 No health centre No health centre No health centre Lack of
cash income
2 Shortage of food Shortage of food No education/school
Absence of markets
3 No education/school Lack of water Shortage of food Lack
of water
4 Lack of water No education/school No farm equipment
Inadequate shelter
5 No veterinary service No farm equipment Lack of water
Lack of democracy
Across the full sample of 1,100 households, lack of health
services was by far the most
frequently mentioned complaint, followed by lack of
education services. Both of these are
the responsibility of government, and restricted access to
health and education has serious
implications for both current livelihoods and future
livelihood potential. Lack of food and
water ranked third and fourth, with lack of cash income fifth
and other problems such as
lack of veterinary services or farming equipment being
specific to the livelihood context,
and no other problems being mentioned consistently across
livelihood systems.
Household survey respondents were also asked if they or any
member of their household had
ever undertaken any livelihood activity in the past that they
do not do any more. Only 45 (5.7
per cent) answered in the affirmative, suggesting a great deal
of occupational immobility
among Ethiopian Somalis, especially in rural areas. The
biggest change was recorded in
Shilabo District, where 22 out of 100 households had stopped
pursuing a range of previous
occupations that included working in the private or public
sector: running a business, working
as a driver, or working for the government. It is likely that
this evidence of disrupted
employment in Shilabo is related to the recent problems

caused by the governments war
on contraband, which, as noted in Chapter 4, has hit Shilabo
especially severely.
In no other rural districts had more than six individuals
interviewed changed their
occupation. In urban Gode, five households had come to
town having left (or been forced
out of) pastoralism, and a further six individuals in Gode had
pursued different livelihood
activities in the past. It is not unexpected that people in towns
can move between
occupations more freely than rural residents. However, the
fact that very few individuals in
either our rural or urban households had moved out of
pastoralism or farming is both a
reflection on the limited employment opportunities in rural
Somali Region and an
indictment of the education system. Lack of formal education
leaves rural Somalis with no
option but to continue pursuing their parents livestock- or
crop-based livelihoods, no
matter how vulnerable that livelihood strategy becomes.
85
Section
Vulnerability
This section addresses the nature and causes of livelihood

vulnerability in Somali Region.
Chapter 8 quantifies vulnerability in terms of two indicators:
a self-assessment of the
households well-being at different points in time, and
mortality in these households since
1992. Chapter 9 considers drought and hunger in Somali
Region; Chapter 10 explores the
nature and impacts of conflict on livelihoods; Chapter 11
turns the attention to issues of
governance and political representation; and Chapter 12
looks at the relationship between
gender and vulnerability.
8
Livelihood
vulnerability and
mortality outcomes
This chapter presents household survey data on self-assessed
vulnerability and on mortality
outcomes.
8.1 Trends in self-assessed vulnerability

Towards the end of the household survey interview, each
respondent was asked to provide
a self-assessment of their households situation, both now and
at several points in the past,
up to ten years ago.4 1 Four categorical responses were allowed,
corresponding to
descriptions of the household as either accumulating, selfsufficient, struggling, or

destitute. The results are presented for each district subsample in Figure 8.1, where, for
simplicity, the two positive responses are combined into a
single category doing well
while the two negative responses are combined into a second
category struggling (which
is a mirror image of those doing well). It is striking how
each diagram presents a nuanced

depiction of livelihood trends in different parts of the region,
and can be interpreted in a
way that corresponds to our knowledge of major events that
impacted on livelihoods in
these communities at different times within the past decade.
The first diagram in Figure 8.1 summarises the responses
across all 1,100 households
interviewed in ten districts. A distinctive but complex pattern
emerges. Comparing the
situation 10 years ago with today (reading the diagram
from left to right), there has been
a dramatic decline in the number of households reporting that
they are doing well, and a
corresponding rise in households that are struggling to
survive. The proportion of
households doing well has fallen from over 90 per cent in
the mid-1990s to about 30 per
cent in 2004/5, while struggling households have risen from
close to zero to above 70 per
cent. This reversal of fortunes has not occurred in a linear
trend, however; instead, the
diagram provides graphic evidence of the volatility of
livelihoods in Somali Region. Around
the time of the 1999/2000 drought emergency, there was a
dramatic spike in struggling
households (almost 80 per cent, even higher than the present
day), which was followed by
a period of recovery for two to three years in 2002/3,
households doing well once again
exceeded households that were struggling before another
drought-triggered setback in
2003/4 caused the rapid deterioration in livelihoods that was
still visible during our
fieldwork in late 2004 and early 2005.
Even allowing for the nostalgia effect (a tendency to
idealise the past, which can
compromise long-term recall data), it seems reasonable to
conclude that life is more difficult
41 This question was first designed and applied in the Wollo Destitution
Study (Devereux, Sharp and Amare
2003). Four possible responses are read out to the respondent,
corresponding to descriptions of the
household as either accumulating, self-sufficient, struggling, or
destitute.
86
IDS
RESEA
RCH
REPOR
T 57
now for most households in Somali Region than it was ten

years ago. Also, the impact of
livelihood shocks on household viability is clearly visible in
this diagram, illustrating the
volatility, unpredictability and hence extreme vulnerability of
livelihoods in semi-arid regions.
The next three diagrams present the same data for our three
pastoralist districts. Gashamo
and Shilabo, both in the north of the region, have clearly been
more severely affected by
the recent drought than Shilabo. Gashamo was the epicentre
of the 2004 drought, and this
explains why almost all households claimed to be
struggling when interviewed (the highest
proportion of any district surveyed). Shinile has also seen a
rapid rise in struggling
households recently, following a strong and rapid recovery
after the 2000 drought. On the
other hand, Shilabo was relatively unaffected by the 2004
drought, but was very severely
affected by the drought of 2000, which struck central Somali
Region more than the north,
and had its epicentre in nearby Gode District. (Compare the
figures for 45 years ago,
when about 60 per cent of Gashamo households but almost
90 per cent of Shilabo
households were struggling.)
The next three diagrams are for agro-pastoralist districts
surveyed. Kebribayah is located in
northern Somali Region, between Gashamo and Shinile, but
has the same rainfall system as
Gashamo (as described earlier, Shinile has a different rainfall
system) and follows a similar
self-assessment trend to Gashamo. Households in Kebribayah
were quite badly affected by
the 2000 drought, but were recovering well until the 2004
drought knocked them back
almost as severely as households in Gashamo. In both
Gashamo and Kebribayah, the impact
of the recent drought was undoubtedly compounded by the
fact that they had been struck
by a drought just three years earlier, and had insufficient time
to rebuild their herds and
flocks before the next drought struck. In the literature on
famine, the evidence suggests
that two moderate shocks in quick succession can be more

devastating on livelihoods than a
single major shock.42 Doboweyn, in Korahe zone, lies between
Gode and Shilabo District,
and its self-assessment pattern is an exaggerated version of

Shilabos. Being closer to Gode,
agro-pastoralist households in Doboweyn were more badly
affected by the 2000 drought
T re nd s i n Sel f - as ses sed Vul n era bi l it y , So mal i R eg io n
T ren d s
in Sel f as ses sed
Vul n era
bi l it y ,
G ash am
o
100%
10 0%
90%
9 0%
Do in g Well
D
oin
g
We
ll
S tr u gg lin g
80%
8 0%
S tr ug g lin g
70%
7 0%
60%
6 0%
50%
5 0%
4 0%
40%
3 0%
30%
D oin g Wel l
2 0%
20%
Do ing We ll
S
tr
u
gg
lin
g
S t ru gg li ng
1 0%
10%
0%
0%
1 0 y e a rs
Today
a go
4 -5 y e a r s a go
2 y e a rs
a go
1 0 y e a
rs
a go
4 -5 y e a
rs
a go
2
y e a r s a
go
1 year a go
Tod ay
1 y e a r a go
(1 9 9
(2 0 0
(2 0 0
(2 0 0
(200
(1 9 9 4 )
(2 0 0 0 )
(2002)
(2 0 0 3 )
4
0
2
3
4
)
)
)
)
)
(2004)
T re nd
s
i n
Sel f - as
ses sed
Vul
n
era bi l
it y , Sh
il a bo
T ren d s in Sel f - as sess ed Vul n era bi l it y , Sh in i le
100%
10 0%
Doi
We ll
Do in g W ell
90%
9 0%
St ru g g lin g
ng
80%
8 0%
S tr u g gli ng
7 0%
70%
6 0%
60%
5 0%
50%
4 0%
40%
3 0%
30%
D oin g We ll
2 0%
20%
Do in g Wel l
1 0%
10%
S tr ug g
lin g
S tr u gg lin g
0%
0%
1 0 y e a rs
Today
a go
4-5
ye a r s
a go
2 y e a rs
a go
10 years
ago
4- 5
years a go
2
year s
ago
1 year a go
Tod ay
1 y e a r a go
(1994)
(2000)
(2002
)
(20
03)
( 2004)
(1994)
(2 0 0 0 )
(2 0 0 2 )
(2 0 0 3 )
(2 0 0 4 )
42 Dyson and Grda (2002: 13) describe this phenomenon as bangbang famines: food crises often come in
pairs so-called bang-bang famines. Having been weakened by
one food crisis a population will be more
susceptible to another, unless circumstances improve.
87

(cont.)
Tre nd s in
S el f - asse sse d V ul n era bi li t y,
T re nd s i n S el f - ass ess ed V u l ne rab il i t y, D ob o we yn
Keb ri ba yah
100%
100%
Do in
g
Well
90%
90%
St r ug g lin g
Doi ng Wel l
S t ru gg li ng
80%
80%
70%
70%
60%
60%
50%
50%
40%
40%
30%
30%
20%
20%
Do in g Well
Do in g W ell
S tr u gg lin g
10%
10%
S
t
ru g
gli
ng
0%
0%
1 0 ye a r s
To day
a go
4 -5
y e a rs
a go
2 y e a rs
a go
10 ye a r
s
a go
4 -5 y e a
rs
a go
2 years
a
g
o
1
year
ago
Tod a y
1 y e a r a go
(1
(2
(2
(2
(2
(1 9 9 4 )
(2 0 0 0 )
(2 0 0 2 )
(2 0 0 3 )
9
0
0
0
0
9
0
0
0
0
4
0
2
3
4
)
)
)
)
)
(2 0 0 4 )
T ren ds i n Sel f - as sesse d Vul n erab i li t y, C he rat i

Tre nd s i
n S el f ass esse d
V u ln era
bi l it y , Ji
gj i ga ( r
ur al )
100%
100%
90%
90%
S t ru gg lin g
Do in g
80%
80%
W el l
Doi ng We ll
St r ug gl ing
70%
70%
60%
60%
50%
50%
40%
40%
30%
30%
S tr ug g lin g
S tr ug g lin g
20%
Doi ng Well
Do in g Well
20%
10%
10%
0%
2
years
To day
0%
4-5 year s ago
1 year ag o
10 year s ago
ago
1 0 y e a rs
Tod ay
(1994)
a go
4 -5
(2 0 0 0 )
ye ars
(2 0 0 2 )
a go
(1994)
( 2000)
(2 002)
(2003)
(2004)
1 year ag o
2 years a go
(2 0 0 3 )
(2004)
Tre nd s in
S el f - asse sse d V ul n erab i li t y,
Tr en ds i n S el f - as sess ed V u ln er ab il i t y, Do l o Od o
100%
Kel af o
100%
90%
90%
Do in g
S tr u gg lin g
Doin g
80%
80%
Well
Wel l
S t ru g gli ng
70%
70%
60%
60%
50%
50%
40%
40%
S tr ug gl in
g
30%
30%
D oin g Well
S tr u gg lin g
20%
20%
Doin g Well
10%
10%
0%
0%
10 year s ago
Tod ay
(2002)
( 2004)
(1994)
4- 5 years ago
(2000)
( 2002)
1 0 y e a
rs
ag o
4-5 year s
ago
2 years
a
go
1 yea r ago
Today
1 year ago
( 2000)
(2003)
2 years ago
(1994)
(2003)
(2004)
T ren d s i n Sel f - a sse sse d Vu l ne ra bi l it y , J ig j ig a ( u rb an )

Tr end s i n S el f - ass esse d V u ln era bi l it y, Go d e
1 00%
90%
100%
90%
Do
in g
Well
80%
80%
Do in g Well Do in g We ll
70%
70%
60%
60%
Doin g Well
50%
50%
St ru g gl ing
40%
40%
St ru g gl in g
30%
30%
Str u gg lin g
20%
20%
St
ru g
gl
ing
10%
10%
0%
0%
1 0 y e a
rs
ag o
4-5 year s
10 year s ag o
Toda y
4-5 ye ars ag o
ago
2 years
a
go
1 yea r ago
Today
1 yea r ago
2 yea rs ag o
(1994)
( 2000)
(2002)
(2003)
( 2004)
(199 4)
(2000)
(20 02)
(2003)
(2 004)
and their recovery was slower. The recent drought has also
affected them quite badly,
reversing the positive trend since 2001.
Agro-pastoralists in Cherati and farmers in Dolo Odo have
experienced similar patterns of
vulnerability over the past decade. Both are located in the
south of the region, and both
reported an improvement in their livelihoods within the last
year the only two districts to
display this encouraging trend. On the other hand, in both
districts the proportion of
struggling households has consistently exceeded the
proportion doing well since the
drought of 2000, which was more severe in central and
southern zones of Somali Region.
Since this major setback in 2000, these southern Somali
communities have been steadily
88
IDS
RESEA
RCH
REPOR
T 57
recovering. The 2004 drought affected them only mildly, and

shows up on the diagrams as a
blip that interfered with an ongoing recovery process
which appeared to be back on track
by mid-2005.4 3
Our sampled households in Kelafo and Dolo Odo are riverine

farmers, whereas those
interviewed in rural Jigjiga are sedentary farmers living on
the Jigjiga plains. The diagram for
Jigjiga farmers displays the same distinctive double crossover pattern as for pastoralist
and agro-pastoralist households in other northern districts
Gashamo, Shinile and
Kebribayah. Most households in rural Jigjiga were doing
well ten years ago, but the
majority were struggling around the time of the 2000
drought, then a recovery period
followed during which those doing well exceeded those
struggling, until the 2004
drought reversed these numbers yet again. The time-lines for
farmers in central and
southern Somali Region are rather different. Farmers living
along the Shabelle river (in
Kelafo) or the Dawa/Ganale river (in Dolo Odo) were
affected by the 2000 drought
almost as severely as pastoralists and agro-pastoralists in
these zones. Although the 2004
drought did not impact significantly on their harvests, most
households in Kelafo reported
deteriorating livelihood conditions within the previous year,
while Dolo Odo households, as
noted, recorded a slight improvement in their circumstances.
The final two diagrams in Figure 8.1 are for the two urban
sub-samples of Jigjiga and Gode
towns. One significant feature of these diagrams is that the
urban samples are the only two
where more households self-report as doing well than
struggling. A second feature to
note is that livelihoods among both urban samples are less
volatile than in the rural districts.
Although the 2000 drought had a visible impact on
livelihoods in Gode, this effect is less
pronounced than in rural districts nearby like Doboweyn and
Kelafo, and there has been
little change in Godes profile from year to year since then. In
urban Jigjiga, well-being is

reported as worsening over time, but at a linear rate rather
than being buffeted by major
shocks at irregular intervals. One possible explanation for the
steady decline in self-assessed
well-being is an influx of people displaced by droughts from
rural areas, having lost the
livestock that are the basis of their livelihoods. Another
possible factor is the nostalgia
effect, but even if respondents in Jigjiga town are
romanticising the past, it is clearly the
case that livelihoods are less volatile, and that people feel
better off, in urban areas.
Figure 8.2 presents the district-level data in the form of a
self-assessment index (which
scales the responses to the self-assessment question between
0 and 1), and also highlights
results for households defined as destitute dependent on
assistance for their survival. The
left-hand figure displays a consistent pattern of declining
well-being from ten years ago to
the present, but with a significant event in the middle of this
period that divides the
trajectory into three phases. The drought of 1999/2000
caused a sudden collapse in wellbeing of varying magnitude biggest in Cherati, Doboweyn
and Dolo Odo, least in the
urban Jigjiga sample followed by a recovery period that
was ended abruptly in most subFigure 8.2 Trends in self-assessment index and destitution,
Somali Region
T r en d s i n
Sel f-a s se ss me n t in d e x,
b y
T r en d s i n Se lf- as se s se d D e sti tu ti o n , b y Di s tri c t
1.0
Di s tri c t
1 00%
G od e
Ji g j i ga U
Go de
Jigjiga U
90%
D ol o O d o
0.9
Do l o O do
K el afo
Ji g j i ga R
Kelaf o
Jigjiga R
80%
C her ati
Ch erati
0.8
D obo w eyn
eyn
K ebr i bayah
Do bow
70%
Keb ri ba yah
S hi l abo
Shi l ab o
S hi ni l e
0.7
Shi n i l e
60%
G asham o
Gas hamo
O ver al l
Over all
50%
0.6
40%
0.5
30%
0.4
20%
0.3
10%
0%
0.2
10 yea rs ag o 4-5 ye ar s ag o 2 yea rs ag o 1 ye ar ag o T o day

ye ar s a go 4- 5 y ear s ag o 2 ye ar s a go 1 yea r ag o T od ay
10
43 One factor that could partly explain why Cherati and Dolo Odo display
uniquely rising numbers of households
doing well at the time of interview is that the southern leg of
fieldwork was conducted last in the
sequence, during the Gu rains of May 2005, while households in
central and northern Somali Region were
interviewed in late 2004 and during the harsh Jilaal dry season of
early 2005.
89
Table 8.1 Mortality in Somali Region, 19912004/05, by

district
District
Households
Deaths
Deaths per Deaths per
affected Male Female Total household
affected HH
Kelafo 51 (51%) 72 55 127 1.3 2.5
Gashamo 61 (61%) 70 59 129 1.3 2.2
Doboweyn 51 (51%) 70 48 118 1.2 2.4
Shilabo 56 (56%) 64 53 117 1.2 2.1
Dolo Odo 58 (58%) 60 39 99 1.0 1.7
Kebribayah 59 (59%) 45 44 89 0.9 1.5
Jigjiga town 43 (43%) 45 27 72 0.7 1.7
Cherati 43 (43%) 41 30 71 0.7 1.7
Shinile 37 (37%) 33 15 48 0.5 1.3
Jigjiga rural 29 (29%) 31 15 46 0.4 1.6
Gode 20 (20%) 18 8 26 0.3 1.3
Total 508 (46.3%) 549 393 942 0.9 1.9
samples by the drought of 2004, which has caused a further
deterioration in well-being
that is sharpest in Gashamo and Kebribayah. This most recent
collapse in livelihoods is very
visible in the diagram on the right-hand side of Figure 8.2,
which records households selfassessing as destitute (unable to cope without assistance) as
being close to 90 per cent in
Kebribayah and well over 90 per cent in Gashamo, but in the
range of only 5 per cent to
20 per cent in all other districts surveyed. One caveat should
be noted: the timing of the
survey immediately after the drought peaked in Gashamo
and Kebribayah but several
months later in other districts will have influenced peoples
perceptions of their situation
at the time they were interviewed. On the other hand, the
timing and relative magnitude
of major livelihood shocks across districts corresponds
surprisingly accurately with what we
know about recent events in Somali Region.
8.2 Mortality
Premature death or excess mortality is one simple but
robust indicator of ill-being. It can

isolate which agesex cohorts are most vulnerable, where the
vulnerable are located, and
which years or seasons are associated with heightened risk.
The household survey
questionnaire asked respondents to report on all deaths that
their household had suffered
since the fall of Mengistus government in 1991.4 4 Almost half the
households surveyed
(n=508/1,098 =46.3 per cent) had suffered one or more

deaths in this period of 1314 years.
The total number of deaths recorded was 942, making an
average of 1.9 deaths in each of
the 508 affected households (see Table 8. 1, which ranks total
mortality by district sub-sample).
Reported mortality was significantly higher in rural districts
than in urban centres. In six out
of nine rural districts, more than half the households
interviewed had suffered one or more
deaths since 1991, with the average being close to two deaths
in each affected household.
The proportion of affected households was highest in
Gashamo (61 per cent) and Kebribayah
(59 per cent), perhaps reflecting tragic memories of the
recent drought which had its
epicentre in these two districts. The number of deaths per
affected household was highest
in Kelafo (2.5) and Doboweyn (2.4). In the urban sample of
Gode, fewer households had
44 The fall of the Derg regime was identified in pilot testing as a
memorable event for recall purposes. Note
that the data on mortality (and all other demographic data) presented
in this report were not collected in a
rigorous epidemiological survey, but as one module in a household
questionnaire survey of 1,100 households,
sampled from a region of four million people. These figures should
not be extrapolated to Somali Region as a
whole, but should be regarded as indicative of relative levels and
trends across household categories.
90
IDS
RESEA
RCH
REPOR
T 57
Table 8.2 Mortality in Somali Region per household, by age

cohort
Age Total By gender By livelihood system
cohort Deaths Per HH Male Female Pastor Agro Farm
Urban
04 559 0.51 0.29 0.22 0.56 0.55 0.65 0.16
59 6 5 0.06 0.03 0.03 0.09 0.0 5 0.06 0.0 3
1014 32 0.03 0.02 0.01 0.03 0.04 0.03 0.01
1519 25 0.02 0.01 0.01 0.04 0.01 0.03 0.01
2029 45 0.04 0.02 0.02 0.06 0.05 0.02 0.03
3039 36 0.03 0.02 0.01 0.04 0.03 0.02 0.06
4059 47 0.04 0.03 0.01 0.04 0.05 0.02 0.07
6079 81 0.07 0.04 0.03 0.06 0.10 0.05 0.10
80+ 5 1 0.05 0.04 0.01 0.06 0.0 5 0.03 0.0 4
Total n/a 0.86 0.50 0.36 0.98 0.92 0.91 0.50
Deaths 941 n/a 548 391 294 276 292 99
suffered deaths (20 per cent) than anywhere else, and the
total number of deaths was
lowest (just 26 reported in 100 households) (Table 8.1).
Four times as many children died in rural households (0.59
deaths of children under five per
household) than in urban centres (0.16 under-five deaths per
household) (Table 8.2). The
highest concentration of infant and child deaths is in farming
households, which also have
the lowest average (mean) age at death just 10.7 years. This
might reflect a higher
susceptibility among farming communities to water-borne
and communicable diseases, due
to their densely settled permanent villages and proximity to
disease vectors such as
mosquitoes. In rural areas in general the average age at death
(15.2 years) is less than half
that in urban areas (33.3 years). When districts are ranked by
average age at death, the two
urban centres of Jigjiga and Gode top the list, followed by the
six pastoralist and agropastoralist districts (where the range is 23.7 years in
Kebribayah down to 12.1 years in
Shilabo), with the three farming districts clustered at the

bottom (8.7 to 11.1 years) (Table
8.3). In both urban samples, a significant number of (male)
household heads had died,
leaving widows who moved to town being unable to survive
in the rural areas.
The number of deaths reported by year exceeded 100 only
three times, all within the last
five years: 2000, 2001 (when deaths for a single year peaked
at 129) and 2004 (Figure 8.3).
Table 8.3 Mean age at death in Somali Region, by location
and livelihood
District Age Livelihood Age (Standard deviation)
Jigjiga town (Urban) 36.4 Urban-based 33.3 (29.4)
Gode (Urban) 24.5 Agro-pastoralist 18.5 (27.7)
Kebribayah (Agro) 23.7 Pastoralist 16.5 (25.6)
Gashamo (Pastoral) 19.8 Farmer 10.7 (21.9)
Shinile (Pastoral) 18.4
Cherati (Agro) 17.5
Doboweyn (Agro) 15.2
Shilabo (Pastoral) 12.1 Location Age
Dolo Odo (Farming) 11.1 Rural 15.2
Kelafo (Farming) 11.1 Urban 33.3
Jigjiga rural (Farming) 8.7
Average 17.2
91
Figure 8.3 Deaths per year, 19902004

1 40
1 20
1 00
80
60
40
20
0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
Figure 8.4 Deaths per household and by season, 19922004

1. 0
3 5%
3 4 .2 %
0 .9 8
0 .9 2
3 0%
0 .9 1
0. 8
2 7 .7 % 2 5%
0. 6
2 0%
1
9
.2
%
1
9
.0
%
0 .5 0
1 5%
0. 4
1 0%
0. 2
5%
0%
0. 0
P a st o ra l A gro- pa s t ora l Fa rm e r U rba n
la al
G u' Hag aa De yr Ji
It is not possible to assess how much of this is due to the

recency effect (recent events
being easier to recall), but these years correspond to episodes
of major droughts and
livelihood stress in Somali Region, so it is reasonable to
attribute many of these deaths to
these traumatic events. In the most recent five-year period
alone (20004), these 1,100
households have suffered over 500 deaths (60 per cent of all
reported deaths since 1990),
at 0.46 deaths per household.
Confirming the fact that mortality is related to environmental
stress factors is the
breakdown of deaths by season. Jilaal is the hardest season in
Somali Region, and this is
reflected in the fact that more than one in three reported
deaths occurred during Jilaal
(Figure 8.4). The main rainy season is Gu, and reported

deaths in this season were also
higher than during Hagaa and Deyr, probably because the
incidence of diseases like malaria
and diarrhoea is higher during the rains.
Figure 8.4 also confirms the locational bias of mortality in
Somali Region, with almost twice
as many deaths per household recorded in rural areas (0.94)
than urban centres (0.50).
92
IDS
RESEA
RCH
REPOR
T 57
9 Drought and hunger

Fieldwork for this study was undertaken in the aftermath of a
severe drought in Somali
Region. This chapter explores the devastating impacts that
drought inflicts on rural
livelihoods in dryland areas, not only for livestock producers
and their families but also for
crop farmers, traders and service providers whose income
depends, directly or indirectly, on
livestock and/or rainfall. Next, we challenge the popular
assertion that droughts in Somali
Region are more frequent and more severe than in the past,
by analysing available longterm rainfall data. Finally, we analyse access to water for
different households, by their
primary source of domestic water berkad, wells, rivers and
piped water and consider the
costs and vulnerabilities associated with each source.
9.1 Drought in Somali Region

Drought is part of the normal cycle of life in arid and semiarid areas, where rainfall is low at
the best of times and abnormally low every few years.
Pastoralist livelihoods are sensitively
attuned to conditions of low and variable rainfall. Droughts
cannot be accurately predicted, but
they are expected and pastoralist systems are well adapted to
drought cycles. It follows that,
while drought is a major risk factor affecting livestock-based
livelihoods, the main source of
vulnerability derives from the inability of pastoralism and
related livelihoods to cope with drought.
In other words, it is not drought as such that makes
pastoralists vulnerable, but factors that
constrain highly evolved drought-response mechanisms,
especially mobility of people and animals
conflict, legal restrictions on trade, and so on. If these
factors deteriorate over time,
vulnerability to drought increases even if the incidence and
severity of drought does not. Various
vulnerability factors are examined in later chapters; here we
focus on drought itself.
Drought was cited as the number one risk to peoples
livelihoods in almost every rural

community visited in fieldwork for this study, whether the
dominant livelihood activity
locally was pastoralism, agro-pastoralism or farming. This
response was undoubtedly
influenced by the fact that the fieldwork took place in the
aftermath of a severe drought in
Somali Region in late 2003 and early 2004. Having warned
earlier during 2003 of rising
food insecurity in the Greater Horn of Africa (GHA), in
November a regional famine early
warning system warned: The late onset of October
December (deyr) season rainfall will
further imperil the food security of pastoralists over the
Somali Region (FEWS NET
2003a: 2). In December, the monthly early warning bulletin
reported:
Successive seasons of poor rainfall in 2003 and prior
years, are compromising food
security in several countries in GHA. An estimated 11
13 million people are
experiencing drought induced food insecurity as well as
livelihood crises, reduced
resiliency in coping capabilities, policy issues, poorly
working markets and disrupted
alternative sources of food and income. Ground
reports indicated scarcity of pasture
and water for livestock with consequences of abnormal
but localized livestock
migrations. (FEWS NET 2003b: 2)
By February 2004, at the height of the Jilaal season, FEWS
NET (2004a: 1) was reporting
that the eastern parts of Somali Region of Ethiopia [are]
currently experiencing droughtinduced food insecurity. The following month, conditions
were worsening: food insecurity
in some countries in the Greater Horn of Africa continues to
deteriorate, largely because of
an increase in chronic vulnerability compounded by conflict
and poor rainfall performance.
Food security conditions in Somali Region require close
monitoring (FEWS NET 2004b: 1).
In April 2004, FEWS NET (2004c: 1) summarised the
situation in Ethiopia as follows: Five
years of rainfall variability, poverty and currently high food
prices have heightened food
insecurity of 7.1 million people currently requiring food aid,
who are mainly pastoralists and
chronically food insecure farmers.
A food security assessment in Somalia in September 2004,
following the Gu rain s
(AprilJune), estimated that 700,000 people were in a state of

humanitarian emergency
93
Box 9.1 Food insecurity in Warder Zone, Somali

Region, August 2004
The food security situation in Warder Zone is poor
and deteriorating due to a poor
performance of the Deyr 2003 and Gu 2004 rains, a
continued presence of large
numbers of livestock from Somalia, and water
shortages due to borehole breakdowns.
There are serious pasture and water shortages in most
parts of the one, particularly in
the eastern districts of Galadi and Bokh and Eastern
parts of Warder District, where
livestock concentrations are higher than normal. Most
of these areas which are berkad
dependent are currently running out of options for
accessing water, and frequent and
confused stress migrations are being seen. Expensive
private water trucking has already
started and in the worst hit areas one 200 lt drum costs
30,000 SoSh (up from
5000 normally). Initial reports pertaining to the
mortality of cattle and sheep are
just coming from the worst affected areas of the Zone.
The security situation seems
to be improving under the ongoing reconciliation and
negotiation efforts. Displaced
households are beginning to return home, but pastoral
resource sharing has not yet
started in the rural areas, indicating that the situation is
still fragile.
or livelihoods crisis. Most of these people live in parts of
Somaliland and Puntland (such as
the Sool plateau) adjacent to Somali Region, which is
experiencing a similar combination of
environmental and governance problems, including
protracted localised droughts, charcoal
production, and civil conflict. This combination of chronic
vulnerability factors and shortterm shocks has been extremely severe for pastoralists in
Somalia. Dominated by pastoral
livelihoods, some areas in the North have suffered
cumulative livestock deaths of upwards
to 60 per cent for goats/sheep, and 80 per cent for camels.
Already there are large
numbers of destitute people who have dropped out of

pastoralism and are now entirely
dependent on social and humanitarian support (FSAU 2004
1). The situation in
neighbouring Somali Region at the same time was almost as
serious (see Box 9.1).
Visible evidence of the most recent drought was provided by
the significant numbers of
animal carcasses cattle, sheep, goats and even camels
seen by members of this study
team in a scoping visit to Gashamo District, the epicentre of
the crisis in Somali Region, in
July 2004. This drought was given the local name Tuur ku
Qaat, meaning carry on your
back when their pack animals died, pastoralists on the
move had to transport everything
themselves.
One immediate consequence of drought is to exacerbate the
seasonal hunger that is a fact
of life in rural communities throughout tropical Africa
(Chambers, Pacey and Longhurst
1981). In Somali Region the hungry season coincides with
the hot dry season (Jilaal), which
runs from January to March and is associated with
heightened vulnerability and higher
mortality rates than any other time of year (as seen above).
In pastoralist areas the hungry
season is normally at the end of the dry season, when milk
availability is low and animals are
in poor condition (SC-UK 2004: 183).
During the most recent Jilaal season before our survey, more
than eight households in ten
reported suffering a shortage of food (n=935/1,094 =85.5 per
cent), while only one
household in seven had sufficient food (n=159/1,094 =14.5
per cent). According to this
indicator of hunger, food shortages were almost universal in
rural districts such as
Doboweyn (100 per cent), Shinile (99 per cent), Dolo Odo
(98 per cent), Cherati and
Kebribayah (97 per cent). Food shortages were less prevalent
in our urban samples of Jigjiga
(27 per cent) and Gode (62 per cent) (Table 9.1), which is not
surprising, since urban
livelihoods are less dependent on rainfall and urban residents
tend to earn higher and less
variable incomes.
Shortages of food translated into rationing of consumption,
following the same pattern
across districts. Only 15 per cent of adults in our survey ate
three meals a day throughout
the preceding year, even during the worst months. More than
half reduced their
consumption to two meals a day, while most adults in

drought-affected Gashamo and
94
IDS
RESEA
RCH
REPOR
T 57
Table 9.1 Indicators of hunger in 2004, by district

District Food Meals per day: Adults* Meals per day:
Children*
shortage 0 1 2 3 Mean 0 1 2 3+ Mean
(%) (%) (%) (%) (%) (%) (%) (%) (%)
Pastoralist 92 3 40 49 7 1.6 2 8 76 13 2.0
Gashamo 93 9 76 12 3 1.1 3 13 80 4 1.9
Shinile 99 0 11 87 1 1.9 1 5 80 11 2.0
Shilabo 85 1 34 48 17 1.8 1 6 68 25 2.2
Agro-pastoral 98 0 41 56 1 1.6 0 15 77 7 1.9
Kebribayah 97 1 65 32 1 1.3 0 18 79 2 1.8
Doboweyn 100 0 31 65 2 1.7 1 10 82 5 1.9
Cherati 97 0 27 72 1 1.7 0 18 69 13 2.0
Farmer 94 1 25 69 4 1.8 2 12 77 9 2.0
Jigjiga rural 89 0 9 82 9 2.0 4 8 75 13 2.0
Kelafo 95 4 33 59 3 1.6 0 10 80 9 2.0
Dolo Odo 98 0 33 67 0 1.7 1 17 76 6 1.9
Urban 45 0 10 28 62 2.5 1 2 25 73 2.7
Jigjiga town 27 0 13 14 73 2.6 0 0 17 83 2.8
Gode town 62 0 7 42 50 2.4 1 3 32 63 2.6
Average 85.6 % 1% 31% 53% 15% 1.8 1% 10% 68%
21% 2.1
* 0 = Sometimes passed a whole day without eating
anything; 3+ means that children
sometimes eat four times a day.
Kebribayah cut down to just one meal a day. A clear
indication that the drought had its
epicentre in Gashamo is the evidence that rationing was most
severe in this district, where
many adults and some children passed entire days without
eating. Conversely, food
consumption in urban Gode and Jigjiga was apparently
unaffected by the drought average
consumption by adults remained well above two meals per
day throughout the year, and
close to three meals for children in Jigjiga. Empirical
evidence from studies of food
insecurity and famine in a range of contexts reveals that

parents tend to protect the food
intake of their children preferentially for instance, when
parents cut down to two meals
their children continue to eat three times daily. Our survey
finds confirmation that this
behaviour is practised in Somali Region. In all districts,
adults protected the food
consumption of children by rationing their own food intake
more severely (Table 9.1).
Although drought and hunger were severe in many
communities in 2004, it was not
exceptional. Somali Region has in fact suffered a series of
droughts since 1999/2000, and
some areas have faced three years or more of below average
rainfall. Earlier droughts were
also mentioned by people interviewed for this study as major
livelihood shocks. In central
and southern districts, the drought of 1999/2000 was said to
be the most severe in living
memory, as it caused substantial human and animal mortality.
People started feeding on the
carcases of dead animals and some even boiled animal skins
to feed on. In Kelafo District,
near the epicentre in Gode, a local name for this drought was
Barbar, meaning
inflammation, because those affected had swollen bellies and
necks.45 In Shinile District,
the same drought was named Soodaf (abandoned),

because all that was left of animals
that died of thirst were their carcases, which were abandoned
to hyenas. The famine of
45 A food crisis in Malawi in 2002 was also nicknamed the swelling by
local people, because many adults and
children suffered from oedema which caused their bellies, legs and
arms to swell.
95
1974 was given the name of Abarti daba deer (long-tailed)

by people throughout the
Ogaden and Somalia, because it was triggered by a drought
that lasted for two years.
There is a widespread perception in Somali Region that
droughts occur more frequently
than in the past. I remember when I was a child, the droughts didnt happen so often.4 6
People also believe that periods between droughts are shorter,

interrupting the recovery
process. Nowadays, it seems theres a drought just after two
or three years. In the past
there was enough time between droughts for restocking before the next drought hit.47
One consequence of recurrent droughts in rapid succession is

that poverty ratchets are
set up, especially in pastoralist households which need longer
to rebuild herds and flocks
than farmers need to return to normal harvests. Livestock
traders also suffer from this
ratchet effect, as their capital is tied up in their animals.
One trader explained how he lost
his bulls in the drought of 2000, so, having less capital,
switched to sheep, only to lose
them in the drought of 2004.
In 2000, I had accumulated some 80 bulls, which I
hoped to sell at a good profit.
When the drought hit, I lost the bulls. They died near
Ali Jama. In 2003 I bought 120
male sheep which I planned to fatten for sale in the
following year. Now am I left
with 40 sheep, which I cannot see surviving this drought.4 8
Another reason why the drought of 2004 had such a heavy

impact on pastoralists in
Gashamo is that its severity was underestimated, so
appropriate pre-emptive measures
such as selling animals before prices collapsed were not
taken by most local livestock
owners.
Before the drought I had 150 goats and 55 cattle, but I
have only 10 goats left now.
We did not think this drought would be so severe. We
have met many droughts
before, and only lost a few animals each time. We didnt
realise until it was too late to
sell the animals that it was such a bad drought.49
A typical feature of drought in pastoralist communities is a

decline in the barter terms of
trade (livestock to grain prices), as the value of livestock

falls while staple food prices rise.
This was documented in Turkana, Kenya in the 1980s (Swift
1989), and also affected
Gashamo in 2004. People are dependent on animals which
are affected by the drought.
Most of the animals are very ill and weak and we can no
longer barter-exchange a grown
sheep for the 50-kilo sack of food.5 0
Although droughts affect livelihoods in rural areas most

directly and immediately, since
livestock and crop production depend directly on rainfall,
traders and service providers
whose business depends on rural incomes are also negatively
affected, as they face declining
demand for their goods and services. Retail stores in rural
settlements lost income, and
several stores were forced to close down. (Most businesses
are struggling to find enough
customers; The families who were our customers are
buying only one quarter what they
used to take.) Another case in point is the market for milk.
While exporters complained of
declining supplies of milk in 2004, people who sell milk
locally complained of declining
demand. During the drought period, the supply of milk goes
down. This month [November]
we are suffering because of limited supplies. This is not
usual; we used to export large
quantities of milk at this time of year. We also struggle to get
buyers.
46 Case study interview, female khat seller, Gashamo.

47 Case study interview, woman stall-owner, Gashamo.
48 A jeble in Gashamo.
49 Case study interview, male pastoralist, Gashamo.
50 Agropastoralist, rural Kebribayah.
96
IDS
RESEA
RCH
REPOR
T 57
This indirect impact of drought on livelihoods has been

labelled derived destitution (Sen
1981), and it has a levelling effect on most people in the
drought-affected area. The last
drought made us all equal in terms of poverty: we call it Sima
the equaliser.5 1 The reason
for this indirect impact on businesses that do not depend

directly on rainfall is that
households facing food stress impose austerity measures
almost immediately for instance
by rationing food consumption and cutting back on nonessential spending and there is
not enough buying power in those parts of the economy that
are not reliant on rain. One
group who suffered the consequences of this coping
behaviour in 2004 were traders who
sell cloth to pastoralists:
I am a cloth seller. It is not a good business these days.
Before the drought, men
would come into the village to sell animals and buy
cloths. Now everybody is poor,
and the men have no animals and no spare money to
buy things like cloths which are
not a necessity.52
A woman who runs a stall selling groceries in Gashamo town

explained how drought impacts
on her business. When there are good rains, we get lots of
good customers. Then the
pastoralists come and buy goods from me. But when the rains
are bad, my business is also
bad. A butcher in Gashamo explained how drought
undermines his business in complex ways:
Because of the droughts, there are periods when there
are few animals like goats.
Then, the prices we pay for these animals are still high.
At the same time, because the
goats are very thin and the quality of the meat isnt so
good, people arent willing to
pay good prices for the meat. So, we can end up making
a loss or just making a little.
In non-drought years, our profits are greater because the
animals are cheaper, and our
customers will pay more because the animals are

healthy and fatter.
Very few income-earning activities are impervious to the
general impoverishment that
drought brings. An intriguing exception is running a tea-shop.
The best business, the one
that is most drought-resistant, is having a tea-shop, because
tea is cheap! There is always a
market for it, even during droughts.53 Another business that
actually profits from droughts
is khat trading and retailing, because unlike other nonessential spending demand for
khat increases when times are hard.
I sell khat. Luckily, khat isnt badly affected by the
drought. In fact, it seems now
because of the drought, men are chewing more khat, so
my business is actually better
than before when men were busy with their animals.54
This is one way in which droughts are blamed for

undermining not only the economic basis
of the pastoralist system but also the basis of the culture and
society, as traditional sharing
institutions are undermined and men who have lost their
animals are reduced to chewing
khat.
These droughts are destroying the Somalis because
theyre destroying our culture. They
are making it impossible for us to practise our
traditions. The droughts are also causing
men to chew so much khat thats also affecting our
traditions. Just like the fact that
the droughts are wiping out the mens livestock
livestock are central to Somali
culture.5 5
51
52
53
54
55
Life history inter view, female teashop owner, rural Gashamo.

Female focus group participant, rural Gashamo.
Case study interview, woman khat seller, Gashamo.
97
Box 9.2 Environmental degradation in Somalia

Evidence indicates that environmental degradation is
accelerating and undermining
current and future livelihoods and food security
throughout Somalia. Uncontrolled
destruction of the environment is severely affecting all
productive areas, not only
pastoral rangelands, but also productive agricultural
lands. In pastoral rangelands,
environmental degradation is resulting from a number
of factors, including excessive
tree clearing for charcoal production and exportation,
lack of rangeland
management, increased sedentarisation due to
proliferation of berkeds, and one of
the worst prolonged droughts in recent times. In
agricultural productive areas, it is
due to uncontrolled and accelerated tree clearing and
charcoal production, invasion
of foreign weeds and noxious plants, encroachment of
sand dunes, and increasing
trends of erratic and below normal rainfall.
Droughts also cause a rise in domestic stress and even
marital breakdown, as men lose their
central economic role and status within the family, and
women lose patience with husbands
who, having lost their animals, do nothing to bring in income
to feed the children.
I have two daughters and two sons, but Im divorced.
The reason I left my husband was
because he was just sitting and not doing anything, not
earning any income. I couldnt
stay with someone who was not doing anything useful.56
The perception that droughts are increasing in Somali

Region, and that their cumulative
impacts are placing intolerable stress on local livelihoods, is
contributing to growing
disillusionment with pastoralism as a way of life.57 All our lives
we wasted in pastoralism and
we have nothing to show for it as we are now herd-less due to
drought.5 8 An ex-pastoralist
family, displaced to Sagarabuur IDP camp on the outskirts of

Gode town by the drought of
2000, spoke about how the recent succession of droughts has

created a new class of
people a herd-less society.
I dont want a pastoralist life anymore because Im too
scared of droughts. The
droughts arent going to stop: there are too many people
and too many animals, and
the land and grazing is getting less. The rains are not as
much: everything is drier and
hotter. Trees that we grew up with have died and will never come back again.59
When asked why the impacts of droughts today are more

severe than in the past, many
people echoed this Malthusian explanation too many
people and too many animals
and the effects of these demographic pressures on the natural
resource base, as people
adopt coping strategies that exacerbate environmental stress,
such as fencing off land and
burning charcoal.
In the past, there were many places where you could
take your animals without going
so far away. There were fewer people so there was a lot
of space. The other problem
today is that some people are fencing off land so there is
less space to roam with
animals. No people used to do that before. People are
fencing off land because they
want to burn it and use it for charcoal burning. This is
also making it difficult to find
land easily for animals.6 0

57 This issue of changing Somali attitudes to the future of pastoralism is
explored later in this report.
58 Womens focus group participant, Shinile District.
59 A woman in Gashamo District.
60 A woman in Gashamo District.
98
IDS
RESEA
RCH
REPOR
T 57
Similar pressure on natural resources is reportedly occurring

in neighbouring Somalia, which is
environmentally similar and, in many ways, economically
integrated with Somali Region (Box 9.2).
9.2 Rainfall trends

Since at least the 1960s, the Food and Agriculture
Organisation (FAO) and other outsider
experts have predicted the demise of pastoralism in Ethiopia,
because of human and
livestock populations exceeding the carrying capacity of the
land. In February 2005, FEWS
NET argued that pastoralist livelihoods in the Greater Horn
of Africa are collapsing:
Continuing drought, increased competition for grazing
lands, conflict, disrupted livestock
trade, and limited regional market options put the
sustainability of the pastoralist livelihood
in question (FEWS NET 2005: 1). In support of this
argument, FEWS NET presented a
rainfall graph for selected pastoral areas (including Gode in
Ethiopia), illustrating that annual
rainfall has been well below 300mm the minimum amount
of annual rainfall needed to
support viable pastures in the Horn for most of the ten
years between 1995 and 2004.
A closer look at long-run rainfall data from Somali Region
reveals a more complex story. It is
true that rainfall in Gode is very low, and that there appears
to be a decline since the mid1990s. In 1997, Gode enjoyed almost 500mm of rain (a
record high since at least 1972), but in
the drought year three years later, just 134mm. However,
over the 30 years for which data are
available, the trend in annual rainfall has in fact been
increasing (Figure 9.1). Rainfall in the
1980s and 1990s was higher than in the 1970s, and
statistically identical to the early 2000s
(Figure 9.2a). Note also that the bar representing the 2000s is
an average for only three years;
this average could rise (or fall) as the decade proceeds. Total
rainfall has exceeded 300mm only
Figure 9.1 Total annual rainfall, Gode, 19572002 (mm)61
500
400
300
200
100
0
1 9 75 19 8 0 1 98 5 1 9 90 19 9 5 2 00 0
Figure 9.2. Long-run rainfall trends, Gode (mm)

80
2 50
1 97 2-19 86
2 00
60
1 987 -20 02
1 50
40
1 00
20
50
0
19 70 s 1 98 0s 19 90 s (2 00 0s ) 0
J an Fe b Ma r Ap r
Ma y J un J ul A ug S
ep O c t Nov D e c
(a) Average annual rainfall by decade (mm)

monthly rainfall (mm)
(b) Average
61 All the rainfall data analysed in this section were provided by

Ethiopias National Meteorological Services
Agency.
99
Figure 9.3 Total annual rainfall, Jigjiga, 19522002 (mm)

20 0 0
16 0 0
12 0 0
800
400
0
1 9 5 5 1 96 0 1 9 65 1 9 7 0 19 7 5 1 9 80 1 9 8 5 19 9 0 1 99 5 2 0 00
Figure 9.4 Long-run rainfall trends, Jigjiga (mm)

1000
140
19 52-1 97 6
900
120
800
100
700
600
80
500
60
400
300
40
200
20
19 76- 200 2
100
0
0
J a n Feb Ma r Apr
Ma y J un Ju l A ug
Se p Oc t N ov D ec
1 9 5 0 s 1 9 60 s 1 9 7 0 s 1 9 80 s 19 9 0 s ( 2 0 00 s )
(a) Average annual rainfall by decade (mm)

monthly rainfall (mm)
(b) Average
three times in 30 years, so by this criterion pastoralism has

been unviable in central Somali
Region for a very long time. Nor was the drought of 2000
exceptionally severe in five years
since 1972 (1979, 1980, 1984, 1992 and 1993) Gode received
less than 100mm of rain.
Analysis of monthly rainfall trends for Gode confirms the
typical lowland Horn of Africa
pattern of two dry seasons (Jilaal and Hagaa), in the first and
third quarters of the calendar
year, and two rainy seasons (Gu and Deyr), in the second
and fourth quarters. Dividing the 30
years of rainfall data into two halves, it is interesting to note
that the average volume of Gu
rains has remained more or less consistent over the years
since 1972, but that the Deyr rains

have generally been higher since 1986 than before (Figure
9.2b). It is this increase in Deyr
rainfall that accounts for the trend increase in annual
precipitation over the last three decades.
Rainfall data for Jigjiga, in the northern part of Somali
Region, are very different to those
for Gode in the centre. For one thing, there is some evidence
of the third rainy season
(Karan) in mid-year that is typical of neighbouring highland
regions (where it is called Kremt).
This means that rain falls almost continuously between April
and September (Figure 9.2b),
which allows long-maturing rainfed crops to be cultivated,
unlike other parts of Somali
Region where only riverine or irrigated farming is possible.
Second, annual rainfall is considerably higher, averaging
630mm per annum for the period
19722002 (more than double FEWS NETs pastoralist
viability threshold) as against
190mm for Gode over the same period. Third, while a slight
upward trend in annual rainfall
was observed for Gode, in Jigjiga the trend since the 1970s
has been downwards. From the
mid-1960s to the mid-1970s, rainfall in Jigjiga was almost
always in excess of 800mm,
exceeding 1,000mm six times and peaking at 1,825mm in
1976. Strikingly, however, since
1976 total rainfall has never once reached 800mm again
(Figure 9.3).
This might seem to provide persuasive evidence for a longterm, irreversible decline in
average rainfall in Jigjiga. On the other hand, 50 years of
rainfall data are available for
100
IDS
RESEA
RCH
REPOR
T 57
Figure 9.5 Sources of water in birka-dependent areas of

Somali Region
Rainy season
Dry season
Source: Household survey (n=232).

Jigjiga, and this reveals that the decade before the mid-1960s
was comparable to the 1980s
and 1990s, with an annual average just above 500mm and no
single year achieving 800mm.
Even a period of 50 years is too short to provide definitive
evidence of long-term trends in
rainfall, but it is enough to suggest that a long cycle may be
in operation, rather than an
irreversible decline. Certainly, both sets of data (for Gode and
Jigjiga) throw doubts on the
strong claims made by FEWS NET (based on just 10 years of
rainfall figures) that
pastoralism is becoming unsustainable in the Greater Horn of
Africa because of rainfall that
is persistently below minimum thresholds of viability.
9.3 Access to water

Access to water is always a major livelihood concern in
lowland areas, for human
consumption (drinking, cooking, washing), livestock

consumption (watering animals), and
crop production. This section examines the various sources of
water available to the
residents of Somali Region, and the cost of water when it
needs to be purchased.
Perhaps because rainfall in Somali Region is very low
(ranging, as we have seen, from over
600mm in the north to under 300mm in the south) and
variable, several alternative sources
of water are available, including rivers, ponds, boreholes,
shallow and deep wells, rainwater
harvesting, berkad communal and private taps, and water
tankers. Different parts of the
region can be characterised as having distinct water
systems, which differ for humans and
animals, and also alter significantly between the rainy and
dry seasons. These water systems
are best defined by how people access water during the dry
season, when water is scarce.
On this basis, four distinct water systems can be identified
among the 11 districts where
fieldwork was conducted: birka-dependent, well-dependent,
river-dependent, and piped
water (taps or boreholes).
9.3.1 Birka-dependent communities

In the most arid parts of Somali Region, notably in Warder
Zone to the east, people and
animals are almost totally dependent on constructed berkad
for their water in the dry
season. In our survey, most households in Gashamo and
Kebribayah, and one-third of
households in Shilabo, are birka-dependent. This dependence
is especially acute during the
dry season. During the rainy season, more diversified sources
of water are available,
including ponds and shallow wells that accumulate rainwater,
but dry up when the rains
end. These water sources are preferred not only for their
convenience, but also because
they are free, whereas access to berkad costs money.
Figure 9.5 illustrates access to water for people in Gashamo
District.62 During the rainy
season, people and livestock share several sources of water,

dominated by standing water in
62 Since the sources of water in each water system are almost identical
for people and livestock, only human
sources are shown here.
101
ponds or lakes, followed by shallow wells. In the dry season,

however, ponds, lakes and
wells dry up and water consumption switches almost entirely
to berkad, for both humans
and livestock. This has three important implications. First,
since standing water is dangerous
to human health unless boiled, and since people share the
same water sources as their
animals all year round, it is likely that shifting from ponds
and wells to berkad represents no
health improvement in terms of access to clean water.
Second, since berkad are often
privately owned while ponds and lakes are open access, water
for both humans and animals
often has to be paid for during the dry season, and this
introduces economic rationing of
water, especially for the poor. Third, control over water
sources also allows berkad owners
to exclude people from access to water and thus surrounding
grazing resources, which can
restrict mobility of animals to migration routes that have
friendly berkad owners.
Alternatively, restricted access to berkad can provoke
confrontation this is in fact a
common cause of conflict in Somali Region.
Pastoralists and agro-pastoralists are more likely to pay for
water than are farmers. In the
arid districts of eastern Somali Region, especially during the
dry season, there is often no
alternative to berkad, and most berkad are fenced off and
privately owned. Berkad are
invariably owned by wealthier people, who earn substantial
amounts of income from selling
water for domestic use and livestock. Conversely, poorer
people spend significant amounts
on buying water. Almost all purchased water is obtained from
berkad, with access and price
being negotiated individually in each case. Prices paid for
water vary seasonally, rising quite
dramatically during the hot dry Jilaal season. There is no
definite amount charged. It
depends on how severe the situation is.6 3
In some places people pay for their water all year round, but
even in these cases prices paid
in the dry season are higher than in the rainy season. One
household in Kebribayah that
collected its domestic water from a fenced pond in the rainy
season of 2004 paid the
owner 10 Birr for each 200-litre barrel. In the dry season
the pond dried up and the

household switched to buying water from a birka, but at the
considerably higher price of 40
Birr per barrel. Also in Kebribayah, several households that
purchased water from a birka all
year round paid 5 Birr per barrel in the rainy season, but 20
Birr per barrel in the dry
season. This household also reported that they and several
neighbours make advance
payments to a local birka owner, to secure their access rights
to water during the dry
season. During the Jilaal there is a shortage of water, so we
will make an advance payment.
The maximum is 100 Birr and the minimum is 10 Birr to reserve that water.6 4
The commercialisation of water in pastoralist areas

introduces a new source of vulnerability
the rationing of water by ability to pay which is
compounded in drought years, when
water is scarce and prices rise to reflect constrained supplies.
In Gashamo, the price paid for
a 200-litre barrel (oil-drum) of water is typically in the range
of 7,000 to 10,000 Somaliland
Shillings (10 to 50 Birr) during the rainy months, and 20,000
to 40,000 Somaliland
Shillings (30 to 60 Birr) in the dry months. However, several
pastoralists interviewed in late
2004 reported paying as much as 100,000 Somaliland
Shillings per barrel (150 Birr) an
exceptionally high price explaining that the drought had
pushed prices up to
unprecedented levels. During the peak of the drought, water
tankers were sent by the
regional government to deliver water to the worst affected
communities, charging a more
affordable 20 Birr for a 200-litre barrel.
9.3.2 Well-dependent communities

In many pastoralist and agro-pastoralist communities where
adequate groundwater is
available, the main source of water all year round is hand-dug
wells. In our household
survey, shallow or deep wells are the dominant source of
water for both livestock and
people, during both the rainy and dry seasons, in Doboweyn,
Shilabo and Shinile Districts.
During the rainy season, wells are complemented by standing
water in ponds or lakes. In the
63 Male pastoralist, rural Gashamo.
64 Agropastoralist, rural Kebribayah.
102
IDS
RESEA
RCH
REPOR
T 57
Figure 9.6 Sources of water in well-dependent areas of
Somali Region
Rainy season
Dry season

Figure 9.7 Sources of water in river-dependent areas of
Somali Region
Rainy season
Dry season

dry season, dependence on wells intensifies, with a few
people accessing water from
berkad, which may even be located in other districts, as
pastoralists move in search of water
for themselves and their animals (Figure 9.6).
A case in point is Lasoole, in Shilabo District. There are
some 25 wells in and around
Lasoole. These wells belong mainly to clans, although some
are owned by individuals. Access
to water from these wells is free and unrestricted for the clans
living in the area, for their
animals, and for those who migrate to Lasoole in search of
water provided that water is
plentiful. It is common practice among Somalis to share
water there is a strong reciprocal
arrangement among and between us.6 5 Selling water from wells is
seen as shameful; even
people with private wells do not deny others from using

them.
The use of wells does sometimes result in conflict, however.
One source of tension arises
from individuals wanting their animals to use the wells first,
as less labour is involved in
drawing water when the water level is high. Preferential
access is given to the well owner,
his family (we found no cases of women owning wells) and
clan members, with strangers
coming last in the queue. A second source of conflict occurs
when someone digs a well too
close to another, as this can lead to a lowering of the water
level of the adjacent well. To
avoid this possibility and ensure properly planned wells,
people are required to consult their
community elders before digging a new well. When
confrontations over water do occur,
the problem is referred to the elders for arbitration.
9.3.3 River-dependent communities

There are only two permanent rivers in Somali Region. The
Wabe Shabelle river has its
source in Oromiya Region to the west, bypasses Gode town
and flows into Somalia. Inside
Somali Region, it provides a natural boundary between Gode
and Afder administrative
zones. Further south, the Dawa/Ganale/Web river system
forms a natural boundary
65 Key informant, Shilabo District.
103
Figure 9.8 Sources of water in urban areas of Somali Region
Rainy
Dry season
season

between Afder and Liban Zones. Most of the communities
living along the riverbanks or
within walking distance depend on the river for their
domestic water needs, and also water
their animals at the river. This is especially the case during
the dry season, when other
sources of surface water (ponds, lakes) and shallow wells dry
up, and the river becomes the
only freely accessible water source.
This pattern of water use is evident from the three
communities in our household survey
that live in the vicinity of either the Shabelle or the
Dawa/Ganale watershed namely, the
residents of Cherati, Dolo Odo and Kelafo (Figure 9.7).
Because they live in arable rather than arid areas of Somali
Region, these farmers generally
enjoy privileged access to rivers for domestic water that are
free of charge unlike
pastoralists, they rarely have to pay for water for drinking,
cooking and washing. On the
other hand, being sedentary and dependent on water for their
livelihood introduces
different forms of vulnerability. As described earlier in this
report (see Chapter 3), riverine
farming in Kelafo District has been threatened by the

construction of a large dam near
Gode town in the late 1990s, which has diverted water from
the Shabelle river,
permanently reducing downstream flow and seasonal
floodwaters. Farmers in Kelafo argue
they are now forced to pay wealthy people with mechanical
pumps to pump river water to
their fields. We are at the mercy of owners of pump owners
who charge exorbitant prices
for irrigating our farms. They are charging us up to a third of
the grain we produce, or 10
Birr per hour.6 6
Other farmers pay pump owners to lift water from the river to
fields located on higher
ground. The traditional crop irrigation system can only
channel upstream river water along
furrows to fields located downstream, but mechanical pumps
extend the range of irrigation
by some distance. People who live further from the river also
have to pay people who own
donkey-carts to collect water for them. Another risk of
depending on river water for
human consumption is its impacts on health. River water in
Somali Region is not treated
and people rarely boil it before drinking it.
9.3.4 Piped-water-dependent communities

Within Somali Region, piped water is provided almost
exclusively in the major urban
settlements of Jigjiga and Gode town. There are very few
boreholes, and virtually no taps,
in pastoralist areas and farming communities. There are many
advantages to drinking tap
water rather than untreated surface water from ponds, handdug wells or rivers. One of the
main benefits is to human health. The incidence of waterborne diseases is much lower in
towns with a clean piped water supply or at least in those
parts of towns where taps and
boreholes are provided and are free or affordable. In informal
settlements on the outskirts
of towns like Gode, people displaced and made destitute by
the drought of 1999/2000
were extremely vulnerable to disease because of their
overcrowded and unsanitary living
66 Community discussion, Kelafo District.
104
IDS
RESEA
RCH
REPOR
T 57
conditions, and thousands of people died from preventable

diseases, many of which were
water-borne (IDS 2002).
Piped tap water for domestic use is not free, but the cost is
predictable and does not vary
from season to season. As we have seen, urban residents have
higher average incomes.
Moreover, they own few livestock, if any, and are not
farming, so they have no need to
purchase water for animals or crops. People in Jigjiga town
with piped water in their
homes pay the municipality for tap water, usually around 30
40 Birr per month.
A third important advantage of piped water is its reliability
over time, especially when
compared to sources of water that derive, directly or
indirectly, from rainfall. In a region
where variability is the norm in so many crucial respects,
vulnerability is reduced when the
risk of variability is minimised. People living in Jigjiga and
Gode towns enjoy the most
consistent access to water of any group in Somali Region. As
seen in Figure 9.8, there is
virtually no seasonal variation in urban sources of water a
marked contrast to all other
water systems in the region.
10 Conflict
Conflict and insecurity are common in pastoralist societies,
for reasons discussed below, and
Somali Region is no exception. Conflict arises, however,
from a variety of sources, and it is
important not to stereotype pastoralist areas in general as
violent and unstable, but instead
to disaggregate conflict in each specific context. This chapter
reviews the multiple causes of
conflict affecting Somali Region, and the various impacts of
violence and insecurity on local
livelihoods.
10.1 The complex nature of conflict in Somali

Region
Conflict and civil insecurity in Somali Region take several
distinct forms, including war (the
legacy of the EthiopiaSomalia war and Somalias civil war);
rebel or militia activity (conflicts

between ONLF and government forces); inter-clan resource
conflicts (over access to water,
grazing and farmland); and inter-regional border disputes
(with neighbouring Oromiya and
Afar). Civil insecurity is fuelled by the proliferation of small
arms throughout the region.
Some conflicts, especially those that occur between or within
clans, are mediated by
traditional conflict resolution mechanisms and compensation
payment procedures. A final
manifestation of physical insecurity is invisible or latent
conflict (disputes between groups
that do not necessarily erupt into episodes of violence), and is
best expressed by the Somali
word aabsi, meaning fear of confrontation.
The war between Ethiopia and Somalia in 19778
(commonly referred to in the region as 77) caused massive population displacement, with up to one
million Ethiopian Somalis crossing
the border into Somalia, where most settled for the next
decade. In 1988, civil war erupted
in Somalia (this is commonly referred to as 8-8), and
approximately half a million refugees
from Somalia crossed into Ethiopia, where they were
accommodated in refugee camps
(Aware camps) run by international humanitarian agencies.
After the fall of Siyad Barre and
the collapse of the Somalia state in 1991, most of the one
million Ethiopian Somali refugees
(from the 19778 war) returned from Somalia and were
assimilated into Somali Region.
Some joined refugees from Somalias civil war, including a
new wave from southern Somalia
in 1991, in the refugee camps, but the majority rebuilt their
livelihoods remarkably quickly,
even though many arrived with few assets. We were a big
family in Somalia. We had a good
standard of living. Then, the [civil] war came, and we were
made refugees. Before we came
here, soldiers destroyed our houses. We came here with empty hands.6 7
A more recent source of internal displacement within

Ethiopia was a referendum in
October 2004 over the border between Somali and Oromiya
Regions, which caused tens
105
of thousands of Oromos and Somalis to move or be driven

out of their communities. Apart
from some immediate emergency assistance, little support
has been provided to these
people, many of whom have lost their homes, land, livestock
and other property, and are
now living in informal IDP camps in both regions.
International geopolitics have always impinged directly and
significantly on livelihoods in the
Horn of Africa, even in rural Somali Region. In August 2003,
the government of Djibouti
expelled over 100,000 foreign residents (labelled as illegal
immigrants), allegedly at the
instigation of the USA as part of its War on Terror. One
middle-aged man interviewed in
Kebribayah was directly affected by this decree. He was
working as a security guard in
Djibouti but was abruptly deported and forced to return to his
village in Kebribayah and take
up farming. Since 9/11, there has been a permanent and
highly visible US military presence
in Somali Region, with two complementary objectives:
winning hearts and minds by
building schools and digging wells for local communities,
and monitoring radical Islamic
movements (They are hunting Al Qaeda).
Another source of conflict and insecurity with an overtly
political dimension is the longrunning campaign by the ONLF for greater autonomy for the
Ogaden. There are regular
skirmishes between ONLF militia and the state security
services, with deaths on both sides.
In some ONLF-controlled areas, communities have tried to
expel ONLF fighters,
complaining that their presence is preventing the delivery of
public services and inhibiting
private investment. In Shilabo town, tensions between the
ONLF and the local community
have resulted in the deaths of several Ethiopian soldiers,
ONLF fighters and residents of the
town (Box 10.1). During our fieldwork in early 2005, these
communities felt strongly that
the government should engage the ONLF in negotiations to
resolve their differences, as
the impasse is affecting their livelihoods and the general
development of Somali Region.68 In
fact, the Government of Ethiopia took steps towards

negotiating a ceasefire with the
ONLF in late 2005.

Most of the violent and latent conflicts within the region
are not political in the larger
sense, but occur between or within clans and concern access
to productive resources,
especially water and land pasture for grazing, arable land
for farming (Gebre-Mariam
2005). Conflict is often seen as endemic to pastoralist
communities, partly because
productive resources are scarce and because, as a
consequence, people move continually in
search of water and grazing for their livestock. With legally
formalised property rights over
these resources either absent or weakly defined, the
boundaries between clan territories
are often contested and when two groups converge on the
same water point or pasture,
conflict often follows. However, it would not be correct to
conclude that pastoralist areas
like Somali Region are lawless and anarchic. In fact, an
elaborate system of property rights
or resource allocation mechanisms exists, and traditional
conflict resolution institutions and
compensation procedures are widely accepted and effectively
implemented. These
mechanisms and institutions are not sufficient, however, to
prevent all conflicts from
escalating into violence.
Disputed territory (e.g. a water point, pasture or farmland)
often becomes inaccessible to
both parties involved in the dispute until it is resolved
which can take years. One direct
impact is to take these resources out of the production system
for the duration of the
dispute, which is a significant loss in a region where natural
resources are already scarce. An
indirect impact is to disrupt mobility one of the most
fundamental characteristics of
pastoralist livelihoods so that people, livestock and
commodities cannot move through
these areas because of aabsi, or fear of confrontation.
Pastoralists have to use different
pasture or water points elsewhere, new migration routes must
be established, or traders
might need to find a circuitous route to avoid the disputed
territory. All of these
68 One minor but significant consequence of ONLF activity is that
fieldwork for our survey had to avoid parts of
Somali Region that are known to be highly militarised and
potentially dangerous. Even so, the fieldwork
teams were stopped and questioned by armed militia and Ethiopian
soldiers on several occasions. In such an
environment, delivery of normal government services becomes
impossible. Even NGOs with negotiated

access face security risks and occasional loss of life.
106
IDS
RESEA
RCH
REPOR
T 57
Box 10.1 Conflict in Shilabo District, 19982005

Balisha 2005: Two Ogadeni sub-clans claimed
control of Balisha kebele and
ownership of Balishas wells. One sub-clan set fire to a
farm belonging to the other,
and physically assaulted two of their men. At the time
of our fieldwork, the situation
remained tense, with rumours that both sides were
arming themselves. Clan and
religious leaders were trying to intervene to prevent
lives being lost.
Bi'idyahanMakahil conflict, 2003: Two clans went
to war over control of the
Ogadeni village of Gohweyne. The Ismail, a sub-subclan of the Ogadeni Makahil
clan, had lived in Gohweyne for 100 years, while the
Bi'idyahan, a subsection of the
Majartein clan, claimed Gohweyne as their ancestral
land, and tried to construct
berkad in the area. Both parties obtained arms from
relatives in Somalia and waged
war on each other. The Makahil had 128 men killed
and 173 injured, while the
Bi'idyahan lost 152 lives and had 190 injured. The
Bi'idyahan won the war and
displaced the Makahil, who fled to Warder District.
Labobari 2001: Conflict erupted between three clan
groups over control for Labobari
kebele and its water wells and pastureland. One of the
three groups procured arms
from Somalia and defeated the other two, displacing
them out of the kebele. Many
lives were lost and many were left destitute. This
dispute was resolved by armed
conflict with no intervention from other clans or the
Government of Ethiopian,
either to provoke the violence or to prevent it.
Godle 2001: The people of Godle felt that the
presence of ONLF fighters in their
community was undermining their relationship with
the government, and decided to
disown the ONLF. When the ONLF refused to leave
the area, the community killed
three ONLF fighters.

Qamuda 2000: There was a local skirmish between
the state and the ONLF when
three young men who were accused of working for the
ONLF were killed.
Shilabo 1999: Two Ethiopian government soldiers
were killed by the ONLF in
Shilabo town, and the government retaliated by
gunning down four people from a
local Ogaden sub-clan.
Jaleelo 1998: A vehicle transporting khat was
attacked by members of a different
clan who intended stealing the khat. In retaliation, the
khat traders clan killed two
relatives of the thieves. This conflict was resolved by
the elders who found the khat
traders clan guilty and fined them blood money.
Source: Key informant interviews, Shilabo District.
adaptations impose additional transactions costs on economic
activities. If it takes traders
and middlemen longer to bring commodities to local
communities and to take livestock to
the market, for instance, their operating costs will be higher.
For example, fighting between
the Bi'idyahan and Makahil in 2003 (Box 10.1) not only cost
an estimated 280 lives, it was
also very damaging to local livelihoods, as the disputed area
which sparked the conflict lies
on the main trading route from Shilabo to Bosaso, so this
route became dangerous and
inaccessible for several months. Some trucks that did attempt
to transport livestock to
Bosaso were attacked and the animals were seized or killed.
Conflict also has the significant indirect impact of
constraining access to productive resources.
In Doboweyn District, a low-level or latent conflict has been
ongoing for four years between
two Ogadeni sub-clans, concerning an area of land that is
sometimes farmed and sometimes
left fallow. In 2001 this area received enough rain for crop
cultivation, and both sub-clans
claimed the right to farm it. An international NGO
(International Committee of Red Cross
and Red Crescent societies) offered to provide farm tools to
local people, which
unintentionally escalated the tension, as each side feared that
the other would seize this
opportunity to settle and farm the land permanently. In order
to avert a conflict both
parties being heavily equipped with small arms the local
government intervened and asked
107
both sub-clans to refrain from farming the land. Despite

being a valuable resource for local
livelihoods, this land has remained unutilised ever since. The
local community is frustrated by
the failure of the government to resolve this stand-off,
complaining that local and regional
officials appear to have no interest in finding a durable
solution.
Conflict can also occur around access to externally provided
resource transfers. A case in
point relates to problems associated with food aid targeting
during the 2004 drought. In
one community in Kebribayah where community targeting
was applied i.e. community
representatives were asked to identify deserving beneficiaries
we were told that one
person was killed. (The conflict arose because of food aid.
When the drought was
unbearable and people were very hungry, they started
fighting each other for the food that
was being brought to the community.)
In parts of Somali Region that are especially insecure,
communities are exposed to conflict
of different kinds. In the case of Shilabo District alone, seven
serious incidents of violent
conflict were reported as having occurred within the past
seven years (Box 10.1). In the same
district, tension is building around control and exploitation of
the Calub natural gas deposits
reputedly one of the largest in the world, but unexploited to
date, due mainly to security
concerns. This is a major potential flash-point for future
conflict, both within the local
communities and between local people, the Ethiopian
government and multinational
businesses. An elder in Shilabo commented that there have
been no discussions with local
communities about the division of future revenue streams
from this resource between local,
national and international interests.
We see foreign companies come and go, and the
government is maintaining a large
army here. No one tells the community what is going
on, let alone consulting them.
The ONLF and others fear that the Federal authorities
want to exploit this resource
for the benefit of non-Somalis, and this is unacceptable.
First of all, we believe this

land is ours. If we are not consulted on how it will be
used, then it better remains
unexploited. However, if we are consulted and given a
fair share, we shall defend this
land even against the Americans!
10.2 Impacts of conflict69
Most households in this survey claimed that their livelihoods

have not been affected by
conflict at all in recent years (n=1,003/1,093 =91.8 per cent),
while a small number said
conflict has affected them but not very seriously (n=32 =2.9
per cent), and one respondent
in 20 said conflict has affected their livelihoods quite
seriously or very seriously (n=58 =5.3
per cent). (Table 10.1). This is a relatively small number but
it is significant, especially when
disaggregated to the district level, given that conflict and
insecurity appear to be
concentrated in certain areas.7 0 Almost one-third of households in
Doboweyn (31 per cent)
reported that their livelihoods have been seriously affected by

conflict, while livelihoods have
been disrupted, either mildly or seriously, in one household in
four in Shilabo (26 per cent)
and one household in five in Kebribayah (19 per cent).
As an indicator of how conflict can impact directly on lives,
respondents were also asked if
any member of their family has been injured or lost their life
due to conflict. A total of 47
conflict-related deaths was recorded, 21 of which occurred in
our 100 Doboweyn
69 The section on conflict in the household questionnaire included the
following questions: Since the fall of
Mengistu in 1991, how has conflict affected your familys
livelihood?; Has any member of your family been
injured, or lost their life, due to conflict?; The last time there was
conflict that affected your family or
interfered with your livelihood, who was involved in the conflict?
70 It is important to note that our sampling frame was restricted to parts
of Somali Region that were
considered relatively stable at the time of the survey (so as not to put
our fieldwork teams at risk), so that Fiq
Zone, for instance, was excluded. This implies that the qualitative
and quantitative findings on conflict, as
reported in this chapter, understate the true extent and impacts of
conflict and insecurity in Somali Region as
a whole.
108
IDS
RESEA
RCH
REPOR
T 57
Table 10.1 Households affected by conflict in Somali Region,

by district
Impact on livelihoods Impact on lives
District None (%) Minor (%) Serious (%) Relative
Relative
i
n
j
u
r
e
d
(
%
)
k
il
le
d
(
%
)
Pastoralist 90 4 6 6 4
Gashamo 96 2 2 0 3
Shinile 100 0 0 0 0
Shilabo 74 9 17 19 9
Agro-pastoral 82 7 11 7 11
Kebribayah 81 18 1 4 9
Doboweyn 68 1 31 13 21
Cherati 96 2 2 5 2
Farmer 98 0 2 1 1
Jigjiga rural 100 0 0 0 0
Kelafo 94 0 6 3 3
Dolo Odo 100 0 0 0 0
Urban 100 0 0 0 0
Jigjiga town 100 0 0 0 0
Gode town 100 0 0 0 0
Average 92 3 5 4 4

households, followed by nine deaths in Shilabo and nine in
Kebribayah. Doboweyn and
Shilabo also reported the highest numbers of family members
injured during conflicts (13
and 19 respectively). Once again, this amounts to only 4 per
cent of households being
affected across the 1,100 households interviewed, but with a
heav y concentration in a few
districts. Across livelihood categories, agro-pastoralist areas
appear to suffer most from
instability and conflict, accounting for two-thirds of conflictrelated deaths. In two of the
three farming districts Dolo Odo and rural Jigjiga by
contrast, no conflict impacts were
recorded at all. This was also the case in the urban centres of
Jigjiga and Gode.
When asked which actors were involved in the most recent
conflict that affected the
respondents, the most common answer was conflict between
our clan and another clan
(n=139/195 =71.3 per cent). After inter-clan conflict, conflict
between militias and the clan
came a distant second (8.7 per cent). All responses that
mentioned militias came from
households in Shilabo District. Intra-clan conflicts (between
sections or sub-clans) ranked
third (5.1 per cent). When asked who benefited most from
that most recent conflict,
politicians and government officials were named by three
respondents in four
(n=152/199=76. 3 per cent), followed by militias (13.6 per
cent), then clan elders (8.5 per cent).
Almost one in four households (n=198/870 =23 per cent)
stated that they had lost access
to farmland that used to belong to them, or grazing land that
they used to use, because of
conflict. This consequence was felt most severely in
Doboweyn (61/100 households), rural
Jigjiga (39), Kebribayah (35) and Shilabo (34). Asked why
they were no longer using land they
had lost, most affected farmers replied that it had been taken
over and was being farmed by
other people now. In a few cases, land that is disputed was
being left unfarmed by either
party, pending a resolution of the dispute. In Kebribayah,
some land that was formerly used
for grazing has been settled and is now being cultivated by
agro-pastoralists. Also in
Kebribayah, several respondents told us that they no longer
use certain migration routes and
pastures because of aabsi the risk of conflict with other
groups that are contesting access

to the same grazing land. We dont go there for fear of being
killed, since the land has
been taken by another clan.
109
Table 10.2 Impacts of conflict in Somali Region, by district

District Loss of Loss of access Loss of Relative Relative
grazing or to water livestock (%)
injured (%) killed (%)
farmland (%) points (%)
Pastoralist 13 7 7 6 4
Gashamo 4 0 2 0 3
Shinile 0 0 0 0 0
Shilabo 34 22 18 19 9
Agro-pastoral 34 29 12 7 11
Kebribayah 35 31 2 4 9
Doboweyn 61 54 31 13 21
Cherati 6 3 4 5 2
Farmer 19 13 1 1 1
Jigjiga rural 39 28 0 0 0
Kelafo 19 12 3 3 3
Dolo Odo 0 0 0 0 0
Urban 0 0 0 0 0
Jigjiga town 0 0 0 0 0
Gode town 0 0 0 0 0
Average 23 15 5 4 4
A similar pattern was recorded for loss of access to water
points, especially berkad and
wells, for livestock or domestic use. The same districts were
affected as those that had lost
access to land, to a slightly lesser extent. Once again, agropastoralists in Doboweyn appear
to have been worst affected. The same reasons were given as
for land: typically the water
point has been taken by a rival clan who denies access to
our survey households, and they
fear that conflict will break out if they try to use that water
point now. In some cases,
berkad and wells that were previously open access have been
fenced off and privatised
either people have to pay to water their livestock there, or
access is restricted for the
exclusive use of the owner and his relatives.
Several households have lost physical property due to
conflict. Eight households in Shilabo

and nine in Kebribayah reported that assets belonging to
them had been stolen by rival
groups during episodes of violent conflict. Two farming
households in Kebribayah and Kelafo
claimed that their granaries had been raided or destroyed
during a conflict, and they had
lost their cereal stocks. Most significantly, though, almost
one-third of households
interviewed in Doboweyn, and one in five in Shilabo, have
lost livestock. Usually the
numbers of animals involved were small (fewer than 10 head,
in 46 out of 59 cases), but in
some households in Doboweyn the respondent reported
losing over 50 head of sheep and
goats. This finding is consistent with evidence from localised
conflicts between pastoralist
groups elsewhere in Africa, where owners of large herds or
flocks are specifically targeted
and their animals seized during raids or attacks on the
community (Deng 2003). In this
sense, conflict can make households with more assets more
vulnerable, not less. This is a
distinctive feature of conflict-related vulnerability, since
livelihood vulnerability is usually
assumed to be inversely related to wealth (i.e. vulnerability
falls as asset ownership increases).
110
IDS
RESEA
RCH
REPOR
T 57
Table 10.3 Institutions for conflict resolution, by district

District Elders (%) Police (%) Gurti (%) District Family
(%) Militia
admin
istrati
on
(%)
(%)
Pastoralist 85 6 6 1 2 0
Gashamo 92 0 0 2 5 1
Shinile 80 0 16 2 2 0
Shilabo 81 17 3 0 0 0
Agro-pastoral 96 0 3 1 0 0
Kebribayah 99 0 0 1 0 0
Doboweyn 92 0 8 0 0 0
Cherati 97 0 1 1 1 0
Farmer 84 3 8 3 1 1
Jigjiga rural 77 2 20 1 0 0
Kelafo 88 2 2 8 0 0
Dolo Odo 96 0 2 0 2 0
Urban 41 53 2 2 1 1
Gode town 43 52 3 1 1 1
Average 79 13 5 2 1 0
10.3 Conflict resolution

identify the primary institutions
involved in resolving disputes that affect themselves or their
community. (Who do you turn
to in cases of conflict or trouble, to solve the problem?)
There was a clear ruralurban
divide in responses to this question. In rural areas, the elders
are overwhelmingly preferred
(over 80 per cent of cases) (Table 10.3). In two districts, a
significant minority turn first to
the Gurti, and in Shilabo which (as we have seen) has been
the location for much conflict

and instability in recent years the police were mentioned by
one in six households. In only
one district, Kelafo, was the District Administration given
credit by more than one or two
households as an institution that can step in to solve
problems.
In both the urban centres of Jigjiga and Gode, by contrast, the
police were named by more
than half the respondents as the primary institution for
conflict resolution, with the elders
relegated to second place. This suggests that a shift is
occurring in urban areas, from
traditional to modern institutions for conflict resolution.
Alternatively, this could be
interpreted as a transition from resolving interpersonal
disputes through the communitys
own internal mechanisms, to delegating that responsibility to
the agents of the state. It is
also plausible that the nature of conflict between people is
rather different in towns, being
dominated by petty crime such as theft and assault. In such
cases, the appropriate authority
to intervene and provide justice is increasingly seen as the
police, rather than the clan. In
rural areas, where conflicts are more likely to arise over
disputed access to water and
grazing, these are issues of inter-clan territoriality and the
elders are still regarded as the
people who should sit together and negotiate a solution.
10.4 Compensation payments

Conflict imposes a variety of costs on affected households,
including death or injury of
family members, loss of livestock and property, and restricted
access to grazing land and
111
Table 10.4 Compensation payments, by district

District Paid Received Paid cash Paid Received Received
livest
ock
cash
livest
ock
Pastoralist 17% 8% 17 26 18 6
Gashamo 2 0 0 0 0 0
Shinile 0 0 0 0 0 0
Shilabo 49 23 17 26 18 6
Agro-pastoral 29% 3% 10 51 0 3
Kebribayah 27 3 6 21 0 0
Doboweyn 58 7 4 30 0 3
Cherati 2 0 0 0 0 0
Farmer 5% 4% 1 13 3 8
Jigjiga rural 0 1 0 1 0 1
Kelafo 14 10 1 12 3 7
Dolo Odo 1 0 0 0 0 0
Urban 0% 0% 1 0 0 0
Gode town 1 0 1 0 0 0
Total 13 % 4 % 29 90 21 17
water points. For those who are deemed to be responsible for
the conflict, a further cost is
the compensation that they are required to pay to the other
party involved.
Qaadhaan is an effective risk-pooling mechanism.
According to Somali customary law
(Xeer), for instance, if a member of one clan or sub-clan kills
a member of another, 100
camels must be paid as compensation if the victim was male
and 50 camels if she was
female. When compensation is due to be paid, the elders
decide on the share to be
contributed by each family in the paying clan or sub-clan,
which divides the cost among
many families. When the compensation is paid, half goes to
the victims household
(analogous to the paying out of life insurance) and the other

half goes to the clan (it is
retained by the elders). In practice, however, compensation is
rarely paid immediately or in
full. Instead, the offending clan makes a nominal payment to
acknowledge their culpability
if they do not make this gesture, tensions can escalate to fullscale conflict and the
balance is a debt that is known to both parties but is paid
slowly, in instalments, or reduced
if the other clan offends against the debtor clan.
A total of 154 households in our survey (14.2 per cent)
reported having paid compensation to
other families or clans because of conflict. Most of these
households live in just two of our
11 districts Doboweyn (58) and Shilabo (49). Significant
numbers of households had also
paid compensation in Kebribayah (27) and Kelafo (14).
Fewer households had received
compensation from others (n=44/1,086 =4.1 per cent).
Interestingly, most of these
households were concentrated in the same communities
although the numbers are less,
almost all the households either paying and/or receiving
compensation live in the same four
districts (Table 10.4).
Typically, compensation paid from one clan or sub-clan to
another is shared among all
members of the paying group (Dia). In the 29 households that
paid compensation in the
form of cash, their share (hagaag) averaged 224 Birr, ranging
from a minimum of 100 Birr
to a maximum of 1,063 Birr. Compensation in the form of
livestock was paid mainly in
cattle (49 households, usually one cow each, but eight cows
in one case from Kelafo). In 26
cases (21 from Shilabo), compensation was paid in camels
(onethree camels per household).
112
IDS
RESEA
RCH
REPOR
T 57
Table 10.5 Relationships with neighbouring groups, by

district
District Good (%) Stable (%) Tense (%) Conflict (%)
Stability index
Pastoralist 84 9 6 1 0.08
Gashamo 91 9 0 0 0.03
Shinile 99 0 0 1 0.01
Shilabo 60 20 18 2 0.21
Agro-pastoral 69 11 17 3 0.18
Kebribayah 75 5 16 4 0.16
Doboweyn 36 24 34 6 0.37
Cherati 96 2 2 0 0.02
Farmer 93 3 3 1 0.03
Jigjiga rural 91 7 1 0 0.03
Kelafo 90 1 7 2 0.07
Dolo Odo 99 1 0 0 0.00
Urban 98 1 0 1 0.02
Jigjiga town 94 3 0 3 0.04
Gode town 100 0 0 0 0.00
Average 85 7 7 2 0.09
In 15 cases, sheep and goats were paid (onefive per
household). Compensation received by
households surveyed was also in the form of cash and/or
livestock, and to approximately the
same value.
Several cases of compensation payment were recorded in our
fieldwork. In one case, a man
from Kebribayah bought khat from another, but there was a
disagreement over the price to
be paid, and one man stabbed and killed the other. Blood
money was asked for, even
though the two men were related to each other. One
household we interviewed from the
affected community did not know the total compensation that
was due to be paid, but he
had personally contributed one cow towards the blood
money.
10.5 Vulnerability to conflict

As a proxy indicator of how vulnerable communities might
be to conflict breaking out in
the near future, the household survey asked respondents to
describe their relationship with
neighbouring groups at the time of the interview. Responses
were clustered into four
categories: good, stable, tense, or conflict.
Encouragingly, relationships were described
as good by over 90 per cent of respondents in 8 out of 11
districts. In Doboweyn, Shilabo
and Kebribayah, however, relations with neighbours were
described as tense by significant
numbers of respondents, and in a few cases even as open conflict
(Table 10.5).71 The
situation in Doboweyn appears to be most tense and should

give cause for concern. As was
seen above, there is a great deal of latent and overt conflict in
Shilabo District. As will be
seen below, Shilabo and Doboweyn are also the two districts
where local people are most
disillusioned with government.72
71 This simple question (How would you describe your relationship with
neighbouring groups at this time: good,
stable, tense, or conflict?) generates useful information and could be
incorporated into a rapid conflict early
warning system.
72 Again, note that highly insecure parts of Somali Region were not
included in our sampling frame, so the true
extent of instability and conflict in the region is under-represented in
this analysis.
113
A simple stability index was constructed that summarises the

responses for each district into
a single score between 0 (complete stability all
relationships good) and 1 (complete
instability all responses conflict). Where the index value
is 0, this district can be described
as stable, with no evidence of either overt or latent
instability. This is the case for only two
sub-samples in our survey Gode town, and Dolo Odo, a
riverine farming community in
southern Somali Region. In six district sub-samples, the
stability index value lies between 0
and 0.1, suggesting a limited degree of instability that is
mostly latent at present, but could
deteriorate into open conflict in the future. Districts with
index values above 0.1 are
Doboweyn (0.37), Shilabo (0.21) and Kebribayah (0.16)
(Table 10.5). In these places, conflict is
either overt or is simmering and could erupt at any time. Note
that this is a snapshot of the
situation at a particular moment. Given the dynamic and
unpredictable nature of conflict in
this region, the situation could (and does) change from one
week to the next, as tension
builds, explodes or subsides in different communities over
time.
Generalising across livelihood categories, agro-pastoralist
communities emerge as apparently
most unstable or vulnerable to conflict in terms of our index,
followed by pastoralist areas.
Sedentary farmers and urban residents are least affected by
conflict, or the threat of
conflict, in our sample. This finding could have been
predicted, given that the source of
much conflict in arid and semi-arid lowlands across the world
is competition over scarce
natural resources among people who are either moving with
their livestock in search of
water and grazing, or claiming a plot to farm where arable
land is scarce, and where no
private ownership or use rights have been established over
these common property
resources. However, it is important to avoid drawing an
overly deterministic association
between livelihood systems and susceptibility to conflict.
There is great variability across
communities pursuing similar livelihoods the stability
index reveals that one of the three
agro-pastoralist districts, and two of the three pastoralist

districts, are extremely peaceful.
Many complex and context-specific factors environmental,
political, socio-cultural affect
whether any particular community will be peaceful or
conflict-prone at any point in time.
Note also that evidence of conflict is not an argument for
privatising rights to land, water
and other natural resources in Somali Region. Negotiated
access property regimes are a
rational response to the harsh and unpredictable environment
in which pastoralists and
agro-pastoralists make their living, and mobility is an
essential component of their livelihood
strategies. Any intervention that restricts the physical
mobility of people and animals, such
as fencing farmland or pastures, or individualising ownership
of water points, reduces
freedom of movement and raises the livelihood risks that
pastoralists face.
11 Governance
political
representation
and
This chapter considers the roles and effectiveness of key

governance structures and
institutions in Somali Region, including the regional
government and the regional Gurti, an
adaptation of a Somali institution that facilitates
communication and negotiation between
representatives of clans. Selected policy processes, notably
decentralisation and the 2005
national and regional elections, are also briefly discussed. We
then present evidence from
our household survey on perceptions of the fairness and
effectiveness of government
representation, at local, regional and national levels.
11.1 Political structures and policy processes

The Somali National Regional State (SNRS) is a product of
the Ethiopian Peoples
Revolutionary Democratic Front (EPRDF) governments
policy of ethnic federalism, which
assigned the major ethnic groups in Ethiopia to nine regional
administrations corresponding
to major language groups. The intention was to eradicate the
domination of one ethnic
group over all others, though critics have argued that ethnic
divisions and polarisation may
have increased as a result of the hardening of ethnic

identities.
114
IDS
RESEA
RCH
REPOR
T 57
The relationship between the Ethiopian state and traditional

political structures in Somali
Region is delicately balanced, and constantly evolving.
Rather than being two distinct
systems, either performing complementary functions or
competing against each other for
authority over citizens, representatives of the state and the
clans intersect and interrelate in
complex ways. While the state holds nominal power and
controls most of the public
resources disbursed at regional level, respondents describe
how the elders and other
traditional leaders also continue to exert considerable power.
At regional level, elders and
elites work in concert with the state, and at local level, where
government administrators
are thin on the ground, most decisions are made by traditional
leaders. These leaders are
often employed by and also advise the state, while many
bureaucrats and politicians in the
regional government are also pastoralists or traders, with
strong affiliations to particular
clans and rural communities. Tensions created by these
multiple roles infuse political life in
Somali Region. Members of government point to the
energetic efforts towards reform and
improvement of political and administrative systems.
The main political and administrative representative of the
federal state in Somali Region is
the regional government, based in Jigjiga town, which
includes the Regional Parliament
and the Regional Administration. At the time of the survey,
respondents, particularly in rural
areas, claimed that the regional government had not yet
managed to deliver much in the
way of development to Somali Region. Most respondents
said that there had been no
discernible improvement in the provision of social services or
physical infrastructure. In
Jigjiga District, elders noted that there were some
infrastructural improvements.
Respondents were concerned that government interventions
to prevent conflicts or
mediate in long-running disputes usually had only short-lived
success, and there was
widespread criticism of the governments performance on

livelihood-related issues (such as
the Saudi livestock import ban and the war on contraband).
People generally perceive the
regional government as being embroiled in politics rather
than serving the people, who
are rarely consulted about decisions taken on their behalf.
Elders see regional bureaucrats as
being mainly young and inexperienced, describing them as
small boys or black hairs
(rather than white hairs like themselves).
The regional budget is always fiercely contested and often
underspent, and in some years
funds allocated to the region have been returned to Addis
Ababa. Observers of the political
scene in Somali Region note that the Regional President and
most members of the
Regional Cabinet rarely stay in their posts for too long before
being removed and
replaced with other officials who, in turn, are replaced one or
two years later. This high and
repetitive turnover of high-level politicians and bureaucrats is
both a cause and a
consequence of failures of accountability at the regional
level. A unique form of neopatrimonialism7 3 dominates politics in Somali Region. Rather
than representing the interests
of Somali Region as a whole, people who are elected or

appointed to positions of power
tend to exercise that power to benefit the interests of the
group that installed them in
their post. Clans with members in positions of power and
influence expect those individuals
to direct resources (public spending, food aid, jobs and
contracts) towards the interests of
the clan, rather than to the region as a whole. Knowing that
their time in power might be
short-lived, these individuals necessarily have to have a short
time horizon and a narrow
vision of what they hope to achieve. (Hagmann 2005).
Whether they are personally corrupt
or not is beside the point the consequence is a distorted
policy process. Political processes
that lack effective means of regulating competition for scarce
resources and of keeping the
civil service separate from patronage are unlikely to achieve
broader developmental
objectives (such as the provision of effective public services)
and sustainable development
outcomes. Even if power is (unofficially) rotated between the
dominant clans, the interests
of marginalised clans and powerless groups (e.g. the Bantu
Somalis of Kelafo) are not
adequately represented by this political arrangement. This

situation is widely recognised
73 Neo-patrimonialism is a term that describes political regimes in
which the collective interests of society as
a whole are subordinated to the much narrower interests of people in
power and their patrons and clients.
(See Chabal and Daloz (1999) and van de Walle (2001), for detailed
analyses of neo-patrimonialism and
clientelism in various African states).
115
within the regional government and the government is

discussing reforms that have the
potential to address these fundamental issues.
One institution that should assist marginalised clans to
overcome their exclusion is the Gurti.
A Gurti is a traditional Somali institution that brings together
representatives of different
clans (or sub-clans within a single clan) as a forum for
discussion and negotiation over issues
of mutual concern, such as establishing territorial boundaries
or mediating in conflicts. In
1999 the regional government created a regional Gurti,
comprising regional and zonal elders
elected by 700 clan elders and religious leaders (sultans and
Ugaz) and representing all major
clans and sub-clans throughout Somali Region. Its members
are paid a salary by the regional
government. This high-level Gurti serves as a formal council
of traditional leaders, with
whom the regional president and government can consult on
major issues and decisions.
As a hybrid adaptation of a traditional institution, the
regional Gurti occupies an ambiguous
political position in Somali Region. It was created with a
view to incorporating the traditional
governance system into the modern one, but the process has
not yet reached a stage where
the institutions responsibilities and powers have been
clarified and codified. Gurti members say
they are only occasionally called upon, usually to provide
short-term conflict resolution services,
but have few formal powers in government. Some feel
frustrated that the Gurti is not given
more opportunity to provide more leadership in the region.
People outside Jigjiga town tend
to see the Gurti members as losing authority in the
community and being compromised by
their close association with the difficulties of Jigjiga politics
and administration.
A new element in the relationship between communities,
their elected representatives and
the state, is the ongoing policy of decentralisation.
Government decentralisation in Ethiopia
has been termed deconcentration by the World Bank (2001)
the extension of state
power from the centre to the periphery, to provide a more
responsive and locally
accountable administration. If and when local councillors and
administrators succeed in
mobilising resources to build schools and clinics, and to meet
other locally articulated
priorities, then decentralisation will benefit the people of
Somali Region significantly. Survey
respondents were concerned, however, that newly
empowered local politicians might
disrupt indigenous mechanisms of conflict resolution,
allocation of rights to land and water,
or community-based natural resource management.
Federal and regional elections were held in Ethiopia on 15
May 2005, except in Somali
Region where they took place in August. European Union
monitors described the May
elections as more transparent and efficient than ever before,
but they raised concerns
about the overall political environment in which the
elections took place (European Union
2005). Opposition parties made large gains. In some senses
the setback that the EPRDF
received in the May elections gave the government an
unprecedented sense of
accountability towards its citizens, and extensive talks were
held between government and
traditional leaders in Somali Region in the weeks leading up
to the election. On the day
before voter registration was due to begin, however, hand
grenades were thrown into
hotels and official residences in Jigjiga, Fiq and Degahbur,
killing at least five people. The
reasons for this action were unclear, but some blamed the
ONLF, which denied the
accusation. The elections which followed on 21 August were
criticised by EU election
observers, who alleged that ballots were being sold on the
black market and that
opposition activists were intimidated and arrested. Since the
elections there have been
further efforts by government to publicise its prioritising of
pastoralist issues.
11.2 Perceptions of political representation

comment on the fairness of their
political representation and the effectiveness of their political
representatives.74 The
percentages of respondents who answered in the affirmative

that they believe they are
74 Do you feel that you are fairly and adequately represented in
local/regional/federal government? and Are
your representatives effective in representing you at
local/regional/federal government?
116
IDS
RESEA
RCH
REPOR
T 57
Table 11.1 Perceptions of fairness and effectiveness of

government representation
Fairly
represented
Effectively
represented
District Local Regional Federal Local Regional Federal
Average
(%) (%) (%) (%) (%) (%) (%)
Pastoralist 30 17 23 17 9 14 18.4
Gashamo 62 32 58 19 6 19 32.6
Shinile 40 24 24 34 25 26 28.8
Shilabo 2 0 0 2 0 0 0.7
Agro-pastoral 20 6 5 18 4 4 9.5
Kebribayah 22 12 13 18 9 9 13.6
Doboweyn 4 1 0 4 0 0 1.6
Cherati 42 3 2 39 2 1 15.0
Farmer 27 10 9 26 9 9 14.9
Jigjiga rural 32 27 26 29 23 23 26.5
Kelafo 22 5 4 22 4 4 10.0
Dolo Odo 27 0 0 26 0 0 8.9
Urban 36 21 18 35 22 16 24.6
Jigjiga town 19 19 24 19 19 21 20.3
Gode town 59 23 12 56 24 12 30.9
Average 27 12 13 23 9 10 15.8
fairly represented and adequately represented in government
structures, at the local,
regional and federal levels are summarised in Table 11.1,
by district. The general impression
that this table conveys is that representation is inadequate at
all levels. Across the six
questions asked, the range of positive responses averages in
the range of 9 per cent (the
lowest result, for effective representation at the regional
level) to 27 per cent (the highest
response, for fair representation at the local level).
These are discouraging findings for government. They
suggest that three-quarters of people
in Somali Region feel disenfranchised (not fairly represented)
and believe that government is

not working effectively to serve their interests (ineffectively
represented). There is better
news for the process of decentralisation: for both fair and
effective representation,
positive responses for the local level were more than double
those for the regional and
federal levels. Significantly, although the regional and federal
administrations were seen as
almost equally ineffectual, the regional administration in
Jigjiga received the worst approval
ratings of all three levels of government.
It is important to note that there is substantial variability in
perceptions across districts, with
positive approval ratings (>50 per cent) recorded in a few
cases for local government in Gode
town (both fairness and effectiveness), and for fairness of
representation in rural Gashamo (at
local and federal levels, but not regional). At the other
extreme, 100 per cent disapproval ratings
were recorded for both fairness and effectiveness of
representation at regional and federal
levels, in three districts: Shilabo, Doboweyn, and Dolo Odo
(Table 11.1). There is no obvious
connection between these three districts Shilabo is a
pastoralist area, Doboweyn is mainly
agro-pastoral, and Dolo Odo is dominated by riverine farmers
except that Shilabo and
Doboweyn are geographically contiguous districts in Korahe
Zone. It is likely that these strong
signals of disapproval from selected districts reflect locally
specific concerns and issues. As noted
earlier, for instance, the Bantu farmers of Dolo Odo feel
politically marginalised because of their
ethnicity. Also, Shilabo District has suffered from several
incidents of violent conflict in recent
years (as discussed above), which the government has not yet
managed to resolve. The people
of Shilabo are also suspicious of government intentions
regarding local deposits of natural gas.
117
In discussions with communities during fieldwork, people

expressed strong and generally
critical views about their relationship and interactions with
government. The following
selection is typical of many comments made that suggest an
absence of effective
government in Somali Region.
We do not know if we have representation. It seems no
one is conveying our
problems to the government.
Nothing changes. Electing leaders is a waste of time.
Government, government what government are you
talking about? We only see the
army, if thats what you mean.
No one talks to us to ask what we need. The
government does not exist here.
The government does nothing at all for us.
One elder in Kelafo District complained that successive
regimes have marginalised the district,
and Somali Region in general.75 He argued that elected local
representatives cannot have any
positive impact, because they are effectively voiceless and

quickly forget their roots in the rural
constituencies: The elected representatives do not look back,
they go and disappear in Jigjiga.
Asked whether decentralisation is making a difference, he
replied that no basic services have yet
been provided, so the local people can see no benefits from
government at either the central
or decentralised level: Decentralisation is a big propaganda
there has not been any positive
outcome that helped us in any way. Others recognised the
potential for decentralisation to
make a difference to local people, but identified serious
shortcomings in its implementation.
Decentralisation can be a good process, but there must
be clear communication
between the different levels. The District must pass
messages to the regional
government and then they must pass it onto the federal
government. At the moment,
decentralisation only means that we have these different
levels, but not that there is
any communication between the levels.76
We cant only blame the government in Addis. Our own

government in Jigjiga is
there, and there is also our local government

representatives in Gashamo, but they
dont help us even though they are our own people.
They arent conveying our
messages and needs to Jigjiga and the federal government.77
Another farmer in Kelafo District spoke for many others in

his community when he
expressed discontent about the lack of delivery of essential
services, despite the promises of
elections and decentralisation: If we can get anything at all
from the government, we
would like schools, health facilities and irrigation equipment.
But we doubt that we shall
ever get any of these. We have been forgotten.7 8
Complaints about being forgotten and invisible to

government were common: We are
in the country but invisible to those who exercise ruling
rights over us. Alternatively,
government was accused of being invisible to local people:
We have not seen any
government or NGO officials, in fact, this is the first time
that someone has come to talk
to us. A more benign interpretation attributed the failure of
the administration to deliver
basic services to lack of government capacity: We think that
even the government itself is
as weak as us and cannot afford to assist.79
75
76
77
78
79
118
Key informant interview, male elder, Kelafo District.

Male focus group participant, Gashamo town.
Male focus group participant, Gashamo town.
Male focus group participant, Kelafo District.
Male focus group participant, Shinile District.
IDS
RESEA
RCH
REPOR
T 57
Another benign view, articulated by many respondents

throughout rural Somali Region, is
that their neglect in terms of basic services is related to their
geographic isolation and
political marginalisation within Ethiopia: We are forgotten
here because we are so far from
Addis, and so close to Somaliland. Others in eastern Somali
Region added that their
proximity to Somaliland means that they receive more
support from NGOs there than from
the Ethiopian government.
We have received more assistance from Somaliland than
Ethiopia. We have received a
veterinary centre; health centre; terracing; restocking of
animals of people who have
lost animals; prevention of soil erosion; repairing
berkad. We have received none of
these things from Addis no schools, no health centre, nothing.8 0
Finally, participants in a womens focus group discussion in

rural Gashamo complained that
they are ignored by politicians, who talk only to men: We
have MPs in Gashamo but they
never come and meet with us. Either they only go to
Gashamo town or else they talk to
the men only.8 1
12
Gendered
vulnerability
Somali society is highly patriarchal, and females in any
patriarchal society face genderspecific risks and vulnerabilities, or experience the effects of
generic risks and vulnerabilities
more acutely than do males. The most extreme outcome of
gendered vulnerability is higher
mortality rates for women and girls, and this chapter presents
demographic evidence from
the 1997 census for hundreds of thousands of missing
females in Somali Region. Our
household survey also finds that life expectancy is
significantly lower for women than men.
Explanations for this discrepancy are sought at the household
level and in the labour

market, where women face constrained employment options.
On a positive note, there is
evidence that women are gaining increasing autonomy, partly
as a consequence of the
recent sequence of droughts, where loss of livestock
translates into loss of male power.
12.1 Gender bias in Somali Region

Although women in Ethiopia enjoy constitutional equality
with men, they are underrepresented at all levels of government federal, regional
and local. In Somali Region there
is currently only one woman MP (there were previously two)
in a Regional Parliament of
over 160 MPs. In the Council of Elders, two women sit
alongside 13 men an innovation
imposed on the Gurti from above. In January 2004, district
elections in Somali Region
returned 356 women out of 3,309 councillors elected (NEB
2004) just 10.8 per cent. This
is despite official instructions from the ruling party to select
more women candidates, with
a target of one-third female representation in regional
parliaments and district councils. If
women are not adequately represented in key institutions that
exercise power, it is highly
unlikely that their concerns will be reflected when decisions
are taken (or not taken) that
substantively affect their lives.
Given the fact that women have been artificially introduced
into political fora in Somali
Region, a related consideration is how effective these women
MPs and Gurti members can
be within such patriarchal institutions. The female Gurti
members clearly do not regard
themselves as fully equal members. One of them told us: We
only listen, and try not to
interfere with the proceedings of the elders on the Gurti
committee. It is unclear whether
the womens deference reflects respect for the age and
seniority of the male elders,
nervousness at being outnumbered, and/or a feeling that they
are intruding on a space that
has traditionally been reserved for men. The paradox is that
Somali women throughout the
80 Male focus group participant, Gashamo town.
119
Figure 12.1 Sex ratios in Somali Region*

220
Cen
sus
200
180
160
140
120
Survey
100
80
60
0-4 5-9 10 -14 15 -1 9 20-24 2 5-29 30 -34 35 -3 9 40-44 4 5-49 50 -54 55 -5 9 60-69 7 0-79 80+
* Sex ratio = males per 100 females.

Horn of Africa have a reputation for being strong-willed and
influential in many arenas
they are famous for mobilising their communities around
political issues, and many women
traders are wealthy and respected yet they appear unable to
break through into formal
political leadership roles.
More disturbing evidence of the consequences of gender bias
comes from the Somali
Region Census of 1997. Disaggregating the population by
sex and by age cohort reveals a
distinctive pattern, with three components. In the child and
adolescent cohorts (020 years
old), males outnumber females. Among young adults (2540
years old), however, there are
more females than males. Among middleaged and older
people (all cohorts over years old),
males once again outnumber females (Figure 12.1). It should
be noted that the census in
Somali Region was difficult and controversial, and had to be
done twice. Some districts
were not visited at all, for security reasons. Nonetheless, it is
difficult to explain away this
systematic pattern of sex ratios as methodological error or
under-enumeration. For one
thing, the distinctive pattern of the Census is mirrored in
demographic data collected for
this survey, which enumerated over 8,600 individuals living
in 1,100 households and found an
almost identical distribution by agesex cohort.
The most skewed sex ratios occur in the older age groups.
Among people over 60 years of

age, the sex ratio peaks at 203, meaning that there are twice
as many men over 60 as
women. In relative terms, gender-biased demographic
outcomes are most skewed against
older women. In absolute numbers, however, since this is a
very young population 59 per
cent of residents of Somali Region are under 20 years old
male bias affects younger age
cohorts most, where the sex ratio averages 129 males per 100
females. The intriguing
reversal that occurs in sex ratios among young adults is also
found in both the census and
our survey data. Among 2540-year-olds, the sex ratio falls
to 79 males per 100 females.
There is no obvious explanation for this anomaly, but it might
be related to gendered
differences in mobility and migration: men in this age group
who have the opportunity are
quite likely to be travelling abroad. It is also possible that
conflict-related deaths affect young
adult men disproportionately, though this effect is unlikely to
be large enough to produce
such a dramatic demographic outcome.
Overall, these census and survey statistics suggest that there
are large numbers of missing
females in Somali Region. Simply adding up the number of
excess males by age cohort in
the census data reveals that there are some 300,000 missing
females under 25 years of
age, or 22 per cent fewer females than males. Conversely, in
the 2539-year-old cohort,
women outnumbered men by 57,000 in 1997, though it is
likely that most of these missing
men are only temporary absentees from the region. Above
the age of 40, however, sex
ratios that peak at 2 to 1 mean that there are a further 65,000
missing women (Figure
120
IDS
RESEA
RCH
REPOR
T 57
Figure 12.2 Missing females in Somali Region (1997

Census data)
100 ,000
75 ,000
50 ,000
25 ,000
0
0-4 5-9 1 0-14 15 -1 9 20 -2 4 2 5-29 3 0-34 3 5-39 40 -4 4 45 -4 9 50-54 5 5-59 6
0-69 70 -7 9 8 0+
-25 ,000
Figure 12.3 Crude life expectancy estimates (household

survey data)
60
Male
51
51
Female
4
8
50
45
41
41
39
38
40
37
33
30
20
10
0
P ast oralist Ag ro p astoral F arm ers Urb an T O T AL
12.2). Most of these excess males reflect the fact that men
live longer than women in
Somali Region.
These gender disparities are also reflected in mortality rates.
A baby boy born in Somali
Region has a 22 per cent greater chance of surviving to the
age of five than a baby girl.
According to census data, life expectancy at birth is 56 years
for males and 53 for females.
As the census report concludes: life expectancy for males is
higher than females, unlike the
life expectancy of most other regions (FDRE 1998). Our
household survey questionnaire
allowed crude estimates to be made of life expectancy based
on self-reported mortality.
These estimates (derived from a sample of 8,600 individuals)
are less reliable than the
calculations presented in the census report (derived from a
regional population of 3.4
million), but they are indicative of relative trends across the
survey population. Our survey
data confirm that men outlive women in Somali Region
(Figure 12.3). Moreover, gender
differentials in life expectancy are greatest in pastoralist and
agro-pastoralist households. The
gap is less in farming households, and in urban households
women actually have a higher life
expectancy than men, suggesting that urban life is less harsh
for women.
121
Box 12.1 Celebrating sons:

discrimination in Somali Region
intra-household
Girl no. 1: In Somali culture, most mothers prefer

to take care of their sons than
their daughters. Most mothers prefer to have boy
babies, so they take better care of
them than their girl babies. When a boy is born, there
is great celebration. This isnt
the case when a girl is born.
Girl no. 2: Most girl babies and children are thinner
that the boys because their
mothers dont treat them and feed them as well as the
boys.
Girl no. 3: In our home, my uncle and male cousins
get the best food in the family.
All the women just serve them, even though they are
thinner than the men and
deserve more food. Even among the children, my
mother and aunts give the small
boys better and more food than the girls.
Girl no. 4: My mother gives my sisters herbs to
drink when they are sick, but buys
proper drugs for my brothers. I dont like it. I wont
behave like that if I have children.
I will treat them the same.
Girl no. 5: When my brother gets sick, they take him
immediately to the Health
Centre or spend money buying tablets from the
pharmacy. When my sister or me
gets sick, they say: shell get better, its not necessary
to take her to the Health
Centre.
Girl no. 6: My mother is always taking my younger
brother to the Health Centre and
buying him drugs even when he only has a slight
cough. But she never takes my small
sister although she is a sickly child and my brother
isnt.
Source: Focus group discussion, female high school
pupils, rural Gashamo District.
12.2 Intra-household discrimination

One plausible explanation for gender-biased well-being
outcomes is intra-household
behaviour that favours males at the expense of female family
members. It is difficult to
obtain rigorous and credible evidence on discriminatory

behaviours, so here we report on
qualitative findings from discussions with older children and
young adults in Somali Region.
In a discussion with high school pupils in rural Gashamo
District, many girls remarked that
their parents display an overt male bias in several ways,
including: (1) son preference in terms
of fertility; (2) male bias in intra-household allocation of
resources such as food; and
(3) favouring sons with access to services such as health care
and education. Many of these
teenagers gave examples from their own family, and most
confirmed that they had had the
same experiences, especially when they were younger.
Interestingly, these girls blamed their
mothers as much as their fathers for neglecting themselves
and their sisters (Box 12.1).
This evidence of internalisation of patriarchal values is not
unique to Somali culture; in many
other cultures women subordinate their own needs and those
of female relatives to the
needs of their husband and sons. The starkest outcome of
these skewed gendered choices is
that women and girls in Somali Region live shorter and less
healthy lives than men and boys.
A large part of womens vulnerability in Somali Region
derives from their limited control
over key productive resources. We dont own anything.
There is nothing that belongs to us
women.8 2 This limited ownership of assets might explain why
women have less of a stake in
the pastoral economy and are therefore more willing to

contemplate a future outside
pastoralism than men, who own all the income-generating
assets and derive most of the
income from them. Even if a woman works hard and gets
something, the man will be the
owner. Thats the problem of us Somalis: women arent allowed to own anything.8 3
82 Womens focus group, rural Gashamo.

122
IDS
RESEA
RCH
REPOR
T 57
Pastoralist women complain specifically about their restricted

ownership of livestock, especially
large stock, which in Somali culture are traditionally the
property of men: Even if a woman
owns animals before she gets married, when she gets
married, the animals become her
husbands property.8 4 This sex segregation in livestock ownership
is changing, however. In
farming communities, such as the Bantu Somalis along the

Shabelle and Dawa/Ganale rivers,
women do not own land in their own right. Tradition dictates
that men control access to all
land, which legally belongs to the state: I know of no woman who owns land in this village.8 5
It is hardly surprising that Somali women feel undervalued

the half of men: If a man is
killed, his family have to be compensated with 100 camels. If
a woman is killed, she is
compensated with only 50 camels. So we are only treated as the
half of men.86 On the other
hand, women take a disproportionate role in domestic

reproduction, while also being actively
engaged in work that contributes to household food
production and income generation.
When we asked a group of women in a farming community
to describe their main livelihood
activities, they replied: Looking after the family, bringing up
children, cooking, fetching water
and working on the farm cutting grass, weeding, cultivating, and helping with the harvest.87
12.3 Khat consumption8 8
As everywhere in the world, financial problems are a

common source of marital stress
between husbands and wives. In Somali Region, where
alcohol consumption is prohibited by
the Muslim religion, chewing of khat is very prevalent among
men it is an almost exclusively
male pastime. Excessive khat consumption is blamed for
exacerbating many problems between
couples, including money troubles. Khat was even mentioned
as the main contributory cause to
several recent divorces: I divorced my first husband after
two children because he wasnt
bringing in any money. He was just spending any money he had on
khat.8 9 The intersection of
drought and khat was blamed by women as undermining

Somali culture and diminishing the
social status, economic contribution and personal self-esteem

of Somali men (Box 12.2).
While women lay the blame on men for becoming idle and
squandering household
resources on khat, men see their increasing consumption of
khat as a symptom of other
problems a response to the depression induced by losing
their animals to drought, feelings
of guilt about being unable to support their families, loss of
self-respect, being displaced to
towns where they are effectively unemployed and idle, and
susceptible to urban
temptations such as khat-chewing (Box 12.3). Like alcohol in
other cultures, khat is seen as a
vice indulged in by unsettled young men, mainly in urban
areas, who are uncertain about
their futures and have too much time on their hands.
More generally, women see men as unable to adapt to
changing circumstances, and living in
the past, and relying on past glories: It should be better to
say I own one donkey today
than to say My father owned 2000 camels. But instead,
they rely on their past glories and
spend time saying how wealthy they used to be. By contrast,
women see themselves as
being more flexible and practical, and better able to live in
the present: Women force
themselves to adapt to changing circumstances.90

85 Woman farmer, Dolo Odo District.
86 Womans focus group, rural Gashamo.
87 Womens focus group participant, Kelafo District.
88 Khat (also spelt Chat or Qat) refers to the leaves of the tree Catha
Edulis. It contains five active ingredients,
the most active being cathine and cathonine. The United States Drug
Enforcement Agency (DEA) has
classified cathonine as a Schedule I drug (most restrictive), and
cathine as a Schedule IV drug. Cathonine is
only present in freshly picked leaves, but turns into cathine after 48
hours (Luqman and Danowski 1976). As a
consequence, khat has to be consumed within hours of being picked.
It cannot be stored or hoarded,
necessitating a rapid distribution system.
123
Box 12.2 More important than food: womens

attitudes to chewing khat
Woman no. 1: Men are different today than before.
In the past, a man had lots of
pride and believed he had to take care of his wife and
children. Nowadays, men
arent as strong and proud as before. When they lose
their animals, they come to
town and just sit chewing khat. They dont care what
happens to their children: they
look to their wife to find a way of earning something. I
blame khat and the many
droughts for mens behaviour. We also used to have
khat when I was young, but men
knew how to use it then. These days men get frustrated
because they lose their
animals, and instead of doing anything useful, they
just sit and chew.
Woman no. 2: Women have worries too, but we
dont chew khat. Women have to
worry about the children; where the food is coming
from, and so on. You can go to
any tea-shop and youll find men drinking tea and
chewing khat. Youll find women
busy, trying somehow to make a living. Men just dont
care. If they have some small
money, theyll use it for khat instead of food for their
children.
Woman no. 3: When men get a small amount of
money, they spend it immediately
on khat. When they come home at night, and the wife
asks for money, and he says
he hasnt any, it causes an argument between them
because she know he has
chosen to spend his money on khat instead of on his
own children. They quarrel all
the time until they divorce.
Woman no. 4: There were many problems between
me and my husband, but it was
mainly because of money. He was also a khat seller
like me, but he used to chew all
his profits away or else give it free of charge to his
friends and relatives. So there
were always arguments between us. Khat seemed to be
more important than his
children and food for his family.
Woman no. 5: Many women divorce because of

khat. Before you used to hear of
one or two such divorces, now it is more and more.
Between us, in the last year,
weve heard of about 13 women who divorced because
of this reason. Weve all
heard rumours of others who are experiencing
problems with their husbands
chewing khat but havent yet divorced.
Source: Womens focus group discussion, rural
Gashamo.
12.4 Gendered division of incomes

In a culture dominated by livestock rearing, Somali men who
lose their animals often appear
to be unable or unwilling to take up alternative incomegenerating opportunities. According
to women, this is a matter of pride: Somali men are too proud
and independent to adjust
to the reality that they need to take up any opportunity to earn
income, even if its only
small money and requires working for someone else. As
one woman who runs a store in
Gashamo town argued:
The trouble with Somali men is that they dont see the
value of earning a bit of money.
Either they must earn big money or they just sit around
not earning anything at all.
Even when it comes to getting a job: most of them
arent interested in working for
someone else. They all want to be employers, but
nobody wants to be an employee. In
every society, there must be some employers, some
employees; some big traders, some
small traders. But Somali men dont seem to be able to understand this.9 1
Many Somali men do not contradict this representation of

themselves, instead seeing their
pride and independence as a virtue of their culture: Other
tribes are happy to do smallscale activities; but we are different. A Somali man feels
humiliated, like he isnt a man if he
91 Female storeowner, Gashamo town.
124
IDS
RESEA
RCH
REPOR
T 57
Box 12.3 Trying to forget our worries: mens

attitudes to chewing khat
Man no. 1: It is frustration that causes us to chew
like this. We are all proud men.
We feel very bad that we arent able to take care of our
families, but we cant do
anything because we only know this life and this
work.
Man no. 2: The problem is that we are trying to
forget our worries. We arent
interested in doing womens work: we are pastoralists
or big traders: thats what
Somali men do.
Man no. 3: Its mainly the young men that chew so
much. When we were young,
we knew khat, but we werent chewing it like the
young men now. When I go to
Gashamo, I see many young men just sitting in the tea
shops chewing khat.
Man no. 4: When Ive gone to towns, even
Gashamo, I see most of the men
chewing khat more than men here. The reason for
their behaviour is because they
are not suited to live in towns and cities. We arent
meant to be settled and living in
one place, and not moving with our animals.
Source: Mens focus group discussion, rural Gashamo.
doesnt have a big herd or if he isnt earning a lot of money from a
big business.9 2 One
consequence of this attitude is that income-generating

activities are implicitly stratified into
high-status, high-return activities, which are reserved for men
(these include livestock
rearing and large-scale trading), and low-status, low-return
activities, into which women are
channelled disproportionately (e.g. petty trading, collecting
firewood). One male pastoralist
articulated this patriarchal view of mens work and
womens work forcefully:
Im not interested in doing minor things like selling tea.
Thats okay for women to do.
If I was interested in something in that line, it would

have to be on a big scale like a
hotel or restaurant. You will never find a Somali man
doing petty trading! I would
rather prefer to starve than humiliate myself by doing womens petty trading.9 3
To some extent these gendered divisions in the labour market

are not as rigid as in many
other societies, and they are breaking down, due to a
combination of social pressure and
economic stress. More women are involved in large-scale
trading now, for instance, while
more men than before are taking up petty activities like
collecting firewood and burning
charcoal. The assertion of male pride by some men is at
variance with the reality that many
others are doing whatever is necessary to support their
families.
More generally, one area where womens lives have changed
is in terms of earning
independent income, which has increased in recent decades
as a direct response to the
crisis of pastoralist livelihoods. (Until about 20 years ago,
women werent doing any work
for cash. Women mainly started working for money because
the of increasing droughts and
the serious losses of animals that most households have
suffered.) As household resources
have dwindled or livestock herds have become more volatile,
so women have taken more
responsibility for bringing food and income into the home, a
responsibility which was
previously reserved for men. (In the old days, women hardly
earned an income the man
was supposed to take of everything. But now women are
forced to do something. In the
past, women werent involved in money matters, not like
now.)
Notwithstanding these changes, a focus group discussion in
Gashamo town revealed that
Somali women still complain about being under the control
of men, at both the household
and community levels: We cant do anything without the
permission of men our fathers,
husbands, brothers and elders. For instance, husbands
sometimes try to prevent their wives
from working outside the home: I know of some men who
dont want their wives to do
125
any work. They want them to remain in the house because

they dont like their wives to
meet so many men in their shops and stalls. On the other
hand, these attitudes are not
universal in Somali Region, and they are also changing: The
drought has forced such men to
change. If their wives dont do anything, the children will
starve. So they are forced to give
their permission for their wives to work.94
Loss of (male-owned) livestock also translates directly into

loss of male power, which
women can and do exploit. Some women faced with
recalcitrant husbands have successfully
lobbied the elders to persuade their husbands to grant them
economic independence,
making the argument that allowing women to work is often
the only way to keep families
together. One woman told us of her success in changing the
attitudes of the elders in her
community:
Because men are just sitting around chewing khat, they
are losing some of their
former power. I wanted to do some business, but my
husband refused me permission.
There was conflict between us, and I decided to go to
the elders and explain my case.
Because he wasnt working or bringing in anything,
they persuaded him to let me
work because I would make a contribution and prevent
the family from becoming
beggars. If he was bringing in money, they would not
have been able to persuade him.
The drought and loss of livestock has brought this change in the elders and the men.9 5
It follows that, paradoxically, droughts and other processes of

impoverishment are
enhancing womens status with respect to men. Another
woman explained how the
balance of power within her home has shifted away from
husbands and towards wives:
In my home, my husband is still the main decisionmaker, but he has to consult me on
all the important and big decisions. They have realised
that we women are very smart
and hard-working and that they are depending on us
more than we are depending on
them. In my home, I am the only one that is earning
anything, because my husband

lost many animals and then sold the remaining animals.
Because most of us are
supporting our husbands and our children on our own,
they have no choice but to
consult us. Some men are still behaving stubborn and
trying to act like the boss. But
they soon realise that when it comes to earnings, we are
actually the bosses. They are
not allowed to take our money from us.9 6
In Somali society, women generally retain independent

control over their own earnings.
(When a woman earns an income, she can spend her money
as she wants. She has
complete control over it because she earned it. She doesnt
have to consult her husband.)
On the other hand, as women acquire more economic
autonomy, they often find that
more responsibility for household expenses food, groceries,
health care, schooling falls
on their shoulders by default (for instance, if men migrate
and send nothing back to support
the women and children), or is devolved to them by their
husbands: In the old days, men
used to be responsible for their families. Now, that has
changed, and women are the ones
responsible for the family.9 7 In some senses, therefore, this
increased autonomy has been
accompanied by an increased burden of responsibility on

those people within Somali society
who already take on more responsibility for domestic
reproduction and household survival
than anyone else.
94
95
96
97
126
Female focus group participant, Gashamo town.

IDS
RESEA
RCH
REPOR
T 57
12.5 Women and the future

It is important to emphasise that many of the statements made
by women throughout this
report came from focus group discussions in Gashamo
District, in the aftermath of a severe
drought. Episodes of severe livelihood stress tend to magnify
intra-household tensions, as
roles and responsibilities of different household members
change and authority shifts. In this
context, there is no doubt that womens critical comments
about men reflect a collapse of
the ethos that men take care of women, just as the fact that
many men are resorting to
chewing large amounts of khat is an abnormal behaviour,
provoked by their inability to
uphold their responsibilities to their wives and children.
Some of these families will not
survive these shocks and stresses, while others might well
revert to their previous way of
operating, with traditional gender roles restored.
Overall, women in Somali Region acknowledge that in many
respects their lives are better
than their mothers lives, mainly because women have more
freedom and independence
today than they had in the past. (Old people always say the
past was better. But in the past,
women were forced to remain in the house. Nowadays
women can do something and earn
a little money. We have more power than before.)
Women also identify modernisation and communications
technology as powerful positive
forces for change in Somali society:
Because of better communication and modernisation,
things are changing and women
are in a better position. We all know of other places and
cultures through TV and
magazines, and can see that there are different
alternatives. Before, our mothers
didnt know anything else, so they just lived like before.9 8
These women are also optimistic about the future, believing

that the recent trend towards
more autonomy will continue for their daughters generation:
Women are now more
independent and free. We are still controlled by our husbands,

but not as much as before. It
will get better for our daughters.9 9
98 Female focus group participant, Gashamo town.

99 Female focus group participant, Gashamo town.
127
Section 4 Responding
to
vulnerability
Responses to vulnerability can be made by households
themselves (coping strategies), by
communities (informal transfers) and by public agencies
government and donors (formal
transfers). These three levels of response are explored in the
three chapters in this section.
13 Coping strategies
This chapter starts with a brief review of key findings from
the empirical literature on
household coping strategies in response to livelihood shocks.
It continues by presenting
survey findings on coping strategies actually adopted by
households in Somali Region, and
concludes with some observations on the gendered nature of
coping strategies.
13.1 Coping strategies in the literature

Households confronted with a livelihood shock that
undermines their access to food can
react in a number of ways. During the 1980s and 1990s, an
extensive literature on coping
strategies, usually in the context of droughts in sub-Saharan
Africa and South Asia,
concluded that these responses can be categorised and tend to
be adopted in fairly
predictable sequences.100 Figure 13.1 divides household responses
to food shocks into those
that protect consumption and those that modify

consumption. Protecting consumption
requires buying or being given food to maintain food intake
levels. Modifying consumption
includes reducing or diversifying consumption, or reducing
consumers by migrating or
sending some household members elsewhere.
The frequency of adoption of different coping strategies
provides an indication of the
sequence in which they are adopted. As a general rule,

strategies that are the least costly
(both economically and socially) and are most easily
reversible, are adopted first, as an
immediate response to a food shock. Examples include mild
rationing of food consumption,
or reducing non-essential spending. Strategies that are most
damaging to livelihoods or
social status, and most difficult to reverse, are adopted last,
after all other survival strategies
have been exhausted. Examples include: selling farmland,
begging, or engaging in illegal
activities.
13.2 Coping strategies adopted in Somali Region

Our survey data confirm that households in Somali Region
tend to follow the general
principles established in many studies elsewhere. Rationing
of food consumption (smaller
portions and fewer meals) is an immediate response to
austerity, and is adopted almost
universally by the 1,100 households in our sample: Before
these droughts, a Somali was
known by eating meat and drinking milk every day. Now, we
eat once or twice a day and
nearly never have meat.101
Another widely practised austerity measure is to reduce nonfood spending, primarily to

release household resources for food purchases. More than
half our sampled families reported
100 The term coping strategies has been criticised for its implication
that poor and vulnerable people can
survive almost any livelihood crisis on their own resources, when the
reality is that malnutrition and child
mortality rates in many places (including Somali Region) are
unacceptably high even at normal times (see
Bradbury (2000) on the normalisation of crises in Africa).
128
IDS
RESEA
RCH
REPOR
T 57
Figure 13.1 A categorisation of household coping strategies

Purchase food
Sell
assets
Use
cash
income
Borrow
Remittances
Protect
Receive
food
Charity
consumption
Food aid
Reduce
Smaller
portions
Livelihood shock
consumptio
n
Fewer meals
/day
Cheaper
staples
Diversify
Modify
Wild foods
consumptio
n
consumptio
n
No meat or
fish
Children
sent to
Reduce
better-off
relatives
consumers
Adults
migrate
Source: Devereux 1993.

cutting spending on non-food items when food is short. The
extent to which these household
resource management strategies are actually implemented

depends on the ability of the family
to raise the cash or food they require from other sources (e.g.
food aid or remittances).
Two-thirds of households surveyed call on support from
relatives and friends, and receive food
and other assistance in times of need (Table 13.1).1 0 2 This is
almost identical to the proportion of
households that received food aid during 2004 (n=749/1,093

=68.5 per cent), although food
aid is more concentrated among rural households (over 80
per cent) than urban households (7
per cent) (see Chapter 15). Conversely, a relatively high
proportion of urban households
(n=87/196 =44.4 per cent) receive informal support from
relatives and friends.
The next category of responses is the adoption of a range of
strategies to raise income for
food purchases. The diversity of income-generating activities
available to people in Somali
Region is highlighted by the dozen sources of alternative
incomes listed under this category
in Table 13.1.103 Most frequently adopted among these is sale of
assets (such as livestock),
followed by borrowing to buy food. Next come two lowstatus activities that are adopted
by the poor at all times, but by the non-poor only in times of
hardship: selling firewood and
selling charcoal. One household in ten admits sending its
children out to work for income.
Small minorities of households resort to potentially damaging
behaviour renting or
selling rights to farmland,1 04 begging and contraband trading
to raise cash for food
purchases when the usual sources of food and income fail.

Under the heading diversify consumption, significant
numbers of households collect wild
foods (18 per cent), hunt (13 per cent), or fish (7 per cent) to
supplement their diets when
necessary and where wild food, wild animals and
opportunities for fishing (mainly in Somali
Regions two permanent rivers) are available. Finally, about
one-third of households alleviate
the pressure on household resources by reducing the number
of mouths to feed, with some
102 A more detailed analysis of the complex nature of informal support is
provided in the next chapter.
103 These are alternative incomes in the sense that these activities are
not usually pursued by these households
in making a living. As discussed earlier in this report, most incomes
in Somali Region derive from livestock
production, crop farming, trading, providing informal services, and
(especially in urban areas) salaried
employment. None of these activities is included in Table 13.1.
104 Since all land in Ethiopia is constitutionally owned by the state,

individuals have no legal right to sell land
over which they have usufruct rights. However, an active land rental
market has emerged in all parts of
Ethiopia where land is scarce. The distinction between renting out
land and selling land is that rights to
farm a plot of land are transferred for one season only in the former
case, but permanently in the latter case.
129
Table 13.1 Household coping strategies in Somali Region

Coping strategy Pastoral Agro- Farmer Urban Total
pastoral
Reduce consumption
Reduce number of meals per day 92% 96% 94% 54% 942
(87%)
Smaller portions 89% 91% 90% 46% 894 (82%)
Reduce non-food spending 60% 64% 55% 39% 608 (56%)
Receive food
Support from relatives or friends 65% 71% 83% 44% 738
(68%)
Purchase food
Sell assets to buy food 34% 44% 58% 32% 464 (43%)
Borrow cash to buy food 38% 30% 41% 30% 383 (35%)
Sell firewood
24% 13% 35% 8% 227
(21%)
Sell charcoal
18% 11% 24% 6% 170
(16%)
Send children to work 14% 12% 11% 1% 112 (10%)
Rely on handicraft 4% 3% 9% 1% 49 (5%)
Begging
1% 4% 6% 2% 38 (3%)
Save fodder for sale 0% 1% 8% 2% 32 (3%)
Rent out animals for food or cash 2% 4% 4% 1% 30 (3%)
Sell land
1% 2% 5% 1% 25 (2%)
Rent out land
0% 0% 6% 1% 21 (2%)
Contraband trading 0% 1% 1% 1% 9 (1%)
Diversify consumption
Collect wild foods 9% 24% 26% 12% 195 (18%)
Hunting
8% 13% 29% 1% 146 (13%)
Fishing for food 1% 2% 23% 0% 74 (7%)
Reduce consumers
Migrate to another country 19% 14% 16% 5% 154 (14%)
Migrate to urban areas 15% 15% 11% 13% 145 (13%)
Migrate to IDP camp 3% 1% 3% 11% 42 (4%)
Strategies per household 5.0 5.1 6.4 3.1 5.1
family members travelling abroad, to a town or to an IDP
camp in search of food or work, or
alternatively to stay with relatives who are not affected by the
drought. One woman in
Gashamo town told us: Two cousins have sent their children
to live with me because they
live deep in the bush and have few animals remaining and little
food.10 5 Another case in point
is a polygamous household in Kebribayah, which divided

due to hunger, during the drought
in 2004, with the second wife taking her two young children
to Jigjiga to live until the
situation in Kebribayah improved. During the drought of
2000, one entire farming family
from Dolo Odo migrated to Luuq in Somalia to stay with
relatives until the crisis had passed.
Farmers adopt a greater number of responses to food shocks
than any other livelihood group:
6.4 strategies per household, against less than half that
number (3.1) by urban households, with
pastoralists and agro-pastoralists falling in between. Across
districts, farmers in Kelafo adopt the
highest number of strategies (7.8) and urban households in
Jigjiga the least (1.7). It is unclear
whether this signifies that farmers have a wider variety of
choices, or a greater need to adopt
105 Female household head, Gashamo town.
130
IDS
RESEA
RCH
REPOR
T 57
Box 13.1 Case study: household drought coping

strategies
What we did in the drought, was cut down on the
number of meals we ate: we
reduced the number from three to two per day. We also
reduced the amount and
quality of food we ate at each meal for both adults
and children.
When we lost animals, we sold maize from the small
of piece of land where we
grow maize, to buy animals. The land is usually just
enough to grow enough for
household needs; usually we dont sell anything.
When things are really bad with the animals, we also
have the option of moving
with the animals. We only do that as a last option,
though. These days, it is more
common for some family members to move while
others remain behind. But when
things are really serious, then we all move.
We still help each other a lot. For example, if
someone dies, and the family is poor,
before going to the grave, each family cooks food and
tea. Each family also gives 10
Birr to the bereaved family. This happens not only in
times of death, but when a
family has a loss of any kind. If people have more, or
in times of non-drought, they
can give one goat to the bereaved family. Because we
only have two cows, I will
either give 10 Birr or cook some food (awino).
Source: Interview with an agro-pastoralist woman in
Kebribayah.
coping strategies because they are exposed to more
frequent or severe shocks than other
households, for instance, or because they are less able to
cope. Evidence presented earlier in this
report on household incomes revealed that farmers are not
poorer on average than other rural
households, though they are less wealthy in terms of asset
ownership (especially livestock).
An additional factor is that farmers are less mobile than
pastoralists and agro-pastoralists,

and in a region as drought-prone and volatile as Somali
Region, mobility is a vital defence
against livelihood shocks. On the other hand, since more than
one farming household in
four in our survey had at least one member migrating to
another country, to an urban area
or even to an IDP camp as a drought survival strategy, this
suggests high levels of livelihood
stress among farmers. When asked what they do in
anticipation of drought, focus group
participants in a farming community in Kelafo District
claimed that they first cut down on
consumption, then slaughter or sell the few animals they
have, and as a last resort migrate
across the border to Somalia.
Box 13.1 describes the main responses adopted by one agropastoralist household in
Kebribayah to the drought of 2004.
13.3 Gendered coping strategies in Somali

Region
Many women have formed or joined groups that provide
mutual support, such as savings
groups. These groups appear to be more common in urban
areas, perhaps because people in
rural areas live in smaller communities and are more mobile.
Urban women also control
more property and are more engaged in trade and business, so
can benefit more from
borrowing and savings facilities. However, some womens
savings groups were found in rural
settlements as well, such as Gashamo town.
I am a member of a savings group with about 23
women. Each woman pays 20 Birr
every day. When your turn comes around, you receive 2
million Shillings. I joined in
order to save money. When I receive my lump sum, I
can buy food and clothes for the
children. That helps me because the money from my
khat business is only small
amounts each day. My turn comes around every three months or so.1 06
106 Female khat trader, Gashamo town.
131
Informal institutions such as rotating savings groups can

provide some support in times of
need. However, they are more effective in cases of
idiosyncratic shocks (e.g. an illness or
death in the family) than covariate shocks (e.g. a districtwide drought). At such times, all
members of the savings group will want to withdraw funds at
the same time. Nonetheless,
savings groups can be seen as a response to vulnerability and
the absence of financial
markets (savings and insurance) and effective formal safety
nets in Somali Region.
There is some evidence that the impacts of livelihood shocks
within the household are
gendered, in the sense that women and men take the strain in
different ways. Asked
whether drought affects women any differently from men, all
the participants in a womens
focus group discussion in an IDP camp answered in the
affirmative, and some argued that
droughts affect women and children more severely than men.
We care for the children and since drought affects the
availability of milk, children
become malnourished, lactating mothers grow weaker,
women have to trek long
distances to fetch water, and some husbands migrate,
leaving women not only to look
after the children but also to provide the household
food. These extra responsibilities
are back breaking.1 07
The reasons why men migrate during difficult times may be

in the best interests of the
family to preserve the households livestock by finding
water and grazing elsewhere, to
search for food and income, or simply to reduce the pressure
on limited household
resources. Nonetheless, a major consequence of mens
absence is increased pressure on
women, who necessarily take on additional responsibilities to
provide for all the household
members left behind.
In times of drought, because men dont know anything
else, the only thing they can
do, is to migrate. My husband has left me to take care of
the children while he is gone
with two sons and the animals. So, droughts increase

womens burden; its easier for
men they just leave. We have to see to everything and everybody left behind.1 08
In the drought, men migrate to other places either

looking for work or with their
livestock. My husband has been gone for more than a
year now. First, he left with the
animals, but he lost some more. Then, he left to try and
find work. But I had to
remain here to look after the children. This happened to other women also.1 09
According to several women, rationing of food consumption

within drought-affected
households is borne disproportionately by women, though
often this decision is taken by
women themselves in their roles as wives and mothers, they
choose to put the needs of
their husbands and children first. In extreme cases, this can
go so far that it has damaging
implications for womens well-being.
I was pregnant during the drought of 2004, and I had to
make the food go further.
Often, I used to go without in order to give more food to
my husband and the
children. The men dont know about those kind of
sacrifices we women make. I had
problems in childbirth because of my poor nutrition.1 10
107 Female focus group participant, Maromadobes IDP camp, Shinile

District.
132
IDS
RESEA
RCH
REPOR
T 57
14 Informal transfers
This chapter describes the range of informal transfers that are
made between relatives and
friends in Somali Region, enumerates the transfers actually
received by households in our
survey, and discusses the limitations of these mechanisms in
supporting vulnerable families
during crises.
14.1 Informal transfer mechanisms in Somali

Region
Somali culture is constructed around complex social
networks that serve a number of social
and livelihood-related functions. These institutions reinforce
clan solidarity, and also provide
a social safety net that delivers some protection against the
extreme variability of pastoralist
livelihoods. Importantly, this social protection is much more
deeply embedded within the
culture than simply gifts among relatives in times of need.
One reason for this is that gifts
and exchanges are not only made on an informal one-to-one
basis; there are complex
religious rules (such as zakaat) and social norms which mean
that food, cash, animals and
labour are regularly transferred at the level of the community
or society. The gift culture is
systematised and compulsory; not discretionary and
voluntary. Even in good years, resources
are shared and exchanged within and between (sub-)clans
through livestock rearing
agreements, compensation payments, feasts and social
ceremonies. These arrangements
build social cohesion, maintain the social order and provide a
level of social protection that
goes beyond social assistance to cope with drought or a death
in the family. They also keep
capital (income and assets) flowing through the pastoralist
economy, providing livelihood
support to poorer community members and acting as a
redistributive safety net when a
shock threatens lives and livelihoods.
To an extent that is unusual in most contemporary societies, a
wide range of informal

transfer mechanisms are institutionalised and operational in Somali
Region.11 1 These can be
clustered into three categories, according to the type of

resource transferred:
1 Redistribution of food: food or grain gifts; ciyi (distribution
of meat to neighbours after
a slaughter); Allah bari or sab (sacrifice made to feed
the poor); awino (cooking food for
the hungry); qharan (contribution of food or animals to
relatives); zero-interest grain
loans.
2 Redistribution of cash: remittances (from relatives living
temporarily or permanently
abroad, especially the Middle East, Europe and North
America); dhibaad (cash gift to
married daughters); other cash gift (e.g. sadaqa); zerointerest cash loans.
3 Redistribution of productive resources: free use of pack
animals (camels or donkeys);
irmaansi or maal (donation or loan of milk animals);
free use of oxen or plough; xoolo
goyn (restocking of poorer relatives); kaalo (gift of
livestock to newly-weds); goob
(watering livestock or tilling a farm for a days food);
free labour; seed loans or gifts.
Some terms apply to more than one form of resource transfer
kaalo, qharan and zakaat,
for instance, all describe contributions or gifts that can be
made in cash, food, or livestock.
Table 14.1 confirms that households in Somali Region
receive a number of informal transfers.
In our rural survey, the average household received three
different types of transfer (poorer
rural households receive six or seven) within the previous 12
months, while urban households,
which are generally less poor and less vulnerable, received
fewer forms of support.
Interestingly, remittances are the only type of transfer
received by more urban households
(38 per cent) than rural households (14 per cent), which
reflects the better connectedness of
urban households with the wider world, and also the fact that
remittances are often used to
help relatives in Somali Region relocate out of rural
livelihoods and settle in towns.
111 Definitions of several of these informal mechanisms are provided in
the Glossary of this report.
133
Table 14.1 Transfers received in Somali Region, 2004/5

Informal transfer Pastoral Agro Farmer Urban Total
pastoral
Distribution of slaughtered meat (ciyi) 71% 71% 64% 48%
704 (65%)
Free labour (goob)
45% 51% 48%
5% 439 (40%)
Donation/loan of milk animals (irmaansi) 39% 47% 39% 7%
388 (36%)
Sacrifice to feed the poor (Allah bari, sab) 34% 36% 32%
22% 347 (32%)
Free use of pack animal (camel, donkey) 16% 44% 17% 1%
229 (21%)
Remittances (from relatives elsewhere) 20% 7% 14% 38%
197 (18%)
Gift to married daughters (dhibaad) 21% 18% 14% 16% 190
(17%)
Restocking poorer relatives (xoolo goyn) 18% 21% 14% 2%
164 (15%)
Gift of livestock to newly-weds (kaalo) 19% 16% 12% 8%
157 (14%)
Free use of oxen or plough 1% 20% 9% 0% 90 (8%)
Zakaat
7% 6% 5% 6% 66 (6%)
Seed gift
0% 3% 12% 0% 44 (4%)
Other cash gift (sadaqa) 5% 3% 3% 2% 37 (3%)
Food or grain gift
4% 3% 4% 1%
36 (3%)
Cash loan (zero interest) 5% 1% 3% 3% 33 (3%)
Seed loan (zero interest) 1% 2% 5% 0% 27 (2%)
Grain loan (zero interest) 0% 0% 3% 1% 12 (1%)
Transfers per household 3.1 3.5 3.0 1.6 2.9
The most frequently mentioned informal transfer is food
(ciyi the customary distribution of
meat to neighbours after an animal is slaughtered). This
practice is almost universal among
livestock owners, but is also common in urban areas where
livestock ownership is low. A
variation on ciyi is Allah bari or sab sacrificing an animal
specifically to provide food for the
poor. This is also one of the most frequently reported
informal transfers, received by one in
three households surveyed.
After ciyi, the second most frequent transfer is free labour

(goob), which was received by
half of all rural households. Among pastoralists, goob takes
the form of watering and
grazing livestock in return for a days food plus tea or khat.
Among farmers, goob involves
helping someone who is labour-constrained (e.g. an older
person or a female-headed
household without adult males) to plough or weed their field,
again in return for a daily
meal and tea or khat. Sometimes this kind of labour-sharing
occurs on an exchange basis
groups of farmers working on each others fields as labour
parties though this traditional
practice is less popular nowadays than in the past.
Transfers of animals are also common between rural
households, but almost non-existent
among urban households. Livestock owners often lend or
donate milking animals to relatives
(irmaansi or maal), to be reared in exchange for their milk
and/or offspring. Alternatively,
wealthier livestock owners with pack animals to spare will
lend a camel or donkey to poorer
relatives. Two other important mechanisms for redistributing
livestock are restocking of poor
relatives who have lost their animals (xoolo goyn), and gifts
of livestock to newly married
couples (kaalo). In our survey, significant numbers of rural
households had benefited from
irmaansi (42 per cent), free use of pack animals (26 per
cent), xoolo goyn (18 per cent) and
kaalo (16 per cent), in the past year. Among agro-pastoralists
(20 per cent) and farmers
(9 per cent), lending or sharing oxen and ox-ploughs is also
practised.
It is interesting to note from Table 14.1 how few informal
transfers are made in the form of loans
rather than gifts. Several gifts of cash, food grains and seeds
were reported. Access to credit is
constrained, from both formal and informal lenders, and only
a few households reported receiving
loans at zero interest from friends or relatives, whether in the
form of cash, food or seeds. Most
loans of food (grain) and seeds, and most seed loans, occur
among farming households.
134
IDS
RESEA
RCH
REPOR
T 57
14.2 Access to informal transfers

An important factor in terms of access to informal support
concerns circles of responsibility
(at the societal level) and the extensiveness of social
networks (at the household or family
level). The circles of responsibility and of support for any
Somali start with his or her
immediate family, then radiate out to the extended family, the
sub-clan, the clan, and finally
to Somalis in general. Having a relative who is living
elsewhere and/or is employed in the
formal sector, and who can be approached in times of need,
can be an important source of
support, especially because salaries are not dependent on
rainfall, so employees incomes are
not correlated with drought. I have a brother who is a
security guard in Jigjiga. He is also
married with many children. But when he can, he tries to give us something.11 2
A similar point applies to remittances, which are a major

source of transfers into Somali
Region from abroad, and provide an independent source of
income for many families, either
on a regular basis (e.g. monthly) or in response to requests
for assistance at times of
livelihood stress. I am not lucky enough to have any relative
living abroad who could send
me anything. I once knew a lady who used to receive
contributions from relatives in
England and the USA. She never had any worries or problems like
the rest of us.11 3 One
household in Jigjiga town comprises a 39-year-old man, his

31-year-old wife, her 18-yearold sister, and their three young children aged 7, 6 and 4. The
husband sells building
materials for a living, earning around 4,500 Birr per month.
They own no livestock and do
no other paid work. However, the wife receives 300 dollars
each month from her brother
living in the USA. This is sufficient to give the family a
comfortable living.
Remittance flows into Somali Region are large and extremely
difficult to quantify, but it
should be emphasised that many families receive remitted
income regularly often in the
form of a monthly money transfer and not only during
emergencies. Interestingly,
remittance income is rarely invested in pastoralism, but
instead is used to finance alternative
livelihoods. Mainly it is spent on: (1) building permanent
structures (mainly houses or stores in
urban areas); (2) working capital for business (including
trading); (3) educating children; and
(4) to support the family though a shock (such as a costly
illness, or a social obligation such
as a dowry payment)11 4 or a community-wide shock such as a
drought.
The problem with calling on employed relatives, or

requesting remittances from relatives
living abroad, is that access to these sources of informal
social protection depends on each
familys circumstances, but is not available to all. If the
extended family cannot offer
adequate protection in times of need, people turn next to the
institutionalised support
provided by their clan.
When I face problems, I first turn to my mother and
father for help. The next thing I
rely on is kaalo I inform elders and close family of my
problem, and they ask for
contributions from everyone.11 5
Being Somali, and Isaq, and of the same sub-clans, they

are the first people we turn
to for help. According to qharan, people will make a
contribution of a goat or cow.
Then, if the problem isnt solved at that level, those who
have relatives outside will
send word to them requesting help. This will still be
qharan.

113 Elderly widow, rural Gashamo.
114 In contemporary Somali culture, there is no fixed dowry payment.
Instead, the amount that parents of the
groom pay to parents of the bride is negotiated on an individual
basis. Traditionally, dowry was paid in
livestock (e.g. 20 camels), but it has become increasingly monetised,
usually being paid in cash (often in US
dollars, reflecting the importance of remittances from the Somali
diaspora abroad). Families with relatives
living in the USA might pay as much as US$5,000, while those with
no relatives abroad will pay just US$
100. This gives some indication of the scale of resources that those
with access to remittances can call on,
either regularly or in times of need.
115 A widow in Gashamo town.
135
Closely related to the immediate circles of extended family

and sub-clan in Somali communities
are neighbours. Several respondents mentioned that they have
relationships of reciprocity with
their neighbours, which are important to maintain as sources
of social insurance in hard times.
A few weeks ago, I had to ask my neighbour for food
because we didnt have anything
in the house. She gave me food for me and the children
because my husband is away
and we didnt have any food. He came back and had
some money he was involved in
charcoal burning with some men in the bush
somewhere. Then we were able to help
my neighbour and her family because they were also in need.1 1 6
14.3 Availability of informal transfers

Despite the range and complexity of Somali informal transfer
mechanisms, and the fact that
they are deeply embedded in Somali society, there are many
reasons why the capacity of
informal redistribution to meet all social protection needs is
limited, especially when
covariate livelihood shocks like drought affect entire
communities. Reasons given by local
people include: declining capacity to provide support;
declining willingness to provide
support; and the shame associated with asking for help. These
can be categorised as
economic, cultural, and social reasons.
Many people in rural Somali Region commented on the
recent erosion of informal social
support systems. Mostly this was attributed to the recent
sequence of severe livelihood
shocks, which they described as equalisers: Everyone in
this community is at the same level
of income, so we couldnt help each other this year.1 17 One
consequence of these recurrent
shocks is that the ability of those worst affected to call on

others for support has been
compromised, because the basis for informal social transfers
has been undermined.
Nobody that I know received any assistance through
xoolo goyn, because everybody
was equally badly affected. The system can only work
when there are some better-off
people in the community, to help the less fortunate. But

now were seeing times
when everybody is in the same bad position. Before,
there were always some people
who were able to help others. It just shows how bad these present droughts are.11 8
Although the systems still exist, they have become less

important and less used
because of the increasingly difficult circumstances. For
example, this years drought,
which was called Sima [the equaliser], made it nearly
impossible for people to make
contributions and help each other since everyone became nearly equally poor.11 9
Before these recent droughts, we all used to help each

other through xoolo goyn,
where everyone used to contribute one animal to help
someone in need. Now,
everyone is equally poor, therefore we dont do xoolo
goyn. Its not because we dont
want to help each other. The old people tell us this is the
first time that people havent
been able to do xoolo goyn. It shows that times have never been as bad as this time.12 0
Even though the institutions continue to exist and do provide

vital support in many cases, their
effectiveness in practice may be diminishing over time. (No
I didnt ask for xoolo goyn. Who shall
we go to? These days everyone is in a bad situation.)
Declining capacity to assist has two
aspects: providing less support, and providing support less
frequently. One implication of this
latter aspect is that people needing assistance can call on their
relatives or clan only once or
116
117
118
119
120
136
Woman in rural Gashamo District.

Agropastoralist in Kebribayah District.
Male pastoralist, Gashamo.
Female focus group participant, Gashamo.
Woman teashop owner, rural Gashamo.
IDS
RESEA
RCH
REPOR
T 57
twice, after which that source of support dries up and these

people are left to cope on their
own. Because we are poorer, even when we can help, we can
usually only help once. But if the
person comes back tomorrow, its not possible to help again.
We are also not able to help with
as much as before.1 21
There is a widespread perception that the assistance people

could draw on in 2004 was less
than in the past. It is not yet clear whether this reflects a
temporary setback a short-term
contraction of social support due to the severity of recent
droughts or an irreversible
trend, because of a permanent shift in the culture away from
traditional informal
institutions of social protection.
Some respondents argue that the erosion of the economic
basis for informal transfers is driving
long-term social change in Somali society. The drought is
causing us to neglect our culture. We
arent able to make ayitu [charitable contributions]. You need
money and an income to do it.
We are losing our culture because of poverty.12 2 A specific
concern raised is that attitudes to
helping each other are changing the circles of responsibility

are collapsing and people are
becoming more individualistic and selfish. This view
resonates with evidence from elsewhere in
Africa, that processes of modernisation are introducing
materialistic values, commercialising
relationships and displacing traditional values of reciprocity
and non-commercial transactions. In
the Somali context, an increasing source of vulnerability is
that the effectiveness of informal
social protection might be eroding, to the extent that the gift
culture is in permanent decline.
This cumulative poverty and increased hunger has also
caused people to become selfish
and cruel. Even if they can help, they are hesitant
because they dont want to take
risks. They dont want to be left without something.
Poverty makes you more cautious.
If these droughts continue, we will lose these practices
completely. This also means
that poor people will be worse off, because they will not
be able to rely on help from
their community. In the past, if you were poor, you
always knew that you could get
support from others.12 3
One man interviewed in Fafaan Valley IDP Camp was

displaced there by the 2000 drought,
when all his 34 head of cattle died. He told us he would have
persisted with pastoralism if
the traditional resource-sharing institutions had assisted him
by restocking some of his lost
animals. He believes that these mutual support mechanisms
have been weakened by
successive droughts and that Somali cultural values have also
shifted against helping others,
except for immediate relatives. It is this negative combination
of economic and cultural
changes that forced him and his family out of pastoralism. He
has no plans to return to
livestock rearing, as he has no asset base and feels that it
would be impossible for him to
restart life as a pastoralist without support from the Somali
community, which he believes is
no longer guaranteed as it was in the past.1 2 4
Finally, there are social costs (shame and stigma) to calling

on the community for assistance,
which discourages some people from asking for help until it
is absolutely unavoidable.
I dont rely on any of the traditional Somali coping
strategies and networks. It is too
public, and there is a certain shame and stigma
involved. With qharan, for example, the
elders announce to the community: Alima needs
assistance. Can you please help by
making a contribution? I would be too shy to have my
business told to everyone:
everyone knows your circumstances and knows you are
struggling. So I only approach
my closest relatives and clansmen.12 5
121
122
123
124
125
Female focus group participant, Gashamo.

Life history inter view, Fafaan Valley IDP camp.
Woman store-owner, Gashamo town.
137
14.4 Gender and informal assistance

Although the provision of informal assistance is often
articulated as being dominated by men
not surprisingly, since men own most of the households
assets, control most of the resources
and usually have ultimate decision-making power within
Somali households womens roles in
supporting extended families in times of need are greater than
is often recognised. Womens
roles also tend to increase during periods of livelihood stress,
precisely because mens ability to
provide for the family declines as household assets decline
(e.g. if livestock die during drought).
This is evident from the case of one woman from Gashamo
town:
If a woman has a shop and her husband has lost his
livestock because of the drought, she
has to maintain the household on her small income. But
it is not only the immediate
family that rely on her. In my own case, when my
husband lost most of his animals, I am
the only one maintaining my six children, and some of
my husbands relatives who also
lost their animals and live out in the villages. Im
helping three other families: because
they are out in the bush villages, their wives arent trading or earning any income.1 26
There are also some informal institutions that exist solely for
women to provide support to
other women. One example is ischar, where women
spontaneously get together and
decide to make contributions to a woman who is in particular
need, for example if her
husband has died. These institutions also assist women in the
community who become de
facto female-headed households when their husbands
migrate during droughts.
Women are also helping each other because of the
drought. We give food to those
women whose husbands have migrated and who are
struggling without anything.
Other women make contributions to give something to
those women who are really
in need. Others try to give loans and credit to poor women.1 2 7
14.5 Zakaat
Zakaat is a mandatory contribution to the poor and the

needy that is made by all Muslims
who can afford it, and is payable during the annual Ramadan.
The minimum level before zakaat
must be paid is known as the nisaab, and is calculated as a
fraction of income earned, produce
harvested and assets owned during the year. For farm
produce, the nisaab is 612 kg; for
livestock, the nisaab is 5 for camels, 30 for cows, and 40 for
sheep and goats (Shuaib, 1991).
Figure 14.1 shows the year when households in each district
sub-sample paid zakaat most
recently, from this year (2004/5) to 4 or more years ago
(2000 or before). In each district,
a number of households have never paid zakaat despite all
being practising Muslims usually
because they are too poor to meet the nisaab. This question
was included in the survey
questionnaire as a way of capturing the impact of the 2004
drought on social obligations. As
might be expected, districts that were worst affected by the
drought saw a sharp decline in
zakaat payments. Only 3 per cent of households in Gashamo
and 10 per cent in Kebribayah
paid zakaat in 2004. This compares with 61 per cent in
Jigjiga town and 62 per cent in
Shilabo, and an average of 39 per cent of households across
the full sample (n=425/1,091).
This finding provides a powerful proxy indicator of the
severity of the drought that afflicted
Gashamo in 2004, where 66 per cent of households surveyed
had paid zakaat in 2003.
Figure 14.1 also tells us something about the distribution of
poverty in Somali Region.
Kebribayah is one of the poorest districts surveyed, in terms
of cash incomes and it has one
of the highest proportions of households that have never paid zakaat at all
(52 per cent).12 8 In
126 Woman teashop owner, rural Gashamo.
127 A woman in Gashamo town.
128 Only Gode town has a higher proportion of non-paying households
(60 per cent), but this is complicated to
interpret as the Gode sample included a cross-section of wealthier
and poorer urban resident s.
138
IDS
RESEA
RCH
REPOR
T 57
Figure 14.1 Most recent payment of zakaat, by district

100%
Never paid
90%
any zakat
80%
4 or more
years ago
70%
60%
3 years ago
50%
2 years ago
40%
30%
Last year
(2003/4)
20%
10%
T his year
(2004/5)
0%
Gashamo, which is wealthier in normal times but

vulnerable to livelihood shocks, only 22
per cent of households have never paid zakaat. Across the full
sample, one-third of
households (n=369/1,091 =34 per cent) had never paid
zakaat. This could be tested as a proxy
indicator for poverty. Simply asking the single question:
Have you ever paid zakaat? could be
a means of identifying the poorest households in their
communities. In our sample, this
indicator produces a proxy poverty headcount of 34 per cent.
In general, urban households are most likely to settle their
zakaat obligations in cash, though
crops maize, sorghum, wheat, rice are also offered, while
pastoralists pay mainly in
animals and agro-pastoralists pay in animals and/or crops. In
Jigjiga town, the average zakaat
payment in 2004 was 884 Birr, though one household with
seven members gave 14kg of
maize. (For each person in the house I paid 2kg.) The
pastoralists of Gashamo made their

most recent zakaat payments most typically in the form of
goats (n=48/68 =71 per cent),
followed by sheep [11/68 =16 per cent), camels [7/68 =10 per
cent), and in one case, a
cow. Among the agro-pastoralists of Kebribayah, payment in
the form of sheep and goats is
also common (n=26/43 =60 per cent), with maize and
sorghum crops being next most
significant [13/43 =30 per cent).
Despite the severity of the livelihoods crisis in Gashamo,
only 4 per cent of households
interviewed reported receiving assistance in the form of
zakaat in 2004. This is an indication
that pastoralists (at least in Gashamo) are generally regarded
as wealthy families that are
more likely to give charity to others than be in need of it
themselves. In Kebribayah, where
chronic poverty appears to be higher, 11 per cent of
households had been beneficiaries of
zakaat. In Jigjiga town, too, 11 per cent of households had
received zakaat in 2004 on
average, 220 Birr per household, plus a sheep or goat and
some wheat or maize. More
generally, there was a widely held view that zakaat has
declined in the face of recurrent
livelihood shocks: Zakaat has been decreasing in the last ten
years mainly because of the
many droughts. 2004 was the final straw.1 29
129 Womens focus group participant, rural Gashamo.
139
14.6 The limits of redistribution?

This chapter has provided evidence of the existence and
functioning of a wide range of
informal redistributive mechanisms within Somali Region.
Some of these are variations on
the traditional sharing and reciprocity arrangements found in
many pastoralist societies,
while others (such as zakaat) are charitable obligations
intrinsic to the Islamic religion.
Together, these mechanisms should provide a fairly
comprehensive informal social security
system. But how effective are these informal safety nets,
especially for those who need
support most (i.e. the poorest) and at times when support is
most needed (e.g. during
drought emergencies)? The idealisation of redistributive
mechanisms within pastoralist
societies and within Islamic culture might reflect an outdated
and romanticised view, rather
than a realistic understanding of how these mechanisms
operate in practice today.
At least three explanations were given for why informal
transfer mechanisms are less able
to deliver effective social protection to the poor and
vulnerable these days than in the past:
1 Less capacity to provide support (because of poverty,
exacerbated by equalising shocks).
2 Less willingness to provide support (because of cultural
change declining social solidarity).
3 Reluctance to ask others for help (because of social costs
shame and stigma).
There are also interactions between the economic and cultural
factors. One IDP in Gode
District explained how livelihood stress is driving cultural
change: Before, people used to
love one another so much so that they shared out the little
they had among the needy. But
now, I do not know why but that kind of love is diminishing,
maybe drought is hardening
peoples feelings.1 30
When asked if they had ever given or received zakaat,

respondents to our household survey
often laughed, and pointed out that contributing zakaat is
constrained in contexts of
widespread poverty: We are too poor for zakaat here. You
have to have at least 30 sheep
before you must give one for zakaat. Which of us here has 30
sheep? Nobody!13 1 On the
other hand, receiving zakaat requires having patronage

connections to wealthier people, or
access to those who collect and distribute zakaat
contributions. According to one key
informant in Jigjiga town:
Charity begins at home. So when you are rich enough to
give zakaat, you first look after
your poor relatives, then others in the clan who need
assistance. After that you might
give something to beggars in the street. If you have no
one to give to, there are people
collecting at the mosque who redistribute to the poor, so
you can give to them.
The problem for the rural poor of Somali Region is that they
are so far removed, both
socially and spatially, from the urban elites who have
significant resources to redistribute,
that they are unlikely to benefit from any redistributive
mechanisms based on patronage or
kin relationships, while the amounts collected at their local
mosques are trivial. As the
people of Kebribayah told us: Zakaat is for rich people and
those who know them; it is not
for people like us.
To conclude, then, the evidence suggests that informal
mechanisms of social protection in
rural Somali Region might well be limited in their
effectiveness in terms of assisting the
poorest and most vulnerable, for two reasons: horizontally,
poverty limits the pool of
resources needed to provide an economic basis for more than
strictly limited redistribution,
and vertically, the process of redistributive transfers
trickling down from wealthier to
poorer individuals appears to stop short of providing
adequate support to those people who
need these forms of social assistance most.
130 Life history inter view, Sagarabuur IDP camp, Gode.
131 Male pastoralist, rural Kebribayah.
140
IDS
RESEA
RCH
REPOR
T 57
15 Safety nets
This chapter focuses on the recent increase in food aid as a
response to a perceived increase
in vulnerability in Somali Region. Food aid deliveries and
receipts are documented, as well as
uses of food aid by recipients. Several concerns are raised,
including mis-targeting and
diversion of food aid; possible disincentive effects on local
production and markets; and a
mismatch between commodities provided and beneficiaries
tastes and preferences.
15.1 Trends in food aid deliveries

Food aid deliveries to Somali Region have increased
dramatically since the drought
emergency of 2000. In the mid-1990s, the proportion of the
regional population who were
declared to be in need of food assistance was negligible just
3 per cent in 1995 and 1996
(Table 15.1). In the food crisis year of 2000, this proportion
peaked at 40 per cent of the
regions population, and since that year it has remained
consistently at around one-quarter
of the population. In terms of numbers of targeted
beneficiaries, this increased from
approximately 100,000 people in the mid-1990s to over one
million people in four of the
last five years a tenfold escalation in food aid beneficiaries.
Two interpretations of this rapid increase in food aid
deliveries of Somali Region are
possible. The first is that this is an appropriate and
proportionate response to deteriorating
food security and rising chronic vulnerability in the region.
The second is that this is an
inappropriate and excessive reaction to the failure of the
federal government and the
international community to respond promptly to the food
crisis of 2000, so that food aid
flows have been maintained at an unnecessarily high level
ever since. The question remains:
why have food aid deliveries to Somali Region stayed so high
since 2000? Has there really
been a tenfold increase in vulnerability in less than a decade?
The answer probably lies between the two extremes.

Evidence presented in this report
suggests that vulnerability has risen during the past decade,
for most people in rural Somali
Region. However, this does not mean that delivering 200,000
tons of food aid per annum
is an effective response. For one thing, the nature of
vulnerability is complex and
multidimensional, and does not inevitably result in shortages
of food at the household level.
For instance, when vulnerability is related to conflict which
constrains livestock access to
grazing and water, rather than triggered by drought that
undermines food crop production,
it is not clear that food aid is either a necessary or sufficient
response. Another concern is
with the delivery of food aid, which is often poorly targeted,
patchy in coverage, and
transfers commodities that are not preferred by local people.
Table 15.1 Food aid deliveries to Somali Region, 19952004
Year People Per cent of Food aid Food aid per Food aid
declared in regional delivered beneficiary per
capita
need of food population (metric tons) (kg) (kg)
1995 110,000 3 n/a
1996 100,000 3 n/a
1997 235,000 7 n/a
1998 366,000 10 n/a
1999 400,000 11 110,545 276 30.29
2000 1,500,000 40 280,106 187 74.52
2001 1,000,000 26 88,658 89 22.90
2002 900,000 23 130,942 145 32.84
2003 1,147,070 28 203,569 177 49.56
2004 1,120,100 26 206,396 184 48.79
Source: Calculated from DPPC data, Addis Ababa.
141
Figure 15.1 Food aid on sale in Jigjiga town, February 2004
This photograph shows wholesalers selling bags of wheat

and corn soy blend (CSB) in Jigjiga
town market, during the Jilaal season of 2004. These are bags
of food aid, sold either by
the intended beneficiaries or by others who obtained this food
illegally. This foreign wheat
was selling at 1 Birr per tin. Across the street, women
retailers were selling local wheat at
1.5 Birr per tin. Their prices had already been pushed down
by competition from food aid,
and several retailers had reportedly been forced out of trading
as their profit margins were
squeezed. Some farmers in the Jigjiga plains have apparently
stopped producing wheat for
the market, unable to compete with free food aid being sold
at below cost on the market.
This alleged disincentive effect on local production and trade
needs to be carefully
investigated.13 2 Firm conclusions cannot be drawn on the basis of
our qualitative evidence. If
true, however, it sends a very negative message to

pastoralists, at a time when the
government is encouraging pastoralists to take up agriculture
and become agro-pastoralists
or settled farmers. Local farmers and traders cannot compete
against free food, even if they
drop their prices, which many are being forced to do.
These problems are exacerbated if food aid is delivered at the
wrong time, so that it
coincides with the harvest season. During fieldwork, farmers
in a community discussion held
in Kelafo District complained about the bad timing of food
aid deliveries. According to these
farmers, a large food aid distribution was made in their
communities just after the recent
harvest, which reduced market demand for their surplus grain
production and caused prices
to fall sharply.13 3 An intriguing contrast was observed between the
attitudes to food aid of
grain farmers and cash crop farmers. This particular village

has a strong market relationship
with neighbouring Somalia, and Kelafo farmers who export
onions and fruit crops noted
132 A recent review of the dependency effects of food aid concluded that
evidence for these negative effect s is
usually anecdotal and inconclusive, and argues for more rigorous
impact assessments to be conducted before
any definitive conclusions can be drawn (Barrett and Maxwell 2005).
133 Even crop farmers who do not produce surpluses tend to sell some
grain after the harvest, to raise cash for
other household needs, so these deficit producers also suffer an
income shock when food prices all because
of competition from food aid handouts.
142
IDS
RESEA
RCH
REPOR
T 57
that demand for their produce actually rises after a food aid
distribution in Somalia, since
recipients sell some food aid to buy vegetables and fruit.
Interestingly, even local people who are benefiting from
receiving food aid handouts
acknowledge that the amount of food aid coming into their
region is excessive and
damaging. At a workshop for this project, held with
pastoralists and stakeholders in Jigjiga
town, five out of six break-out groups that discussed this
issue concluded that food aid is
not a solution to food insecurity in Somali Region, except
possibly during emergencies. On
the contrary, they said, it is undermining local farmers and
distorting local markets.
Concerns were raised that Somali Region might become
another victim of the chronic
food aid dependency that affects other regions of Ethiopia,
especially the highlands.
15.2 Food aid targeting and distribution

In response to reports of extremely high global acute
malnutrition (GAM) rates for Gashamo
in early 2005 (19 per cent GAM, for instance, exceeding the
emergency threshold of 15 per
cent), food aid donors such as WFP argued that this could not
be caused by food deficits,
since large volumes of food aid were being delivered to
Somali Region, but must instead be
due to poor water supplies, causing illness and deaths from
diarrhoea and related diseases.
This might be at least partly true in which case emergency
water and sanitation
programming should be prioritised, rather than putting an
exclusive focus on food
programming or it might suggest that food aid targeting
procedures are seriously
inadequate, or that food aid is being diverted before reaching
its intended beneficiaries.
The distribution of food aid in Somali Region can be seen as
following a cascade
approach, from the centre down to targeted households.
Problems with food aid
distribution were recorded at several levels: from the
warehouse to the community, from

the community to the household, and from the household
head to individual household
members. At each stage, mis-targeting or diversion draws
some food out of the pipeline
and contributes to reducing the likelihood that food aid will
reach the most needy, for
whom it is intended.
15.2.1 From warehouse to community

Many rumours are circulating in Somali Region alleging
corruption in the distribution of food
aid. Some of these were verified by more than one
independent source, though the scale of
food aid diversion is impossible to quantify. Early in 2005 the
Somali regional DPPB Head and
his Deputy were sacked for corruption, following allegations
that DPPB staff were selling
substantial amounts of food aid and pocketing the proceeds.
One woman in Dolo Odo told us
she has heard stories of the local administration in Dolo town
selling food aid meant for local
people. She suggested that this needed to be investigated and,
if true, the personnel involved
should be punished and sacked. The fact that a flourishing
market for food aid wheat and corn
soy blend (CSB) exists throughout Somali Region that is not
legally prohibited there are no
restrictions on beneficiaries selling their rations facilitates
this corruption, as there is no way
of tracing where food aid being sold came from, or who sold
it.
The targeting procedures used to identify communities and
select households who are
eligible for food aid are not transparent to local people, many
of whom complained that they
are overlooked in food deliveries, without knowing why.
People in one drought-affected
agro-pastoral settlement in Kebribayah, for instance,
commented that they are aware of food
aid going to neighbouring communities, and cannot
understand why their own community is
overlooked: We see the food trucks going past, but they
never stop here.
In other cases, people complained that the amount of food
they receive is very variable
they have no idea what rations they are entitled to receive, or
how often and that food
aid deliveries are erratic and unpredictable. One household of
three adults and six children
in Doboweyn was given 50kg of wheat as food aid in 2003,
but only three cupfuls of
wheat in 2004 (about 11/2kg). A female-headed household in Dolo
Odo reported receiving
food aid only once a year, and only in small quantities 2kg
of wheat per household
143
member, like all other households in the village. (Table 15.2,

which shows that households in
Dolo Odo received, on average, just 7.7kg of food aid in
2004/05, supports this assertion.)
She described this amount of food as insignificant.
A few households living in border districts such as Gashamo
or Shilabo reported collecting
food aid from neighbouring Somaliland, where they generally
received much larger
allocations of food. Two households received 200kg and
another two received 250kg of
grain during 2004.
15.2.2 From community to household

Once food aid is delivered to communities it is often handed
over to senior community
members or a committee, to decide how to allocate this food
among community members.
The assumption behind community selection of beneficiaries
in pastoralist societies is that
their insider knowledge, combined with a traditional ethos of
sharing, will ensure better
targeting of the neediest people than outsiders can achieve
with crude proxy indicators and
limited local knowledge. However, community-based
targeting fails to take account of
power dynamics within the community, which can result in
elite capture of external
resources and the (deliberate or inadvertent) exclusion of
certain marginalised groups (e.g.
members of minority clans).
There were many complaints of bias and appropriation of
food aid by people within
communities who have responsibility for allocating it.
(When the food aid reaches the
elders they take half of it and the rest disappears: we dont
know where it goes.) In one
informal IDP camp on the outskirts of Gode town, a poor
family told us they have never
received any food aid, even though regular deliveries are
made to the camp: Only those
with relatives in the committee distributing food get food.13 4
Women in Doboweyn claimed
that the allocation of food aid within communities is done by

men, an arrangement which
they felt was unfair in several ways. First, women
understand better than men who are the
destitute and disabled in our community.13 5 Second, they accuse
those responsible for food
distribution with diverting much of it for their personal use,

and even selling some in
Doboweyn town. Third, these men were accused of favouring
members of their clan and
close friends in food aid distribution, thereby denying the
neediest who usually come from
less powerful clans.
15.2.3 From households to individuals

A final reason for the failure of food aid to reach individuals
who are the intended
beneficiaries (such as undernourished children) relates to
decisions taken within the
household about how this food is allocated. Such decisions
might reflect a rational
utilisation of household resources for the benefit of all family
members, or they could be
described as a misuse of scarce resources by powerful
household members that deprives
weaker household members of essential food.
Some women complained that their husbands squander food
aid by selling it (e.g. for khat),
but they blamed the distribution system for handing this food
to men, rather than to
women who are responsible for feeding the children.
Our problem is that the men always get the food aid.
When my husband gets it, he
sells half and uses the money to buy khat. It would be
better if the food is given to the
women, since food is our responsibility. There isnt a
single woman who would sell
food aid to buy khat. If she sells, it will be to buy other
food items for the home,
especially the children.1 36
134 Life history inter view, Sagarabuur IDP camp, Gode.

135 Womens focus group, Doboweyn.
136 A woman in rural Gashamo.
144
IDS
RESEA
RCH
REPOR
T 57
15.3 Receipts of safety-net transfers

Table 15.2 shows the proportion of households in each of our
11 districts that received food
aid in each of the last three years. In total, just over twothirds of households received some
food aid in 2004/5 (n=749/1,093 =68.5 per cent), Slightly
fewer numbers had received food
aid in 2003/4, and significantly fewer in 2002/3 (just over
one-third of households), which is
consistent with DPPC figures on trends in metric tons of food
aid allocated to Somali
Region (compare with Table 15.1).
Comparing across districts, urban residents of Jigjiga and
Gode are highly unlikely to receive
food aid (less than one in ten) and are much less likely to
receive food aid than households
in any rural community. This is an indicator of good outreach
to rural areas. Five out of nine
rural districts Shinile, Dolo Odo, rural Jigjiga, Cherati and
Doboweyn enjoyed almost
blanket coverage, with more than 90 per cent of households
getting some food aid in
2004/5. Surprisingly, both Gashamo and Kebribayah, despite
being worst affected by the
2004 drought, had lower than average coverage among rural
districts (where access to food
aid averaged 82 per cent). Kebribayah (43 per cent) was the
only rural district where less
than half the households surveyed received food aid in the
previous year.
In most cases, households that received food aid in the past
year (2004/5) had been given
two allocations, though some had only one and a few had
received three rations. In
previous years, too, each beneficiary received only one, two
or (at most) three allocations
over the year. This infrequency of delivery raises questions
about the effectiveness of the
food aid programme in Somali Region as a safety net
intervention, and how much impact
on household food security it can possibly achieve. Since the
trends in numbers of food aid
recipients year on year are fairly consistent across districts,
this suggests that this is a regular
general food distribution programme at the regional level,
rather than a flexible response to

local needs assessments that scale up or down as local
conditions change from year to year.
This conclusion accords with findings from a study of food
aid targeting in Ethiopia, which
found that the main determinant of food aid deliveries to a
region is whether that region
had received food aid the previous year (Clay, Molla and
Habtewold 1999) institutional
momentum and established pipelines dictate food aid
allocations more than localised
estimates of food deficits and household consumption needs.
Notwithstanding the possibility that food aid deliveries may
be relatively predictable and
regular, concerns about their limited effectiveness are
compounded by evidence on the
relatively small amounts of food aid actually received by our
sample households. The final
column in Table 15.2 shows the kilograms of staple cereal
(wheat) received by each
household each time they were given a food ration in 2004/5.
The mean of 14kg is very low
the median is close to 10kg and only in one district
(drought-affected Gashamo) does
the average exceed 20kg. Doubling the average ration to
allow for two rations per year
gives an annual food transfer per household of 28kg an
almost trivial amount, certainly
not enough for anyone to live on and not enough even to
make much impact on household
food security. This 28kg amounts to just 15 per cent of our
estimated allocation (from DPPC
data see Table 15.1) of 184kg per beneficiary in 2004.
Where did all this food aid go?
Two explanations can be suggested for this anomaly. The first
is that food aid is being
informally redistributed among a much wider group than the
targeted beneficiaries which
amounts to approximately 25 per cent of the regional
population (Table 15.1) while over
80 per cent of households in our rural survey reported
receiving food aid in 2004/5. In
many instances it appears that all food aid delivered to the
community was divided among
all households, rather than being allocated to the poorest members of the
community.1 37 As
a result, the rations that each household received were greatly

reduced. The second
explanation relates to problems of food aid distribution, as
described above: diversion of
food by officials, or appropriation of food by community
leaders. The effect of both is a
137 One problem with community-based targeting is that communities are

often reluctant to select the poorest
20 (say), and to exclude ever yone else. To avoid the social tensions
that this can provoke, communities
sometimes decide instead to divide external resources equally among
everyone.
145
Table 15.2 Household receipts of food aid, by district

District Received Received Received Amount (kg)
this year last year 2 years ago of food
(2004/5) (2003/4) (2002/3) received
(2004/5)
Pastoralist 84% 78% 42% 20.9
Gashamo 77% 72% 4% 29.2
Shilabo 78% 63% 40% 18.3
Shinile 97% 98% 82% 15.1
Agro-pastoral 75% 61% 28% 12.2
Kebribayah 43% 28% 8% 18.0
Doboweyn 90% 89% 44% 12.8
Cherati 93% 68% 33% 5.7
Farming 88% 83% 59% 10.5
Jigjiga rural 94% 89% 84% 10.1
Kelafo 75% 66% 31% 13.8
Dolo Odo 96% 94% 63% 7.7
Urban 7% 4% 5% 14.4
Jigjiga town 10% 6% 6% 13.8
Gode 3% 1% 3% 15.0
Total/Mean 68.5% 61.1% 35.9% 13.9
No. of households 749 668 392 749
reversal of the intended food aid targeting outcome: instead
of 26 per cent of the
population receiving 184kg each, our survey finds that 69 per
cent of the population
received just 28kg each.
The questionnaire also asked if the household had received a
range of other safety-net
transfers during the previous year, including: food-for-work,
cash-for-work, faffa
(supplementary food for young children) seeds and tools,
livestock (restocking), cash transfers
and microcredit. There were no positive responses on any of
these programmes, with two
exceptions: 11 respondents in Shinile reported participating
in a cash-for-work project (no
cash-for-work was mentioned in any other district), and
seven households in Gashamo and
four in Kebribayah reported receiving faffa, presumably as an

emergency response to the
recent drought in those two districts.
This suggests that safety nets or social protection
programming for Somali Region is
remarkably undiversified and unimaginative, being driven
almost exclusively by general
distributions of food aid. In highland Ethiopia food aid is also
dominant (though this is
changing with the introduction of cash transfers on the
Productive Safety Net Programme,
launched in January 2005), but the delivery instruments
include food-for-work completely
absent in Somali Region as well as gratuitous relief.
15.4 Uses of food aid

Although most food aid is reportedly consumed at home by
the beneficiaries and their families,
some of it is not, for one of two reasons: either it is not liked
(the food provided is not
preferred), or it is not needed (people have more pressing
priorities than food). In the first case,
families that need food aid but do not like what they are
given usually exchange it (by barter or
146
IDS
RESEA
RCH
REPOR
T 57
sale and purchase) for other food crops. A second reason for
selling food aid is to raise cash for
non-food expenses. Recipients often sell some or all of their
food aid to meet their cash needs
not all of which (e.g. khat) is necessarily in the best
interests of the household.
Much of the food aid that is sold is immediately converted
into preferred local cereals
maize or sorghum or other food items. (Because we dont
like the wheat, I sell it to get
money for other necessities like sugar.) Although this is a
rational and understandable way
of converting an unpopular food into an acceptable
alternative, because there is little
demand for qamadi1 38 (food aid wheat) the price is low and the
terms of trade with other
foods are unfavourable. One kilogram of qamadi typically

buys about two-thirds of a
kilogram of local wheat. Moreover, the price of food aid
fluctuates with place, demand,
supply and seasonality. In Warder in October 2004, a 50kg
bag of food aid was worth
120 Birr, but by December the selling price had fallen to only
35 Birr. This is therefore a
very inefficient way of transferring either food or income to
poor and vulnerable people.
Because qamadi is deeply unpopular, other beneficiaries
simply feed this wheat to their
animals, so that there is no direct impact on human food
consumption at all. We could not
eat it because it was dry and hard to eat. So we gave it to our
livestock.; Some people give it
to their goats. I dont have any goats, otherwise Id also give it to
the animals.1 39 Especially
during the dry season and in drought periods, food aid can
provide a useful source of feed for
livestock. A mature female camel (hasha) might be fed 34kg
of wheat a day. Though this
might seem an extravagant use of food intended for human
consumption, at this feeding rate
the camels will give milk, and it can sustain livestock through
the drought period. This use of
food aid to sustain animals through drought periods may also
point to the potential for a
market in subsidised hay and fodder. In districts like
Gashamo, in north-eastern Somali Region,

a brisk market already exists in the trucking of water for
livestock during the dry season.
15.5 Beneficiary tastes and preferences

The food aid distributed in Somali Region is mostly dry
wheat flour in 50kg bags, known by
locals as qamadi. As noted, most people in Somali Region
strongly dislike the taste of this
foreign wheat, and find it hard to eat. We dont get
nutritious food, only wheat flour
which is too dry for human consumption; I have a baby
who I cant feed wheat. There is a
perception that this wheat might be preferred by
highlanders, but is not appropriate for
Somalis: We dont like the wheat that we receive as food
aid. We are not Highlanders!
Maybe they like it, but we dont.1 4 0
Instead of qamadi, most people expressed a preference for

rice or pasta, or even local wheat.
They are also very dissatisfied with the lack of diversity of
food aid. (In fact, beans and cooking
oil are also distributed, but in smaller quantities and less
frequently than wheat flour.)
Instead of wheat, wed prefer other things like rice,
sugar, everything good like normal
people; maybe even biscuits occasionally! Just because
we are poor, doesnt mean we
dont like such food, or that we dont like a variety of foods.1 4 1
People need a variety of foods to live; the government

needs to know what foods
people need and like, and not just give them what they
think. We have never been
able to tell the government what we prefer.1 4 2
One woman went further, accusing the government of being

indifferent to the likes and
dislikes of the Somali people. On the other hand, it is true
that delivering less preferred
138 Qamadi literally means wheat, but has become a derogatory term
for imported food aid wheat.
142 A widow in rural Gashamo.
147
inferior foods during a food crisis can be an effective selftargeting mechanism. If the
really poor people are eating qamadi rather than selling it
or feeding it to their animals,
this is a sign that the transfer is reaching at least some of the
most vulnerable.
The really poor people are forced to eat qamadi even
though we Somalis dont like it.
But the government thinks that because we are poor, we
shouldnt get the food we
prefer, and we should be happy with whatever they give us.1 43
Some respondents are aware that food aid in Somaliland is

provided in the form of locally
preferred foods such as rice, pasta and oil, and they question
why they cannot be given the
same. Another popular request was for humanitarian
assistance to be provided in the form
of cash, as this is more flexible and allows beneficiaries
greater choice. (I would prefer to
receive other food like rice, or better still, to get cash.) This
opinion gives some support to
the shift in the current Productive Safety Net Programme,
away from food aid for the
predictably food insecure and towards unconditional cash
transfers. The best is to give
money so that people can decide themselves what to buy.
Also, with money people can
send their children to school, and they can also help others in need with loans, and so on.1 44
15.6 Alternatives to food aid

Many pastoralists, interviewed in the aftermath of the
drought of 2004, argued against
continued deliveries of emergency food aid, and in favour of
assistance that would usually
be classified as rehabilitation (e.g. restocking) or
developmental safety nets (e.g.
employment opportunities). What we need is help with
reconstructing the livelihood we
used to live not food aid. Restocking, veterinary facilities,
health care, employment
opportunities, education for our children thats what is important to us.14 5
The argument for restocking was made mainly by men.

Women often questioned the logic
of restocking as a safety net or rehabilitation intervention,
given the drought-proneness of
Somali Region, and argued instead for support to diversify
their livelihoods away from

pastoralism.
Maybe it is a waste of money and time to restock in this
region, when there are
always going to be droughts. Its like throwing money
away. Instead, we all, women
and men, need help with alternative activities so that we
can stop depending on
pastoralism only.14 6
Women and men have different needs. Im sure our men

will tell you we need restocking.
It is true that is important, but that is what the men need,
it isnt what women need. We
will tell you we need schools and a health centre and
credit. We would also like assistance
to start projects that will generate an income and get us markets.14 7
In farming and agro-pastoralist communities, complaints

about the inappropriateness of
food aid were equally common and vociferous, and requests
for alternative forms of
assistance were dominated by agricultural inputs and access
to markets. What we need is
farm tools, help with diggers to plough our lands, seeds and
fertilisers, and market for our
produce not food handouts!1 48
143
144
145
146
147
148
148
A woman in rural Gashamo.

A widow in rural Gashamo.
A male pastoralist in Gashamo.
A male agro-pastoralist in Kebribayah.
IDS
RESEA
RCH
REPOR
T 57
16 Health services
This chapter reviews survey evidence on causes of death in
Somali Region, and concludes that
the majority of these deaths are preventable. Failures of the
public health service are responsible
for thousands of premature deaths every year, and the limited
accessibility and poor quality of
health services, especially for people living in rural Somali
Region, are clearly demonstrated.
16.1 Causes of death in Somali Region

Mortality data from the household survey have been
discussed in earlier chapters, but here
we consider the causes of deaths reported by respondents.
The most common cause was
described as a short or long-term illness (58 per cent). It was
not possible in the survey to
identify the nature of fatal illness or disease. There is no way
of telling, for instance, how
many of these deaths from illness are AIDS-related. Illness
was followed by death in
childbirth (23 per cent), malnutrition (8 per cent) and old age
(5 per cent). Smaller numbers
of deaths were attributed to accidents, snake bite or attacks
by wild animals, famine
which respondents mentioned separately from malnutrition
and conflict, where conflictrelated deaths include two fatalities from landmine
explosions (Table 16.1).
If deaths due to old age or illness can be described as deaths
from natural causes while
recognising that a large proportion of deaths from illness are
preventable (and should have
been prevented) it follows that three in five deaths reported
(n=585/935 =62 per cent)
were natural. This leaves a strikingly high proportion of
deaths (n=354/935 =38 per cent)
that could be described as unnatural and premature (i.e.
deaths in childbirth, deaths from
malnutrition, famine, accidents and conflict), in addition to
deaths from preventable illness
and disease (Figure 16.1).
The rates of premature and preventable mortality are
shockingly high: according to this survey

evidence, five times as many people die in childbirth as die
peacefully in old age in Somali
Region. Deaths of mothers and babies in childbirth accounted
for almost a quarter of all
deaths reported (n=218/935 =23.2 per cent), and almost twothirds (n=218/354 =62 per
cent) of unnatural deaths those that were not attributed to
illness or old age. This is an
indictment of the limited availability and poor quality of
reproductive health services in Somali
Region. It also reflects the generally poor health and nutrition
status of women in the region,
which raises the risks of complications during pregnancy and
in labour. Some plausible reasons
for this state of affairs were provided by one midwife
interviewed in Gashamo town (Box
16.1). The most important factor mentioned is the inadequate
diet of women, especially
during pregnancy, which weakens themselves and their
babies, often fatally. Evidence from
Table 16.1 Causes of death in Somali Region
Cause of death Pastoral Agro- Farmer Urban Total
deaths
pastoral
Short illness 93 (32%) 104 (38%) 71 (27%) 33 (33%) 301
(32.2%)
Long illness 81 (28%) 60 (22%) 67 (25%) 31 (31%) 239
(25.6%)
Died in childbirth* 60 (21%) 69 (25%) 73 (27%) 16 (16%)
218 (23.3%)
Malnutrition 29 (10%) 8 (3%) 32 (12%) 3 (3%) 72 (7.7%)
Old age 9 (3%) 17 (6%) 12 (4%) 7 (7%) 45 (4.8%)
Accident 8 (3%) 9 (3%) 2 (1%) 9 (9%) 28 (3.0%)
Snake bite/wild animal 3 (1%) 6 (2%) 3 (1%) 0 (0%) 12
(1.3%)
Conflict 7 (2%) 4 (1%) 0 (0%) 0 (0%) 11 (1.2%)
Famine 2 (1%) 0 (0%) 7 (3%) 0 (0%) 9 (1.0%)
Total 292 (100%) 277 (100%) 267 (100%) 99 (100%) 935
(100%)
* Died in childbirth includes both mothers and babies.
149
Figure 16.1 Deaths from natural and unnatural causes, 1992

2004
40%
30%
Di ed i n chi ld bi rt h
S ho rt i ll ne ss
35%
Lo ng i ll ne ss
on /f ami ne
M aln ut r it i
25%
Acci de nt
Ol d age
30%
Ani mal s/ snak es
20%
25%
Co nf li ct
20%
15%
15%
10%
10%
5%
5%
0%
0%
P ast o ral Agro F armer Urban
Pa st ora l Ag ro
Farm er Urb an
famines elsewhere suggests the risks of miscarriages,

stillbirths and maternal mortality
increase during drought episodes, as maternal nutrition is
compromised. There is also
substantial evidence that undernourished mothers deliver low
birth-weight babies, and since
low birth-weight is a contributory cause of stunted physical
growth and impaired cognitive
development in later life, this contributes to intergenerational transmission of poverty. One
policy implication from this literature is that it is vital to
invest in the nutrition of adolescent
girls and young women, to break the cycle of mother-to-child
poverty.
Deaths from natural causes (illness and old age) are highest
in urban areas, while deaths
from unnatural causes are highest in rural areas. All deaths
due to snakes or wild animals,
conflict and famine, and 96 per cent of deaths attributed to
malnutrition, occurred in rural
areas (Table 16.1). Most of the famine deaths, and almost
half the deaths from malnutrition,
were recorded in farming communities, with a concentration
in Kelafo District, which was
badly affected by the drought-famine of 2000. Interestingly,
one-third of deaths caused by
accidents occurred in Gode and Jigjiga town, probably
reflecting the hazards of urban life,
such as motor cars and electricity.
16.2 Access to health services

In recent years the government of Ethiopia has invested
heavily in improving public services in
Somali Region, especially health and education facilities. To
date, though, urban residents still
enjoy better access to basic services, such as clinics and
schools, than do rural communities. This
finding is hardly surprising. All over the world, densely
settled urban communities tend to be
better provided with basic services, because large numbers of
people living in close proximity
allows for economies of scale and cost-effective delivery. It
can also be difficult to persuade
doctors, nurses and teachers who have usually been trained
in towns and cities to relocate to
rural areas which offer few of the comforts of urban living.
This issue has constrained the delivery
of public services, and the implementation of decentralisation
programmes, throughout Africa.
What is surprising is the extent to which rural residents of
Somali Region are deprived of access
to essential public services. In the case of health facilities, the
access gap between urban and
rural provision is almost a mirror image. While 96 per cent of
urban households in our survey
answered yes to the question Is there a health clinic in your
community? and only 4 per cent
answered no, in rural areas only 6 per cent answered yes
while 94 per cent said no (Table
16.2). Even the 6 per cent of affirmative replies are slightly
misleading, because the figure for
Shilabo is unusually high among rural districts, at 28 per
cent. In five out of nine rural districts,
100 per cent of respondents reported that there is no health
facility in their community at all. In
other rural districts, a clinic was found in one or two
communities of the many communities
visited. In Kelafo District, for instance, 16 per cent of
households have access to a clinic, but
elsewhere in the same district, a community discussion
revealed that a population of over
10,000 people in Afdub and neighbouring villages do not
have a single clinic. Participants in this
discussion highlighted this problem and made an urgent
appeal for a clinic, arguing that:
Because there is no health facility here, most people rely on religious healers.1 49
149 A woman in Kelafo District.
150
IDS
RESEA
RCH
REPOR
T 57
Box 16.1 Case study: midwife, Gashamo town

There are three midwives here, but Im the most
experienced. I have delivered 150
babies. The saddest thing is when the baby is delivered
stillborn. I have had some women
who have died in childbirth, and that is also very sad.
The main reason for women dying
like this is because they are too thin and work too hard
until they have the baby.
Especially in the bush, you see women who are very
thin, who have very little to eat, so
they arent strong enough to deliver the baby. Other
times, even if the baby is born alive,
it is so small when it is born because the mother hasnt
good and enough food during
her pregnancy. Sometimes, then, the baby also dies
because it is so under-nourished and
small. Then, the mother also doesnt have enough
breast milk to feed it, so it cant really
survive. Because there have been so many droughts in
the last few years, 95 per cent of
womens nutrition is very bad. They tell me they give
their share of any food to their
children, and only have a little bit. If not, they give
most of the food to their husbands.
We midwives dont charge for our services, but the
families do give us something
anything they can afford. Sometimes, instead of the
women giving me something, I
have to give them something, like food, because I can
see how poor she and her
family is. I can see if I dont give her something, she
will die, because she is so thin.
Source: Case study interview, midwife, Gashamo
town.
In rural Dolo Odo District, no communities visited have a
health centre of any kind, and people
complained about the lack of health-care professionals.
Malaria and tuberculosis were named
as the most prevalent diseases in these riverine farming
communities, followed by water-borne
diseases. According to a local key informant: Since clear and
clean water is not available, the

people are using river water. This reduces human productivity and quality of life.15 0
Table 16.2 Access to health facilities, by district

District Health clinic No health Distance to Immunised
in community (%) clinic (%) nearest
clinic (km)* children
Pastoralist 12 88 36 236 (24.4%)
Gashamo 8 92 n/a 57 (33.3%)
Shilabo 28 72 38 62 (12.4%)
Shinile 0 100 33 117 (27.5%)
Agro-pastoral 2 98 26 232 (19.6%)
Kebribayah 6 94 n/a 38 (18.9%)
Doboweyn 0 100 17 73 (15.6%)
Cherati 0 100 35 121 (24.4%)
Farming 5 95 14 298 (35.4%)
Jigjiga rural 0 100 1 94 (38.5%)
Kelafo 16 84 18 69 (31.8%)
Dolo Odo 0 100 13 135 (36.0%)
Urban 96 4 1 297 (49.4%)
Jigjiga town 95 5 1 132 (64.7%)
Gode 97 3 1 165 (34.0%)
Total 23% 77% 21km 1,063 (30.6%)
* Distance = median estimated distance in kilometres, if no
clinic in local community.
150 An elder in Dolo Odo District.
151
Box 16.2 Case study: midwife, Shilabo District

Tamad Buran is a 40-year-old woman with eight
children. Two of her sons are
boarders at schools in Kebridahar and Shilabo town.
Tamad lives with her husband
and their six other children in Lasoole, Shilabo
District. The children look after their
30 sheep and goats. Tamads father was a pastoralist,
but his animals were inherited
by his brother, leaving Tamad with nothing. Apart
from her domestic responsibilities,
Tamad works as a midwife and a tea-stall owner,
selling tea to pastoralists who come
to Lasoole to water their animals, and to travellers
going to Kebridahar and Shilabo.
Tamads mother was also a midwife.
Tamad holds a midwifery certificate from Kabridahar
hospital, having attended a threemonth course. The training covered the administration
of local anaesthetic, stitching,
and provision of blades and gloves to minimise risks.
However, Tamad uses no tools or
equipment in her service, except for sterilised gloves
that she gets for free from
Shilabo clinic. She gets no other assistance from the
government or development
agencies. Tamad does not charge for the service she
provides but people offer her
payment in kind, with gifts of cloth, sugar, or
sometimes a sheep or goat.
On average, Tamad helps deliver about 30 babies a
year. When performing deliveries,
complications sometimes arise babies die in
childbirth, or women bleed profusely
and lose dangerous amounts of blood. Women whose
lives are at risk are
transported by camel or vehicle (if one is available),
either to Kebridahar hospital,
about 70km from Lasoole, or to Shilabo clinic, 16km
away.
Source: Case study interview, midwife, Lasoole,
Shilabo District.
The questionnaire also asked respondents to estimate the
distance to the nearest health
facility. Although the actual distances given in Table 16.2
may not be accurate, there is a clear

gradient between livelihood systems. Pastoralist communities
claimed to be furthest from a
health clinic (on average, 36km), followed by agropastoralists (26km), then farmers (14km),
with urban households being closest to a clinic or hospital
(usually within 1km). In
communities that are located close to Somalia or Somaliland,
many households reported
going across the border for treatment, because these are the
closest health facilities available
to them and access is not restricted to local citizens. They
treat us nicely and we get cured.
The specific dangers for women of not having access to
reproductive health care were
identified in relation to deaths in childbirth, as discussed
above.
We dont have a health centre here. Many women die
when they have complications
in childbirth. Every few weeks, a woman from this area
has such complications and
some do die. The young girls see this and say they want
to live in town so they can
have access to hospital when they give birth.1 51
It is encouraging that many children in Somali Region have

been vaccinated, even where
health centres are not located in their vicinity. This is because
there are mobile vaccination
teams that tour the region, announcing in advance when they
will be visiting each
community or settlement, and mothers are generally keen to
have their children immunised.
There are mobile teams that visit the communities for
vaccinations, like National
Immunisation Day, and the campaign against polio. There is also
the Eye Mobile.1 52 Across
our 1,100 survey households, almost one in three children

have been vaccinated
(n=1,063/3,478 =30.6 per cent). Most of these children have
been vaccinated two or three
times, against polio, measles, diphtheria, pertussis and
tetanus (DPT), and tuberculosis (BCG
vaccine). On the other hand, immunisation coverage is far
from comprehensive, ranging
152 An agro-pastoralist parent in Kebribayah.
152
IDS
RESEA
RCH
REPOR
T 57
Figure 16.2 Perceived quality of health services, Somali

Region
100%
11%
V
e
r
y
P
o
o
r
80%
26%
P
o
o
r
71% 67%
60%
22%
A
d
e
q
u
at
e
40%
G
o
o
d
41% 14%
17%
20%
E
x
c
el
le
n
t
17%
14%
1% 0%
Pastoralist Agropastoralist Urban

from half of urban children (49.4 per cent) to just one in five
agro-pastoralist children (19.6
per cent) in our survey (Table 16.2). The two districts with
the lowest immunisation coverage
Shilabo (12.4 per cent) and Doboweyn (15.6 per cent) are
also the two districts with the

highest scores on our instability index (as discussed in
Chapter 10). This is significant
because it follows that other districts that are more insecure
(and were not covered by our
household survey, sometimes for this reason) probably have
even lower outreach from
mobile immunisation services.
Apart from physical accessibility, another constraint to
uptake of health services is their cost
(financial access). Vaccinations are given free of charge,
but seeking treatment at clinics is
expensive. According to respondents in Jigjiga town, there is
a 20 Birr consultation fee just to
go to the clinic. The cost of treatment or drugs varies
depending on the illness, but is rarely
less than 25 Birr and can be as high as 250 Birr. There is also
a range of informal health
service providers traditional healers, midwives,
circumcisers and Koranic healers whose
charges are variable, but are either lower than formal health
providers or at least negotiable.
Some households are resorting to traditional medicines
because they cannot afford the
price of drugs at clinics. (We are using traditional health
treatments because we cannot
afford to go to the clinic, because of financial problems.)
The recent droughts may well
have contributed to this process of reverting to less effective
but cheaper sources of health
care. This would only have compounded the health problems
associated with the drought,
however, and suggests a strong case for protecting access to
affordable health care at all
times, but especially in emergencies.
16.3 Quality of health services

The quality of health services in rural areas of Somali Region
is considered by local residents
to be much poorer than health facilities in urban areas. In the
pastoralist and agropastoralist communities of Gashamo and Kebribayah
Districts, over 80 per cent of
households interviewed rated local health services (where
these were available) as very
poor or poor, compared to 37 per cent of households in
Jigjiga town, where 42 per cent
of respondents described their health services as good or
excellent, against 14 per cent in
Gashamo and no households at all in Kebribayah (Figure 16.2).1 53
Although the provision of clinics, doctors and drugs appears

to be inadequate in both
quantity and quality throughout most of Somali Region, some
complementary health
153 Since this question was only asked in three district s during the first
leg of fieldwork, these findings should be
regarded as indicative rather than representative of Somali Region.
153
Box 16.3 Case study: HIV/AIDS womens group,

Gode town
Since the year 2000, when it became apparent that
several members of the local
community were displaying symptoms of HIV/AIDS,
womens groups have been
established in each kebele of Gode District. Seminars
were held to sensitise the
community and training was provided for the groups.
Among other things, the
groups organise seminars to raise awareness on the
causes and dangers of HIV/AIDS
to individuals, their families and society. They visit
schools and weddings, organise
plays and distribute information posters and free
condoms. They also provide a special
diet for those affected, if their families lack the means.
Orphaned children are
referred to SOS Childrens Homes. These groups
receive assistance from government,
private businesses and food aid distributors.
A discussion with one HIV/AIDS womens group in
Gode town revealed that the
spread of AIDS in Somali Region is believed to be
caused mainly by negative attitudes
and behaviour. The disease is seen as a taboo and is
rarely acknowledged by those
infected with HIV or their relatives, for fear of
stigmatisation. Initially, denial was also
based on a false perception: We used to think the
disease was common only among
non-Muslims. According to these women, awareness
is improving but so is HIVprevalence, mainly because of high levels of
promiscuity and resistance to condom
use. The message is getting across but still people are
not taking prevention
measures. Some of the affected do not disclose their
status and continue spreading
the disease.
Prostitution also spreads the disease, and the problem
of prostitution was blamed on
highlanders who have migrated to Gode, as well as
IDPs and refugees from
Somalia. The women also pointed out that the problem
of HIV/AIDS is most
common among young men aged 16 to 40. This group

consumes a great deal of khat,
a stimulant that reputedly induces feelings of euphoria
and sexual desire. Poverty and
recurrent droughts are also blamed for accelerating
HIV transmission indirectly. There
is so much poverty now as a result of successive
droughts, and this causes women to
engage in prostitution so as to make ends meet.
Source: Qualitative survey focus group discussion,
Gode town.
services seem to be performing rather better. A case in point
is the HIV/AIDS groups that
have been established by the governments HIV/AIDS
Secretariat and the Education Bureau,
with support from international NGOs. These groups aim to
raise awareness and provide
direct assistance to people affected by HIV and AIDS in
Somali Region (Box 16.3). Though
their coverage is limited to date, they are performing an
important function and are
expected to extend to other communities in the future.
17 Education
Pastoralist communities across the world face problems in
accessing education services,
partly because of their high mobility together with the low
population density of arid areas,
which makes the delivery of public services logistically
complex and more expensive than in
densely settled sedentary communities, but partly because
pastoralists tend to be politically
marginalised, and so have little power to influence
government spending allocations. Somali
Region in Ethiopia is no exception. This chapter first presents
evidence of poor education
outcomes, in the form of low self-reported literacy rates.
Next we examine the main
supply-side cause of low literacy rates, namely inadequate
provision of schools. Several
demand-side factors also determine education outcomes,
including parental attitudes
towards educating their children (which are significantly
gendered), costs of education, and
perceptions about the quality of education services provided.
154
IDS
RESEA
RCH
REPOR
T 57
Table 17.1 Literacy rates, by district and gender

District Gross literacy Literacy rate Male literacy
Female literacy
(%)* (%)** rate (%) rate (%)
Pastoralist 12.0 13.7 22.7 4.8
Gashamo 7.6 9.9 15.5 4.4
Shilabo 17.2 19.0 33.8 5.2
Shinile 10.7 13.0 20.7 4.9
Agro-pastoral 11.3 11.4 19.6 3.6
Kebribayah 9.2 8.1 12.4 3.8
Doboweyn 8.3 11.8 23.2 1.2
Cherati 15.9 14.3 23.1 6.3
Farming 10.4 12.5 19.8 5.1
Jigjiga rural 12.4 15.4 22.9 7.0
Kelafo 9.0 10.6 16.4 5.1
Dolo Odo 10.0 12.1 20.8 3.2
Urban 49.5 49.9 67.6 32.9
Jigjiga town 55.0 47.6 65.9 30.8
Gode 44.7 52.8 69.8 35.8
Total 18.2 19.5 29.5 9.8
* Gross literacy = percentage of individuals able to read and
write in total population.
** Literacy rate = percentage of individuals able to read and
write in population aged 15 and
above.
17.1 Literacy
The questionnaire survey included a self-reported question on
basic literacy for every
household member (Are you/he/she able to read and
write?). Out of the total population
of 8,817 people in the 1,200 surveyed households, 18.2 per
cent (n=1,605) were able to read
and write at the time of the survey, and 81.8 per cent
(n=7,212) could not. As might be
expected, literacy is much higher in urban areas than rural
areas. In Jigjiga town, over half of

the surveyed population (55 per cent) said that they can read
and write, while in Gode
town, the figure was just under half (45 per cent). In rural
districts, by contrast, literacy rates
are extremely low, with less than one person in ten able to
read and write in Gashamo (7.6
per cent), Doboweyn (8.3 per cent), Kelafo (9.0 per cent, and
Kebribayah (9.2 per cent)
(Table 17.1).
The prevalence of illiteracy is not significantly different
between districts that are dominated
by pastoralists (12 per cent literacy), agro-pastoralists (11 per
cent) and farmers (10 per cent).
It is intriguing that farming communities report such low
literacy rates, given the argument
that it is impossible to provide mobile groups of pastoralists
with schools and clinics. Clearly,
other factors must explain the inadequate provision of
education facilities to farmers in
Somali Region, among these being the marginalisation of
ethnic Bantu within the region.
As one official told us, people who live in distant localities
and have little political voice
demand less and complain less. In the July 2005 regional
budget debate, the problem of
unequal disbursement of public resources between woredas
was recognised, and steps were
taken to ensure that fair amounts were allocated to each
woreda.
Apart from the urbanrural divide, the most striking
differences in literacy rates are by
gender. On average across our sample, three times as many
adult males (29.5 per cent, or
almost one in three) as adult females (9.8 per cent, or less
than one in ten) can read and
155
write. In urban areas the ratio is two to one (68 per cent of
males and 33 per cent of
females over 15 years old), but in rural areas the ratio is
between four to one and five to
one (2023 per cent of males versus just 45 per cent of
females). Female literacy rates are
as low as 1.2 per cent in Doboweyn and 3.2 per cent in Dolo
Odo. The highest rates for
female literacy are just 6.3 per cent in Cherati and 7 per cent
in rural Jigjiga. While the low
literacy rates overall suggest problems with accessing
education throughout rural Somali
Region, the malefemale differentials suggest that limited
supply of education services is
compounded by limited demand for educating girls. Both
these supply failure and
demand failure factors are considered below.
17.2 Access to education

Several parents interviewed for this study expressed a strong
preference for educating their
children, and equally strong frustration at their inability to do
so, due to failures in the delivery
of education services. Our children are our assets and
educating them is our priority: but
where are the schools?15 4 It is certainly true that the federal and
regional government has
made efforts in recent years to extend access to education

services in Somali Region, including
building more schools, experimenting with mobile education
services, and supporting NGO
efforts to provide alternative basic education. Mobile schools
and NGO initiatives are
important in pastoralist areas, given the difficulties that the
Ministry of Education faces in
persuading teachers to live in remote communities. Low
education levels are also selfreinforcing: where illiteracy rates are high the pool of
potential teachers is very small.
Many children throughout rural Somali Region are sent to
Koranic schools instead of to
formal schools, either because parents choose this form of
education for their children or,
more often than not, because no other choice is available.
(There is no school here, only
one Koranic school.) In Kelafo District, a womens focus
group told us that they would send
their daughters to school if there was one, but since the only
education accessible to this

community is provided by one Koranic teacher, they send
their sons and daughters to him
instead. A typical village visited in Shilabo District has no
school except for the informal
Koranic school, where enrolment is over a hundred students.
Some Koranic schools do
teach other subjects apart from Arabic and the Koran. Of
course, since it is an Islamic duty
to teach children the Koran, many parents continue sending
their children to Koranic
teachers (e.g. as evening classes, or as a pre-school for very
young children) even where
formal schools are available.
In some communities, school buildings have been
constructed but the school is not
functioning because of a lack of teachers. In rural Gashamo
these are known jokingly as
hyena schools because wild animals live in the empty
buildings at night. (Only hyenas
study there!) In places where schools are functioning,
parents complain about shortages of
qualified teachers. Even in Gode town, where there are
several formal schools providing
better quality education than in rural areas, lack of specialist
teachers means that certain
subjects are not taught (see Box 17.1).
Some participants in a community discussion in Shilabo
District send their children to live
with relatives in urban areas, specifically so that they can
attend school. This is not an
unusual way of ensuring access to formal education for
children, but it does depend on
having relatives based in town who are willing to take in
children for extended periods.
Another alternative to building schools throughout rural
Somali Region is to provide
boarding schools to which children can be sent, while their
parents continue to pursue their
pastoralist lifestyle. Parents are often more reluctant to risk
sending their daughters to
boarding school than their sons (as discussed below), and this
may partly explain the gender
differentials in education participation and outcomes reported
above. Another reservation is
that boarding schools are more expensive than day schools,
because boarders must be
provided with accommodation and meals.
154 An agro-pastoralist parent in Kebribayah.
156
IDS
RESEA
RCH
REPOR
T 57
Box 17.1 Case study: school headmaster, Gode town

This school in Gode town is 30 years old. It was first
constructed as an intermediate
school, but ten years ago a new wing was added that
serves as a high school. There
are currently 1,252 students, only 220 (17.6 per cent)
of whom are girls. Most of the
students come from pastoral backgrounds, followed by
agro-pastoral families and
traders. Only 5 per cent of the students are not
Somalis. Enrolment has been rising,
but there is a bias towards boys (a 20 per cent rise in
recent years) rather than girls
(a 5 per cent increase). Increasing enrolment rates
reflect the communitys growing
recognition that education is empowering people at all
levels. They appreciate it
when they see an ethnic Somali administrator, doctor,
nurse, teacher or bureau
head. The headmaster explained the low girl
enrolment in terms of Somali cultural
practices that include early marriage of daughters and
parents prioritising the
education of sons. The school is trying to redress this
imbalance through awarenessraising to highlight the benefits that derive from
female education. The headmaster
believes that societys attitudes towards educating girls
are changing, which explains
why female enrolment is increasing, albeit at a slower
rate.
There are 35 teachers at this school all men.
According to the headmaster, female
trainee teachers typically fail to attain the minimum
qualifications required to teach.
Also, to qualify to train, one must possess at least a
high school certificate, and there
are fewer Somali girls than boys graduating from high
school, so the pool of
potential female teachers is much smaller. Training is
conducted at the Teacher
Training Institute in Jigjiga and the Alemaya
University Correspondence School.
Because of generally low levels of education in Somali

Region, the school faces
teacher shortages and some subjects have no teaching
staff at all. Subjects taught at
the school include Somali, English, Amharic, Physics,
Chemistry, Biology,
Mathematics, Geography, History and Civic
Education.
Shortages of teachers are compounded by a heav y and
growing demand for places at
the school.
We are overwhelmed by applications for enrolment
and we have already gone
overboard as most classes accommodate three times
the usual number. There are
shortages of teachers, classrooms, lack of science
laboratory, text books, library you
name it!
Source: Key informant interview, Gode town.
The nearest school is in Gashamo town. There arent
any children from here in school.
If we want them to go to school, we will have to send
them to boarding school. If
there was a school nearby, we could send them because
they would be staying at
home, so we wont have all the extra expenses of boarding school.1 5 5
The shortage of schools in rural Somali Region may be

understandable, given the low
density and high mobility of the regions population, but it is
a source of many complaints.
Interestingly, some communities with no formal education
services are taking the initiative
and mobilising resources for education from among their
members, relatives abroad, or by
lobbying NGOs. During fieldwork for this project, for
example, members of the study team
attended the inauguration ceremony of a new Ugaz (religious
leader) in Guradamole
District, which brought together families from seven subclans as well as NGOs and
dignitaries including the local Member of Parliament.
Community leaders took the
opportunity of this unusual gathering to hold a public fundraising drive towards the
construction of a primary school building for the community.
The argument was that once
the building was in place entirely paid for by the
community itself the regional
157
government would be shamed into sending teachers. The

public pledges that were made in
the next hour amounted to several camels, cows, sheep and
goats, plus more than a
thousand Birr and over a hundred dollars.
In another case, a primary school in rural Gashamo has been
established and is running
entirely through donor and NGO support. The Japanese
government paid for the building
costs, the British government has sponsored the furniture,
and Hope for the Horn is
subsidising the schools running costs. The headmaster of the
school has ambitions of
opening a secondary school in the same community, by
calling on remittances from the
Somali diaspora: We receive no money from the Ethiopian
government. We have launched
an appeal to Somalis abroad for funds for the secondary
school we hope to open here in
September 2005.15 6
17.3 Costs of education

Costs of education are relatively low in Somali Region, but
are higher for formal schooling
than Koranic schools. Government schools are not supposed
to charge fees, but parents in
Jigjiga town noted that they have to pay several indirect costs
to send a child to school
registration fee, uniform, books while the Koranic school
charges only a nominal tuition
fee. The estimated cost of sending a child to primary school
in Jigjiga is 125175 Birr per
term, while Koranic school costs just 15 Birr. Even where no
fees are required, many
parents can not afford to send their children to school,
because the non-fee costs are too
high. In one Kebribayah household with three children under
12 years old, none of the
children was attending the local formal school, because of
lack of money to buy the
uniforms and educational materials our children need.
In rural communities where cash incomes are lower than in
town, even modest expenses can
be prohibitive, and can result in children being deprived of a
formal education. I would like to
send all my children to school, but cant afford to.15 7 It is only
lack of money that prevents
me from sending my children to school. In some places local
NGOs have established nongovernment schools which do charge fees, though these are
often subsidised by the NGO.
School fees are 195 Birr, of which the student actually pays
135 Birr and Hope for the Horn
pays the balance. The parents also have to pay for their school uniforms.15 8
One parent in rural Gashamo argued that pastoralists need

free boarding schools to be
provided, since the costs of boarding can exclude children
from pastoralist families getting
an education. (As discussed below, there are also reasons, not
related to cost, why boarding
tends to exclude girls more than boys.)
We need schooling to be free. We need to be able to
educate our children, but at the
moment the fees and different costs make it impossible
for us to do so. We need
boarding schools which are free so that children not living in Gashamo can also attend.15 9
Koranic teachers charge variable fees, and often parents

make voluntary contributions rather
than paying fixed fees. Contributions can be made in cash or
in kind, and are often
differentiated according to ability to pay wealthier families
offer larger contributions.
Koranic teachers are also more flexible than formal schools,
and are typically sympathetic to
conditions of hardship that might afflict their clients. In the
year before the drought of
2004, for instance, it cost families in one Gashamo
community 20,000 Somali Shillings to
send a child to the local Koranic school. During the drought
they were excused having to
pay fees, because of the general hardship caused by the
drought. (At this time people dont
156
157
158
159
158
Primary school principal, rural Gashamo.

Male pastoralist in rural Gashamo.
Primary school principal, rural Gashamo.
IDS
RESEA
RCH
REPOR
T 57
Box 17.2 Case study: Koranic teacher, Shilabo District

Mohamed is a 40-year-old man who runs the only
Koranic school in Lasoole village,
Shilabo District. He has two wives but no children.
Mohamed started teaching the
Koran 20 years ago. His father was a pastoralist and
after he died Mohamed inherited
seven camels and ten shoats. However, all these
animals perished during the drought
of 1992, and Mohamed decided to look for something
else to do outside pastoralism.
After considering various options, he chose teaching
because, unlike pastoralism, it
has no risks. He also claimed that he did not want to
go around the community
looking for assistance, as he preferred to look after
himself.
Mohamed currently teaches 110 students, of whom 70
are boys and 40 are girls. He
charges 4 Birr per student per week and 30 Birr per
student after completion. Most
students take three years to master the Koran.
Mohamed recognises that it is difficult
for students to gain meaningful employment only with
Koranic education and he hopes
that formal, secular education will also be provided, so
that children can have a better
chance in life. As things stand now, the only
opportunity open to most local children is
to look after their parents livestock after completing
the course at his school.
Source: Qualitative survey interview, Shilabo District.
pay anything.) In another Gashamo community a family that
contributed one goat to the
Koranic school per child each year as a payment in kind
withdrew their children in 2004,
after the drought decimated their livestock.
17.4 Attitudes to education

The education deficit in Somali Region has a demand-side
as well as a supply-side
explanation. Limited access to formal education is matched

by ambivalent attitudes to
educating children. Part of this ambivalence reflects
scepticism about the value of
education, but part of it is socio-cultural. Attitudes to
education in Somali Region are highly
gendered, in at least three ways. First, as a generalisation,
women see the benefits of
education more clearly and more positively than do men.
(Somali women want their
children to be educated!) Second, while mothers offer
persuasive arguments for educating
their daughters, fathers tend to believe that there is more
value in educating sons. Third,
almost all parents are reluctant to send their daughters to
boarding schools, because of
concerns for their physical safety.
Some Somali men appear to be indifferent to the idea of
sending their children to school. I
dont believe in educating my children. Education isnt going to
help them.1 60 One woman
who is sending all five of her children to school told us that

she is personally paying for all
the costs: When I asked my husband to help with school
fees, he said: Why should they go
to school? Why bother sending the girls to school when they
will get married into another
family? We are pastoralists; we didnt go to school, why should our children?1 61
Somali women recognise the need to change attitudes, not

only to encourage the
government to provide more schools, but also to promote a
more positive attitude to
educating children among their husbands, brothers and
fathers. Its not only the Ethiopian
government that is the problem; its also our own men who
dont see the need for
education.16 2 Women argue that Somali men dont see the value
of education because they
see the future as being like the past looking after animals
for which literacy is
unnecessary. Men are stubborn and cant move with the changing
times.1 6 3 A high school
160
161
162
163
A pastoralist man, displaced by drought to Gashamo town.

A woman in Gashamo town.
A woman in Gashamo town.
Female store-owner, Gashamo town.
159
pupil in Gashamo town said: Some of our parents dont see

the benefits of education,
because they have seen people with education still being
unemployed.
Conversely, many Somali women see the future for their
children as lying outside of pastoralism,
and for this, they regard education as essential. In the future,
it is only education that will keep
people out of poverty.16 4 In a focus group discussion in Shinile
District, women mentioned
several reasons as to why education is increasingly important

in the contemporary world.
Education empowers people and gives them more
options in life.
Unlike the drought which can finish all our livestock,
nothing can take away what has
been learnt.
These days, you need to be able to read and count. The
world has changed its no
place for an illiterate person.
Children should go to school to learn the English
language, which is important to
know these days.
We need a school so we can send some of our children
to school and they can take
care of us in our old age if they are educated and have
jobs.
Life is easier for educated people.
Elsewhere, pastoralist women made the argument that
education is important for helping
people to be better pastoralists. Education is very good for
any human being, including
pastoralists. Education can even help them to be better
pastoralists to understand better
about weather and animals.1 65 However, the impacts of recent
droughts on rural livelihoods
appears to be causing a rethink. The immediate impacts of

the 2004 drought on education
were negative teachers and children both failed to attend
school, at least temporarily.
Since the drought, teachers cannot teach due to hunger, and
students cannot attend
school, also because of hunger.1 6 6 On the other hand, the
obvious and arguably increasing
risks associated with pastoralism have shifted parental

attitudes in favour of education, as
they recognise that educated people have more diverse
livelihood options. Education is

perceived as a means of improving access to sources of
income outside of pastoralism that
are unlikely to be affected by drought.
Previously, if someone asked me to send my child to
school, I would think that person
is insane. Now, my attitude has completely changed,
because education can sustain
better than pastoralism.1 67
These days, it is better if the women have some way of

earning an income to assist
the family in times of drought. But its even better if
they have an education and a job
that doesnt depend at all on the drought. Thats the only
way that people are going
to survive in the future.1 68
Children themselves expressed a preference for going to

school, rather than herding livestock.
I dont go to school, but I would like to. I would like to
be a teacher. I dont want to
be a pastoralist because the animals always die easily
when theres a drought. Its a
very hard life, and I dont like it. I prefer to live in town,
not in a village. My friends are
164 Female store-owner, Gashamo town.
166 Focus group participant, Shinile District.
167 Female focus group participant, Maromadobes IDP camp, Shinile
District.
168 Female store-owner, Gashamo town.
160
IDS
RESEA
RCH
REPOR
T 57
the same like me. None of us like tending the cattle and
walking in the sun all the
time. Children who go to school are always clean and
tidy, not dirty and ragged like us
who look after cattle. I want to be able to read and write
so that I can read
newspapers and travel to far places.1 6 9
When asked about their parents attitudes to education,

female high school pupils
participating in a focus group discussion mentioned several
reasons why Somali parents
might be reluctant to educate their children, especially their daughters.17 0
Most parents believe that its not important and

necessary to send girls to school
because they will leave the family when they marry.
If my father was still alive, as his daughter I wouldnt
have been sent to school. He was
a traditionalist, and many Somali men are
traditionalists. If our mothers made the
decisions and had the means, I believe all of them
would send us to school.
In contrast to the conservative attitudes of many fathers, most
mothers seem determined to
educate their daughters, recognising that education will
provide the next generation of
women with opportunities and independence that they
themselves did not enjoy. Several
participants in a womens focus group in Gashamo town
confirmed that they had made the
decision to send their daughters or grand-daughters to school,
often without either
financial or moral support from their husbands.
Although I would like all my grandchildren to be
educated, I can only afford to send
the eldest girl to school. So, Inshallah, one day if she is
a divorcee like me or a widow,
shell be able to manage without a husband, not like
me!
These days, people need education more than they need
animals. I would prefer my
daughter to marry an educated man and live in a house
in a city instead of marrying a

man with 100 camels, but be living and moving in the
bush.
It is better to educate a girl than a boy because
daughters will help you forever
because they are more sympathetic than boys. When
girls get married, they do leave
home and go to their husbands home, but they continue
to support their mothers
whenever they can. When a boy gets married, he forgets
that he has a mother and
only concentrates on his wife and his own children. So,
because I dont have enough
money to send all my children to school, I decided to
send my two daughters to
school while my sons stay at home.
One constraint to female education is the necessity for most
Somali schoolchildren to
become boarders, especially at secondary school level,
because of the limited number of
schools and the mobile lifestyle of pastoralists. Many parents
both mothers and fathers
expressed concerns about the vulnerability of their daughters
to sexual predation by male
pupils and teachers in a co-educational school system, and
especially in boarding schools,
which parents do not regard as secure environments for
teenage girls. One secondary
school in Gashamo town has 120 students, only 11 (9.2 per
cent) of whom are girls.
According to the principal, girls usually attend primary
school up to fourth grade, at which
point their parents often withdraw them. Because its mainly
a boarding school, they fear
for their daughters safety, although we try to reassure them.17 1
Some parents suggested
that single-sex schools should be provided for girls.
169 A 10-year-old boy in rural Gashamo.

170 Female high school pupils, rural Gashamo.
171 Key informant interview, school principal, Shilabo District.
161
Figure 17.1 Perceived quality of education services, Somali

Region
100%
9%
Very
Poor
22%
80%
50% 56%
Poor
60%
23%
Adeq
uate
40%
25% 19%
Good
43%
13% 20%
25%
Excell
ent
13%
2%
0%
Pastoralist Agropastoralist Urban

You must please report that we Somali women are not
against education! We just
fear for our daughters safety when they are adolescents.
Thats why we often send
our daughters only to the lower classes, and withdraw
them when they reach
adolescence.1 72
Many parents are scared for the security of their

daughters, especially if they are in a
boarding school away from home. If girls didnt have to
board, more parents would
send their daughters to school.
We would like schools for girls only. If the school is for
girls, then we wont even be
bothered if its a boarding school. We will send them gladly, knowing they are safe.1 73
17.5 Quality of education services

Another factor that influences the uptake of public services is
the perceived quality of those
services. Where schools are not regarded as providing a good
and relevant basic education,
parents will not pay money and lose the contribution that
children in Somali Region make
to domestic work (household chores, herding animals), by
sending their children to school.

According to our respondents, the quality of formal education
services varies greatly between
rural and urban communities. In Jigjiga town, almost half of
our interviewees (45 per cent)
described education services as good or excellent, but in
pastoralist Gashamo District only
13 per cent held this positive view, and none at all in agro-pastoralist Kebribayah (Figure 17.1).174
On the other hand, 75 per cent of respondents in Gashamo

and Kebribayah perceive the
education services that are accessible to their children as
poor or very poor, whereas only
31 per cent of respondents in Jigjiga town chose these
pejorative descriptions.
Because of this negative perception of education quality in
rural areas, parents might choose
not to educate their children, even in those rural communities
where schools are accessible.
The lesson for policy-makers is that bridging the education
deficit in Somali Region has two
aspects providing more schools, and improving the quality
of education services.

174 Since this question was only asked in three districts during the first
leg of fieldwork, these findings should be
regarded as indicative rather than representative of Somali Region.
162
IDS
RESEA
RCH
REPOR
T 57
Section 5 Conclusions
and
implications for policy
In the two concluding chapters of this report, we start by
presenting findings from our
research on how Ethiopian Somalis themselves see the future,
specifically concerning their
livelihood options to continue to pursue pastoralism as in
the past, or to abandon
pastoralism and sedentarise or urbanise. The final chapter
explores options for policymakers, not in narrow prescriptive terms, but in terms of
broad directions which
developmental policies for Somali Region could follow.
18 Somali attitudes to
the future of
pastoralism
The severity and frequency of recent droughts have exposed
deep divisions about the future
of pastoralism in Somali Region, along both gender and
generational lines. In general,
Somali men remain strongly committed to a future in
pastoralism, even if they have lost all
their animals, while women and children are increasingly
attracted to alternative lifestyles
that are urban-based and more settled. Men can see no
alternative to keeping animals it
is all they know how to do and it is intrinsic to Somali
culture. (We are people of
animals.) Women, on the other hand, see pastoralism as a
hard life that generates little
reward for them: pastoralism is not Somali culture; it is
mens culture. Children speak of
living in the twenty-first century, with telephones and

computers, not like our ancestors.
These fundamentally opposing views about future trajectories
for rural people in Somali
Region are the source of some tensions. Men accuse women
who do not share their
commitment to the pastoral way of life as betraying the
culture. A focus group discussion
with men in Gashamo town was very revealing of mens
attitudes in this respect (Box 18.1).
When invited to comment on the fact that women are
increasingly expressing a preference
for a sedentary life in town over a mobile life in the bush,
men were dismissive of these
sentiments and often made insulting remarks about such
women: Women want to settle
because they are not as strong as men; Women these days
want a comfortable life with
luxuries. They have become softer than women in the olden
days; Young women are lazier
than the women in the past.1 75
Many women, on the other hand, are deeply sceptical about

the future of pastoralism in
Somali Region. Dozens of Somali women who were
interviewed or participated in group
discussions during our fieldwork expressed dissatisfaction
with the pastoralist way of life,
and a preference for settling down in town. We are tired of
living like our animals. It is no
way for human beings to live. Several factors account for
this. One point that must be reemphasised, however, is that many of our focus group
discussions with women were held in
Gashamo, soon after the severe drought of 2004 had left
many rural families destitute and
displaced them to town. In this unusual and highly stressful
context, it is hardly surprising
that these women were disillusioned with a livelihood system
that appears to be failing.
Women elsewhere in Somali Region, especially in
communities unaffected by recent
droughts, did not express such strongly negative opinions.
Women in Gashamo are concerned that pastoralism is
increasingly unviable, and they have
seen the possibility of a better life easier, settled, less risky,
healthier for themselves and
their children in urban areas. In a group discussion on this
subject, women revealed
themselves as thinking creatively about options for the future
(Box 18.2). They also asserted
175 Male focus group participant s, Gashamo.
163
Box 18.1 People of animals? Somali mens attitudes

to the future of pastoralism
Man no. 1: All of us here are pastoralists, although
you are finding us all here in town
and not out in the bush with our animals. But that is
just because of our problems
with our animals. I have lost more than 100 animals
over the last years and droughts. I
only have a few animals left, and my eldest sons are
taking care of them in the bush.
Man no. 2: I also lost many animals in the last
drought and now. Soon, I wont
have anything to give to my sons.
Man no. 3: My sons arent interested in keeping
animals. I have to force them to go
and look after them. They are more interested in
schooling, but I cant afford to send
them.
Man no. 4: These days, most of the boys arent
interested in being pastoralists. When
I was a boy, all I wanted was to own my own animals
like my father and uncles. My
sons are not like that. I know they want to live in town
and arent interested in
anything to do with livestock. Its just because they
have no alternatives and I force
them to do it. Otherwise, they would simply leave and
go to a big town like Hargeisa.
Man no. 5: The droughts are making it difficult, but
I dont think we should leave it
all behind. My son wants to carry on this way of life
even though its difficult. Its all
that he knows.
Man no. 6: We are people of animals. As Somalis,
we cant live in towns and cities.
We live for our animals. We wouldnt agree to it.
Man no. 7: A Somali man will do everything to
protect his animals and his livelihood.
We dont want to live in towns. It is in our blood to
move with our animals when we
need to.
Man no. 8: Women may want to settle because they
want to trade in towns.
Pastoralism is a difficult life, but it is our culture.
Maybe men are more faithful to the
culture than women.
Source: Mens focus group, Gashamo town.

that men are more resistant to change, and less adaptable.
Several women challenged the
view that Somali culture is necessarily and inextricably
linked to rearing livestock. (There are
Somalis all over the world and theyre not pastoralists.)
Some women even suggested that
pastoralism should be seen as part of mens culture rather
than Somali culture. More
generally, these women rejected the suggestion expressed by
many men that Somali culture
is invariant over time, and that there are no alternatives to a
pastoral way of life. In other
places, like Hargeisa, I have seen Somali women driving
cars. That is a sign of freedom. It
also shows us that Somali women in different places behave
differently and that the Somali
culture is different in different places.17 6
Although most women in Gashamo expressed a preference

for a settled way of life, which
was perceived as more comfortable, less risky and less hard
work, three distinct positions
could be identified: (1) complete sedentarisation; (2)
spreading risks by dividing the family
between urban and rural livelihoods; (3) remaining in
pastoralism. Very few women chose
the third option. Conversely, strongly positive views were
expressed in favour of
sedentarisation, especially moving to towns which were seen
as holding many attractions
and no disadvantages: People never starve in towns, only in
places like this where we are
forgotten. Retaining a rural way of life but educating some
children was advocated as a
compromise between these two extremes.
One powerful argument for sedentarisation is that people who
are settled and live in towns
have a better standard of living and a more comfortable life.
According to one woman who
164
IDS
RESEA
RCH
REPOR
T 57
Box 18.2 No good future in pastoralism? Somali

womens attitudes to the
future of pastoralism
Woman no. 1: Because of the many droughts these
days, its not easy to be a
pastoralist. Now, the droughts are more severe than in
the past.
Woman no. 2: But the men dont understand this in
the same way: they are too
stubborn. They want to carry on despite the
difficulties. But because it is us women
who are suffering and not able to send our children to
school, we are the ones who
see there is no good future in pastoralism.
Woman no. 3: If droughts continue like this, you
wont find pastoralists in the future,
although, of course, we wish we could continue to be
pastoralists. But the
circumstances are changing and making it difficult to
continue.
Woman no. 4: We have all grown up as pastoralists,
so we know that life. There are
things that we like about it when things are going
well and you can sell your
animals but there are also many difficulties with that
life even when there are no
droughts. I remember when I was a small girl, I used to
walk far with my family and
our animals for many weeks at a time. Sometimes,
we never had enough food and
water; sometimes we had to drink filthy water. We
were always tired; it was very
dusty. There wasnt any chance of attending school or
clinics. It was a hard life, even
where werent droughts. So, if I had to choose that for
my sons or education, then I
wouldnt choose pastoralism.
Woman no. 5: As Somalis, its our culture, like
Islam, to be pastoralists. Without it,
we are losing a part of our culture. So, it would be
better to educate some of our
children, and send others with the animals. In that way,
we can continue with our
traditions, but can also change to meet the changing

times with those of our
children that are educated.
Woman no. 6: I dont think we should leave
pastoralism completely, but we need to
realise that it will decrease in the future because of the
land and droughts. So, we
need to also have alternatives like work and
education.
Woman no. 7: The children these days dont have an
interest in livestock. Thats the
other problem: we cant force them to be pastoralists.
In the past, all boys wanted to
have many animals; now, they want jobs and dont
want to live out in the bush
anymore. Maybe they want animals, but they dont
want to care for them
themselves: they think of employing someone to look
after the animals while they
live in town while having a good job.
Woman no. 8: For girls, there isnt a life in
pastoralism. We dont own the animals,
but work very hard. And then, if you divorce like me,
you are left with nothing. You
did all the work for your husband and the livestock,
but youre left with nothing. He
wont give you a single animal to take care of his own
children. So, I think its better
for girls to be educated and get jobs and forget about
pastoralism.
Woman no. 9: When we talk of pastoralism being
part of Somali culture, its actually
part of mens culture. We get nothing out of it. So, I
agree that the men will try and
carry on with it, but the women are losing interest.
Source: Womens focus group, Gashamo District.
visited her sister in Hargeisa: We had a bathroom, toilet and
water in the house. We had
electricity. It was such a nice and easy life. I would like to
live in one place like that.
A second powerful motivation for settling down is to access
basic services, especially health
and education, that are largely inaccessible in rural Somali
Region. Many women recognise
the difficulties that government faces in providing services in
isolated mobile communities,
and suggested that they need to choose between living
without services in the bush, or
settling down in order to access these services.
165
In the past, we didnt know about these services and we

didnt need them or want
them. Thats why we were not bothered to settle down.
Now because we know
about these things, some of us want them. So, we have
to make the decision to
either continue with pastoralism or settle in one place.
We Somalis live all over the place. Some of us live in
the bush, others are walking
looking for grazing. Its not possible for the government
to provide us with these
services all over the Region. We are too scattered. If
people want access to these
services, then they will have to settle in one place.
Other women who are committed to continuing in
pastoralism were less willing to accept
that they should abandon their way of life simply to access
essential services, and felt that
the government has a responsibility to make schools and
clinics more accessible in rural
areas. For myself, I prefer pastoralism to going to towns. We
want to continue being
pastoralists but we also want decent facilities like health
centres and schools. We dont
want to leave pastoralism just to gain access to these
services.
The case for livelihood diversification was made mainly on
risk-spreading grounds: that the
high and possibly rising riskiness of livestock-based
livelihoods is making it increasingly
necessary to divide our families between those who continue
with our livestock and those
who will be sent to school and will settle in one place and
find work. The advantages of
having family members earning income outside pastoralism
were expressed by several
respondents:
Its better to divide our children between different
activities between pastoralism
and education and jobs.
We are still pastoralists. So, I think we should send
some of our children to school and
some to look after the animals. In that way, our culture
can survive while, at the same
time, some are being educated and getting work and
sending money to help us. In
that way, we are spreading our risks.

I have six children and decided to send three for
livestock-rearing and three for
education.
I would like to send one son to school in Gashamo.
With Allahs help, he will find a
job one day and help us. The rest of us will take care of
our livestock. If the droughts
continue, then he will be able to rescue the rest of us.
Asked to vote on whether they preferred to continue with
their mobile pastoralist way of
life or settle down and live in town, all 11 women in one
focus group and all except one
very old woman in a community discussion with 18 women
(both in Gashamo) voted for
being settled. The old woman said: I want to die and be
buried here. A more typical
response was this one: We wouldnt mind saying goodbye to
the animals and the droughts
and hardships.1 77
In one settlement in Gashamo, a community meeting

generated an interesting debate
between men and women, which highlighted some
fundamental points of disagreement
(Box 18.3).
Ultimately, the future of pastoralism depends not just on the
attitudes of people who are
currently practising pastoralism as their livelihood system,
but even more crucially on the
attitudes and aspirations of future generations. Many women
pointed out that their children
and grandchildren are on the verge of rejecting pastoralism as
a way of life, because of a
combination of push factors (the increasing difficulties of
pursuing a livestock-based
livelihood in Somali Region) and pull factors (the lure of
urban living): It is difficult to find
children these days that are interested in pastoralism, even
boys. My own grandsons arent
177 Womens focus group participants, rural Gashamo.
166
IDS
RESEA
RCH
REPOR
T 57
Box 18.3 Better to have an education than a hundred

camels? A gendered
debate on the future of pastoralism in Somali
Region
Woman no. 1: You can lose your animals, as a
pastoralist, but you can never lose your
education. If we were educated, do you think we
would be sitting here and suffering
like this? Youll never see an educated person suffering
like this. Its better to have an
education than a hundred camels.
Man no. 1: Men will never agree with that. For us,
there is nothing better or more
valuable than having many camels and livestock.
Without our animals, we are not
Somalis. Our animals are an important part of our
culture and our way of life. Even
those of us who are farming still have animals because
we cant give up our animals.
Woman no. 2: But what about the droughts? If you
had education plus animals, if you
lost your animals, you could still use your education to
find a job or be a big trader. Do
you want your sons to struggle like you? Do you want
them only to know animals?
Man no. 2: No, I wouldnt bother to send my
children to school even if there was
one in the village. Instead, I would teach them
everything about our animals and how
to avoid the bad effects of droughts.
Woman no. 3: I think all the women of this village
will agree with us and send their
children to school. Most of the men will agree with
him because they are all the
same and only believe in their animals.
Man no. 1: We cant do anything about droughts
because they are sent by Allah.
There are more droughts now than when my father was
a young boy. There is a
problem of too many people and too many animals on
too little land. But it doesnt
mean that we have to give it up.
Woman no. 4: Women are more practical and
realistic than men. Men realise that

their circumstances have changed, but they say we are
still going to continue. We
women say, these changes mean that we also have to
change our livelihoods. We
look at other new alternatives.
Woman no. 5: Our first need is for a school. If our
children can get educated, we
will be able to give them alternatives to pastoralism.
Even if we die so poor, at least
our children will have a chance of a better future.
Man no. 3: Before a school, we need help with
restocking. Once we have animals
again, we will be able to resume our livelihoods. Then
we will be able to take care of
our families and children.
Source: Mixed-sex focus group discussion, Gashamo
District.
interested in helping with animals any more: they want to
live in towns and have nothing to
do with animals; Our children dont want to live like we do.
They want more from life than
we have; Some young girls have seen television, and say
they want to have that kind of life
that they see there.
Children who participated in discussions at local schools
confirmed these sentiments.
We are living in the twenty-first century. We cant be
expected to live like the
nineteenth century, like our ancestors. We should move
in time with the changing
times. We cant behave as if things like computers and
phones dont exist. Since we
know about these things, we also want to have them. In
the old days these things
didnt exist, so our parents just continued living as their
ancestors. For our generation,
things are different and we want different things. None
of the children in this school
want to be pastoralists. We have been sent to school in order to get a better life.1 78
178 Female high school pupil, rural Gashamo.
167
19 Implications for
policy
The major challenges facing Somali Region are political
rather than technocratic, and many
policy prescriptions (such as investing in education and
health services) are obvious and
common sense. But identifying what to do is only part of
the challenge; how to
implement these interventions to deliver effective social
protection and development
programmes is another matter altogether. This final chapter
does not provide a checklist of
policy recommendations that the government of Ethiopia and
its donor partners should
adopt in order to reduce livelihood vulnerability in Somali
Region. Instead, given the
complexity and insecurity of this highly politicised policy
context, this concluding chapter
asks four topical policy-related questions and suggests broad
directions for strategic thinking
around policy design and implementation:
1 Is pastoralism unviable in Somali Region?
2 How to plan for unpredictability?
3 Is sedentarisation the solution?
4 What kinds of safety nets are needed?
19.1 Is pastoralism unviable in Somali Region?

Livelihoods in Somali Region have suffered a series of
shocks in recent years, some natural
and some policy-related, which in combination have been so
severe that resilience and
ability to cope have been dangerously compromised.
Probably the most devastating
livelihood shock that Somali pastoralists and farmers face is
drought, and the sequence of
low rainfall years that started in 1999/2000 have been so
frequent with insufficient time
in between for herds and flocks to reconstitute that some
observers interpret this as a
long-term decline in rainfall (indicative of global climate
change) and are predicting the end
of pastoralism in the Greater Horn of Africa.
Drought is not the only risk that people face in dryland areas.
This report has demonstrated that
the people of Somali Region experience multiple sources of

vulnerability. The implication is that
a multi-pronged approach to managing risk and reducing
vulnerability is required. A common
failing of past rangeland development projects in Africa was
to focus on single issues and
interventions in isolation, such as addressing animal health
problems with veterinary services, or
addressing the under-provision of education and health
services with mobile schools and clinics.
These measures are important, and both would contribute
greatly to improving the well-being
of people in Somali Region, but they are not enough. There
are political, institutional and social
issues that also need to be addressed by the people of the
region themselves.
Two interpretations can be drawn from the evidence
presented in this report. The first is
that pastoralism in rural Somali Region is becoming
increasingly unviable, and that more
intervention is therefore needed, to support (or control)
pastoralists. This might include
building abattoirs and requiring livestock exports to be
processed according to international
phyto-sanitary standards, or it might involve settling the
pastoralists (facilitating
sedentarisation, as the government terms it) to better provide
services such as schools and
clinics. This argument is reinforced by indicators suggesting
that rural people in Somali
Region have extremely poor outcome indicators (in terms of
mortality, for instance), and
that provision of basic public services (health and education)
was seriously inadequate at the
time of our survey in sedentary as well as mobile
communities.
An alternative interpretation is that pastoralist livelihoods are
as resilient as they have ever
been, but that pastoralists are operating under an increasingly
difficult policy environment,
which makes it impossible for their usual adaptive strategies
to be implemented. Droughts
might be the trigger for recent crises, but increasing
vulnerability to drought is the
underlying problem. While conflict, border closures and
import bans all fuel the view that
168
IDS
RESEA
RCH
REPOR
T 57
pastoralism is unviable, the solution might be to lift these

constraints and allow the
traditional marketing system to continue (it is, after all,
relatively efficient and
competitive). According to this argument, giving people
options is the way forward, not
restricting options by settling pastoralists and closing down
opportunities for trade.
The instincts of the Ethiopian government are to intervene
and regulate, while the instincts
of pastoralists are flexible and contingent adapting to
unpredictable circumstances is the key
to pastoralist survival (Scoones 1995). The conflict that this
fundamental difference in
worldview provokes is inevitable, and is sometimes
mistakenly represented within Ethiopia as
a clash of ethnicities and religions. Accommodating both
worldviews within a unitary modern
nation-state is an ongoing political challenge. The
government is proposing or introducing a
series of solutions to what it perceives as the crisis of
pastoralism. Paradoxically, these
solutions sedentarisation of pastoralists, centralisation of
livestock and meat marketing,
and regularisation of informal cross-border trade (the war on
contraband) directly
contradict the pastoralists need for flexibility. While many
genuinely believe that these
policies are the only logical way forward for Somali Region,
pastoralists and traders interpret
these interventions as restricting their opportunities and
closing down options. Whether
livestock-based livelihoods remain viable in future is a
question only the people of Somali
Region male and female, young and old can answer, but
they need the freedom and
flexibility to give it a fair chance, and to change and adapt in
ways that they choose.
19.2 How to plan for unpredictability?

Pastoralism in Somali Region is characterised by instability,
unpredictability and variability. It
follows that all policy interventions by governments and
donors should follow two
principles: (1) reduce vulnerabilities, and (2) expand options,

to spread risk. Instead of
imposing solutions, external actors should play a facilitating
role, increasing opportunities for
local people to make their own decisions about their future.
Vulnerability is usually defined as a product of two factors:
exposure to risk, and ability to
cope (Chambers 1989). Rural livelihoods in Somali Region
are exposed to an extraordinary
number of diverse risks including drought, conflict, and
policy shocks against which local
people have devised a range of risk management and coping
strategies. Key to their ability
to cope is flexibility and mobility, while ability to cope is
undermined by constraints on
mobility of all kinds physical, economic, social.
In the context of pastoralist livelihoods, physical mobility
relates principally to the freedom for
people to move with their animals in search of grazing, water
and markets, including across
borders. In Somali Region, pastoralists and traders face
multiple constraints to physical
mobility. These include the Saudi livestock import ban, the
Ethiopian governments border
closures and its impounding of vehicles transporting
livestock and commodities, the
inaccessibility of certain pastures and water points to clans
because of conflict or aabsi (fear
of conflict). Most of these constraints are not natural or
environmental, but are related to
policies and politics. Positive policy changes and political
interventions could alleviate or
remove these constraints.
There are many positive steps that the regional, federal or
local government of Ethiopia
could take immediately, to facilitate the physical mobility of
people and animals in Somali
Region. These include:
Negotiating at government-to-government ministerial
level with Saudi Arabia to lift
the import ban against Ethiopian livestock.
Formalising the role of indigenous conflict resolution
mechanisms and arbitrate (where
appropriate), to resolve conflicts between (sub-)clans,
given that decentralisation
provides opportunities for direct engagement of MPs
and administrators in local issues.
Resuming the dialogue with the ONLF to agree a
ceasefire in the short term, and
negotiate a peaceful resolution of the conflict between
the government and the
ONLF in the long term.
169
Ending the war on contraband, and instead put in

place better policies to support
both trade and revenue-raising objectives, for the
benefit of pastoralists, traders and
public services.
Many would accept that contraband trade and informal crossborder livestock marketing
need to be regulated, but this should be managed in a
negotiated process that is
consultative and transparent. The arbitrary seizure of property
and apparent randomness of
actions against Somali traders and pastoralists has introduced
new sources of uncertainty to
local livelihoods, and has raised vulnerability unnecessarily,
sometimes to intolerable levels
traders and small enterprises are going out of business, which
has no obvious benefits to
anyone. It is time to replace these strategies with new and
more effective approaches.
Economic mobility refers to the ability to move out of one
livelihood activity into another,
either temporarily or permanently. Very little economic
mobility was observed in this
research, and most of the mobility that was recorded was
downward. If all the familys
livestock die during a drought, for instance, the available
livelihood options are limited and
almost always inferior, in terms of both income generation
and social status. Such options
include: resorting to petty activities such as collecting
firewood and burning charcoal for
sale, or moving into an IDP camp and becoming dependent
on food aid. None of these
livelihood strategies is preferred they are forced rather than
chosen and reflect a level of
vulnerability that has exceeded the familys ability to cope.
Achieving upward economic mobility requires access to
capital, which might be provided by
a relative living abroad in the Somali diaspora, and will allow
business enterprises to be
initiated or expanded. Alternatively, upward economic
mobility can be achieved through
formal employment, but most jobs require education at the
very least, functional literacy
and numeracy whereas most Ethiopian Somalis are
illiterate.
What can the government of Ethiopia do to facilitate upward
economic mobility and

minimise the risk of downward economic mobility in Somali
Region? Positive policy options
include:
Prioritising the delivery of good quality primary
education for all children in Somali
Region, increasing the emphasis on innovative
approaches such as mobile schools, and
providing single-sex boarding schools to encourage
female enrolment.
Promoting the development of financial intermediation
services credit and savings
to provide working capital and loans for families that do
not have relatives remitting
income from abroad.
Reducing the regulations and red tape that inhibit the
registration and operation of
private enterprises in Somali Region as elsewhere in
Ethiopia.
Expanding peoples options should be a guiding principle of
all interventions aimed at
improving livelihoods and reducing vulnerability. Basic
education (functional literacy) is a
prerequisite for making a living in most economic sectors
outside farming and pastoralism
and can also raise the returns to farming and pastoralism
yet Somali Region has the
lowest primary enrolment rate of any region in Ethiopia.
Cultural resistance to educating
children, especially daughters, may be part of the
explanation, but without adequate access
to schools parents are denied even the possibility of making a
choice. Major investments are
needed in rural education infrastructure, as well as in the
recruitment, training and
deployment of teachers.
The imperative of providing decent education cannot be
exaggerated, and there is evidence
from our fieldwork of strong demand for education,
especially among women and children.
But of course, education takes time, and this is not a strategy
that addresses the needs of
families in Somali Region that have already dropped out of
pastoralism or are on the verge
of doing so. In the short term, many of these families need
urgent social assistance (possibly
but not necessarily in the form of food aid), and they also
need urgent livelihood
170
IDS
RESEA
RCH
REPOR
T 57
assistance to support their return to a sustainable livelihood.

The appropriate form that
livelihood assistance should take depends on a judgement
about the future viability of
pastoralism. If the present sequence of droughts is regarded
as a temporary downturn in a
long-term cycle (such that average rainfall will soon return to
higher pre-drought levels),
then standard approaches to post-drought rehabilitation are
appropriate, such as restocking
of livestock herds and flocks, or providing seeds and tools for
farming. If, however, the
recent droughts are taken as indicative of an irreversible
decline in rainfall, to levels too low
to maintain livestock-based livelihoods, then any efforts to
support pastoralism are entirely
inappropriate. During our fieldwork we heard stories of
families who had been assisted by
an international NGO with livestock restocking after the
2000 drought, only to lose these
animals in the drought of 2004.
The fact is, nobody knows how much rain will fall in Somali
Region this year, or next year,
or in five or ten years time. Given this uncertainty, it is
inappropriate to impose policies on
people that either keep them trapped in pastoralism or force
them to exit pastoralism.
Flexibility is the key to pastoralist survival in an
unpredictable environment, and policymakers need to learn from the pastoralists, and to plan for
unpredictability to allow for
alternative scenarios that might or might not occur, or that
might occur for some people
but not others. Ideally, for families for whom pastoralism
does become unviable, exit
strategies from pastoralism will be available that are up and
out rather than down and
out. Education to enable the next generation to find jobs in
town is an example of an up
and out strategy. Becoming chronically dependent on food
aid handouts in an IDP camp is
a down and out outcome.
There is at present a concern among policy-makers to break
the cycle of dependence on
food aid in Ethiopia, and it is very important to avoid the
institutionalisation of food aid in

Somali Region. Nonetheless, food aid will inevitably be
needed for humanitarian reasons
for IDPs, or for people affected by drought or conflict and
for these groups food aid must
be well targeted and it should be delivered regularly and
predictably. As far as possible, food
aid should be used in a way that supports livelihoods rather
than undermining them. School
feeding or food-for-education could be introduced in support
of the delivery of education
services. Public works projects (which could pay workers
with cash wages rather than food
rations) could focus on the construction and maintenance of
basic social infrastructure
given the enormous infrastructure deficits in Somali Region
such as school buildings and
health centres, or water points that provide safe drinking
water. Economic mobility depends
on physical and human capital formation, and social
protection interventions can be used in
ways that support this.
Social mobility is determined mainly by culture, and is
therefore less amenable to policy
intervention than physical or economic mobility. In the
specific socio-cultural context of
Somali Region, some groups are more constrained and more
vulnerable than others.
Women, for instance, have heavier domestic responsibilities
and less decision-making power
within the home than do men. Outside the home, they tend to
be canalised into lowincome, low-status economic activities. But this is changing.
Some of the biggest and most
powerful traders in Somali Region are women. Driven by
processes of social change and
accelerated by the recent droughts, where loss of livestock
has reduced many men to
chewing khat, women are assuming more economic
responsibilities. In cases where their
husbands are reluctant to allow them to work outside the
home, some women are
appealing directly to the elders, and forcing social change in
that way.
Certain ethnic groups and clans in Somali Region appear to
be socially excluded and
politically marginalised. These include the Bantu Somalis
of Kelafo and Dolo Odo, and the
Issa Somalis of Shinile District. In some cases this lack of
social status and political power
translates into higher levels of poverty and vulnerability, and
should therefore be taken into
account by policy-makers. For instance, these groups might
need special attention in terms

of social protection leaving the allocation of food aid and
other social transfers to
community-based
targeting
might
reinforce
their
marginalisation and exclusion. Whether
(and how) government policies and donor or NGO
interventions should aim explicitly to
overturn the social hierarchy in Somali Region, or at least to
reduce social inequalities, is
171
debatable. Gender quotas have already been introduced to

raise womens representation in
regional political structures; perhaps similar affirmative
policies could be applied to other
groups identified as socially and politically vulnerable.
19.3 Is sedentarisation the solution?

Pastoralism has been a dominant way of life in the Horn of
Africa for thousands of years. It
would be a remarkable coincidence if the early years of the
twenty-first century happen to
be the precise historical moment when pastoralism has finally
become unviable, and
Ethiopias pastoralists urgently need to be sedentarised. Yet
this is precisely the view of
many technical experts and some policy-makers, who believe
that there is no future for
pastoralism in Somali Region, and that nothing less than a
transformation of rural
livelihoods is required. In 2001, the government articulated
the following Pastoral
Development Policy:
Phased voluntary sedentarisation along the banks of the
major rivers as the main
direction of transforming pastoral societies into agropastoral system, from mobility to
sedentary life, from rural to small pastoral towns and
urbanisation. (FDRE 2001: 5)
As was seen in the previous chapter, some pastoralists
(especially women and children) are
becoming disillusioned with pastoralism as a way of life, and
many are attracted to a more
stable existence with access to services and modern
amenities. But this potentially
increasing support for settling down does not necessarily
mean that pastoralists have a
vision of themselves or their children becoming crop farmers.
Pastoralists generally have
negative attitudes towards farming for a living (Scratching
the ground with a hoe is no life
for a Somali!), and some pointed out that farmers in Ethiopia
are even worse off than
pastoralists.
We look at the problems of the farmers in the highlands
and we ask why the
government hasnt solved their problems. Every years

tons of food aid goes to those
farmers, who are supposed to be growing their own
food.
Why are the farmers always telling us to become
farmers like them? We never tell
them to become pastoralists like us!
For many pastoralists (but not all), abandoning livestock
rearing to become a farmer would
be seen as downward mobility in both the economic and
the social sense. Farmers and
pastoralists famously have little respect for, or understanding
of, each others way of life. But
this is not universal, and attitudes are changing. Many
pastoralists are cultivating some food
crops as insurance against the need to buy food in the market,
and the boundaries between
pastoralists, agro-pastoralists and settled farmers are
increasingly blurred in the complex
network of livelihood systems that are pursued in different
parts of Somali Region. To
describe all Somali pastoralists as resistant to sedentarisation
is not true. Many have already
become part-time or permanent farmers, often as a response
to losing too many livestock
to drought. One has no control over livestock as the
persistent drought in this region is
very unkind to pastoralism. Land is better because it will
always be there and can be left
behind for future generations.17 9
Most of the agro-pastoralists in Somali Region were

pastoralists until one or two
generations ago, when they decided to diversify their
livelihoods by adding crop farming to
livestock rearing. This ability to adapt as circumstances
change is a feature of pastoralists
livelihood strategies, and should be encouraged. On the other
hand, the annexation and
enclosure of land for farming is depriving pastoralists of high
quality grazing land along
riverbanks and in valley bottoms and reducing access to
resources which are particularly
179 Farmer in Gode District.
172
IDS
RESEA
RCH
REPOR
T 57
important at times of drought. Any increase in farming raises

competition over access to
land and water between herders and farmers, and restricts
pastoralists options further.
As an alternative to encouraging pastoralists to become
farmers, the government could
facilitate phased voluntary urbanisation instead. Many
pastoralists who have lost their
livestock and dropped out of livestock rearing have moved to
district capitals like Gode or
Gashamo town, or to the regional capital of Jigjiga. This
report has shown that urban
residents of Somali Region are not only better off in terms of
income than rural residents,
they are also much less vulnerable. They are less likely to be
exposed to devastating
livelihood shocks, they are healthier and their children are
more likely to be immunised and
educated. Existing policies tend to discourage urbanisation
and keep people in rural areas,
where they remain chronically vulnerable to drought and
other livelihood shocks. On
balance, urbanisation which strengthens linkages with the
rural economy and thus
complements rather than replaces pastoralism looks like a
more attractive alternative than
sedentarisation within rural Somali Region.
19.4 What kinds of safety nets are needed?

Somali culture has developed a range of redistributive
mechanisms that serve to build social
cohesion in good times and spread risk in bad times. These
indigenous giving and sharing
mechanisms remain important, but are coming under
increasing stress. Recurrent covariate
shocks are undermining the capacity of community members
to support their relatives and
friends through crises. Social change may be undermining the
willingness of people to assist
each other as much as in the past. People who belong to
marginalised groups, or do not
have relatives in the diaspora to send remittances, are
excluded from the circles of
responsibility that provide informal social protection to the
poor and vulnerable.

Similarly, formal social assistance mechanisms, dominated in
recent years by emergency food
aid distributions, do not always reach the most vulnerable
and food insecure, for whom they
are intended. The distribution of food aid follows a vertical
channel, from federal to regional
to district administration levels, after which it is disbursed to
clan elders at the community
level, who are supposed to divide this food among the
neediest households in their area.
However, at each stage some proportion of food aid is
diverted to other purposes such as
paying government workers or lost to corruption. At the
community level, elders
distribute some food aid to needy community members but
allocate the rest to political
agendas. (When it reaches the elders they take half of it and
the rest disappears; we dont
know where it goes.) At the household level, women accuse
their husbands of selling food
aid to buy khat. Foreign wheat provided as food aid is
disliked by Somalis, so tons of this
wheat is sold on local markets, undermining farmers and
traders.
The challenge for policy-makers designing and implementing
safety-net programmes in
Somali Region goes beyond the familiar difficulties of
targeting the vulnerable in pastoralist
communities. These targeting difficulties include:
1 Pastoralists are more mobile than sedentary farmers
which complicates registration
procedures and the delivery of resources and public
services.
2 Ownership of assets (especially livestock) is partly clanbased and fluid, rather than
individualised as in farming households which
complicates the assessment of
individual wealth.
3 Because pastoralist livelihoods are characterised by cycles
of accumulation, collapse and
rebuilding, their wealth and vulnerability are highly
variable which makes
identification of the poor problematic at any point in
time.
4 The range of informal transfers that operate in pastoralist
societies (remittances,
zakaat, etc.) reduces the vulnerability of households
that benefit from these social
support systems, but are difficult for outsiders to
quantify.
173
A bigger challenge than targeting accuracy is ensuring that

substantial proportions of
resources allocated to safety nets in Somali Region are not
diverted to clan politics, as
allegedly happens with much of the federal budget allocation
to the Somali regional
government. Another challenge is to design programmes that
reflect beneficiary priorities
(rather than reflecting the availability of qamadi, which most
Somalis do not want), and
support local livelihoods rather than undermining them.
When asked if they agree that food
aid is an appropriate response to drought, farmers in Kelafo
argued that support for farming
is a more sensible safety net than food handouts: If we had
irrigation facilities, we would
be providing food to others, not collecting food aid!
Stabilising food production in
anticipation of drought shocks is a more creative risk
reduction strategy than importing food
aid after a shock has occurred.
It is also important to recognise that many other shocks
threaten livelihoods in Somali
Region, and to devise appropriate social protection strategies
to address these various
threats. While the famine of 1999/2000 was triggered by a
severe drought (and
compounded by the late delivery of food aid), the impacts
were intensified by underlying
vulnerability factors, including conflicts within Somali
Region, and the ban on livestock
imports imposed by the Gulf States since 1998. This
combination of trigger factors and
underlying vulnerability factors implies that a famine
prevention strategy for Somali Region
should include at least three components:
1 Drought risk management: pre-emptive measures (such as
weather insurance) should
be explored; as well as guidelines for off the shelf
actions that can be initiated (such
as de-stocking programmes) when early warning
indicators suggest an imminent
emergency.
2 Conflict resolution mechanisms: the regional government,
local MPs or government
officials should work more effectively with clan elders
and traditional institutions to
negotiate a ceasefire with the ONLF and end unresolved

disputes between clans over
contested land.
3 Livestock marketing support: the Ethiopian government
should negotiate with Saudi
Arabia to lift the import ban against Somali livestock,
and should promote trade that
supports local livelihoods while generating government
revenue for improved services
in Somali Region.
If this agenda for addressing vulnerability was implemented,
food aid would be mobilised
only as a last resort. Instead, droughts and conflicts have
persisted, and the Government of
Ethiopia has not yet found a way to facilitate legal crossborder trade. Instead of effective
drought risk management, conflict resolution, and support for
livestock marketing, Somali
Region has received a million tons of food aid since the
famine of 2000. This is doing little
or nothing to reduce the vulnerability of rural livelihoods in
Somali Region, and may instead
be making things worse.
174
IDS
RESEA
RCH
REPOR
T 57
Annex:
Household
questionnaire
INTRODUCTION
1) Greet the person you are interviewing, and introduce yourself.

2) Mention which elders and administrators you have met before coming to
this household.
3) Explain where you are coming from: You are coming from the Pastoralist
Communication
Initiative (PCI), which is not an organisation that is implementing projects
or delivering
assistance like food aid. Instead, PCI is trying to listen to the voices of the
people of Somali
Region, and to communicate what people are saying to the Gurti, the
regional government,
and agencies that are making decisions concerning the development of the
Somali Region.
4) Explain the purpose of the study: This research project aims to capture the
stories and
experiences of the people of Somali Region, concerning their livelihoods
and some of the
difficulties people are facing to make a living. Although some problems
are obvious, like the
droughts and conflicts affecting the Region, every household is unique.
So we need to talk
to very many households, to understand what is happening at the
grassroots level.
5) Ask if the person you are speaking to has any questions for you before
continuing.
6) Ask if the respondent is willing to be interviewed. If they agree, start the
interview. If the
respondent is not willing, do not ask any of the questions and move to the
next household.
HOUSEHOLD IDENTIFICATION
ID Name Code
Zone
INTERVIEW IDENTIFICATION
Name of Interviewer:
Interviewer Code:
Woreda
Month:___________
Date of Interview: Day:______
Kebele
Start Time:
Village
End Time:
Clan
Sub-Clan
Checked:
Name of Data Entry Clerk:
Rer
175
A.1. HOUSEHOLD PROFILE
(1) Is this a polygamous household? (circle one) Yes: 1 No: 2

(2) Is this a female-headed household? (circle one) Yes: 1 No: 2
(3) When was your household formed? (write year)
ID
How related
Can he/
Labour
Male [M] or
Age
Years of
Has s/he been

If absent:
Code
to head of
school
present in the
household?
household in
Female [F]
she read
capacity
(age in
What is he/
& write?
Where is
(circle one)
he/ she?
(write code)
(write number
completed
complete
(write
she doing?
the
last
week?
(circle
years)
code)
(writ
e
code)
(writ
e
code)
one)
(circle one)
(4) (5) (6) (7) (8) (9) (10) (11) (12) (13)
M F Yes No
Yes No
01 1 2
1 2
1 2
02
1 2
1 2
1 2
03
1 2
1 2
1 2
04
1 2
1 2
1 2
05
1 2
1 2
1 2
06
1 2
1 2
1 2
07
1 2
1 2
1 2
08
1 2
1 2
1 2
09
1 2
1 2
1 2
10
1 2
1 2
1 2
11
1 2
1 2
1 2
or
00
if
none)
12
1 2
1 2
1 2
13
1 2
1 2
1 2
14
1 2
1 2
1 2
15
1 2
1 2
1 2
16
1 2
1 2
1 2
17
1 2
1 2
1 2
Codes: How related to head of household?

Codes: If absent, where?
what
Codes: Labour capacity

Codes: If absent,
is
he/she
doing?
01 = household head
1 = elsewhere in rural Somali
1 = young child (too young to work)

Region
1 = visiting
02 = wife
2 = working child (herding livestock; doing
2 = family
reasons
2 = elsewhere in
rural
03 = son / daughter of head or wife
domestic chores; childcare; may be
Ethiopia
(funeral,
caring for
04 = son-in-law / daughter-in-law
hired / fostered out)

sick
etc.)
family,
3 = Jijiga town
05 = grandson / granddaug hter
3 = adult (able to do full adult workload)
3
=
married
into ot her
4 = other urban
centres in
06 = father / mother of head or wife
4 = elderly (not able to do full adult
household
Ethiopia
07 = brother / sister of head / wife
workload)
4
=
looking
for work
5 = neighbouring
Somali
08 = other relative of head/ wife
5
=
working
elsewhere
5 = permanently disabled (unable to work)
countries
09 = adopted
6
=
feeding/
IDP camp
6 = chronically ill (unable to work for past 3
6 = the Middle
East
10 = non-relative / servant
7 = away
for
business
months or more)
7 = the West (e.g.
UK, US)
8 = away
for
schooling
8 = other____
___________
9
=
other____
________
176
IDS
RESEA
RCH
REPOR
T 57
A.2. DEATHS IN THE HOUSEHOLD
(14) Since the fall of Mengistus government in 1991, has your household
suffered any deaths? *
(circle one)
Yes: 1 No: 2
* (If the household was formed more recently, ask since your household
was formed)
If NO, go to section A.3. If YES:

Please tell us who died What was their relationship to the household head?
Were they male or female? How old were they when
they died? When did they die (year and season)? What contribution were they
making to the households work? What was the
cause of their death?
How related to
Male / Female
Age at
death
Year of
Season of
Cause of
death
household head?
death
death
(circle one)
(age in complete
(write code)
(write code)
(write year)
(15) (16) (17) (18) (19) (20)
1 2
1 2
1 2
1 2
years)
(write code)
1 2
1 2
1 2
1 2
Codes:
Codes: Relationship to household head
Codes: Cause of death
Note:
Season
If less than one year

01 = household head
(>3 months)
01 = long illness
old at time of death
02 = wife
months)
1 = Gu
02 = short illness (<3
(including died at
2 = Hagaa
03
=
die
d
in
chi
ldb
irth
03 = son / daughter of head or wife
birth), write 00.
3 = Deyr
(wom
en
/
infants
)
04 = son-in-law / daughter-in-law
4 = Jilaal
04
=
acc
ide
nt
05 = grandson / granddaughter
5 = Dont
know
05
=
sna
ke
bit
e
or
wil
d
ani
ma
ls
06 = father / mother of head or wife
06
=
ma
lnu
trit
ion
07 = brother / sister of head / wife
07
=
fa
mi
ne
08 = other relative of head/ wife
08
=
co
nfli
ct
bet
we
en
cla
ns
09 = adopted
09
=
co
nfli
ct
wit
hin
the
cla
n
10 = non-relative / servant
10
=
co
nfli
ct
bet
we
en
ind
ivi
du
als
11
=
oth
er
typ
e
of
co
nfli
ct
12
=
lan
dm
ine
ex
plo
sio
n
13
=
old
age
14
=
oth
er:
__
__
__
__
__
__
__
__
__
_
177
A.3. CHILD IMMUNISATION

(21) Do you have a card on which vaccinations (tallaal) for your children are written down? Yes: 1
No: 2
If YES, record the details from the vaccination card for each child born in the last five years.
(22) (23) (24) (25) (26)
Type of
Child 1 Child 2 Child 3 Child 4 Child 5
Vaccination Day Month Year Day Month Year Day Month Year Day Month Year Day Month
Year
BCG
Polio
OPV 0
OPV 1
OPV 2
OPV 3
DPT
DPT 1
DPT 2
DPT 3
Measles
If NO vaccination card:
(27) (28) Please tell me if your children under 5 years old received any of the following
vaccinations
(29) (30) (31) (32) (33) Did the child If YES,
receive any
A BCG vaccination
When was the
vaccination
An injection
how many
Polio vaccine
DPT
(an
Child
against TB, that is,
first
polio
injection
given
to prevent
times?
an injection in the
vaccinations to
prevent them
given, just after
measles?
(write the
at
(drops in the
vaccine
the
mouth)?
same time as the polio
from getting
how many
drops)? If YES, how
left arm that caused
If YES,
birth or later?
number)
a scar?
many times?
times?
diseases?
Child 1
Child 2
Child 3
Child 4
Child 5
1
1
1
1
1
2
2
2
2
2
1 = Yes
3
3
3
3
3
1
1
1
1
1
2
2
2
2
2
3
3
3
3
3
1
1
1
1
1
2
2
2
2
2
3
3
3
3
3
4
4
4
4
4
5
5
5
5
5
1
1
1
1
1
2
2
2
2
2
3 1 2 3
3 1 2 3
3 1 2 3
3 1 2 3
3 1 2 3
1 = Yes, 1 time
after
time
1 = Yes
4
4
4
4
4
5
5
5
5
5
1
1
1
1
1
2
2
2
2
2
3
3
3
3
3
1 = Just
1 = Yes, 1
1 = Yes
2 = No
birth
2 = No
2 = No
2= Yes, 2 times
2= Yes, 2 times
2 = Later
3 = Dont know
times
3 = Dont
3 = Dont know
3 = Yes, 3
3 = Yes, 3 times
3 =
know
No
4 = Yes, 4 times
= No 0 times
polio
4
vaccine
5 = No 0 times
5=
Do
nt
kn
ow
178
IDS
RESEA
RCH
REPOR
T 57
B. HOUSEHOLD LIVELIHOOD
ACTIVITIES
Next, Id like to ask you how your household makes its living.
In the last 12 months (between now and the same month last year),
which types of work or activity did the members
of your household do, in order to earn food and income?
Who worked at each activity?
Is this activity seasonal or permanent (all year round)?
For every month worked on this activity, approximately how much
income was earned?
Did anyone
Which member
in
Curr-
your
of
Average
household
household
ency
Livelihood Activity
does it?
earned
do this activity
income
(write
in the last year?
(Circle
apply)
one
(Cir
cle
all
that
per
mo
nth
code)
only)
(34) (35) (37) (38)
LIVESTOCK PRODUCTION
Rearing & selling animals (camels, cattle, sheep, goats) Yes
No
3 4
Selling dairy products (milk, butter, ghee, cheese) Yes
No 1
4
Selling meat (from own livestock) Yes
No 1 2 3 4
Selling hides and skins (from goats, sheep, cattle) Yes
No 1
4
Selling wool (from sheep)
Yes
2 3 4
Rearing chickens
Yes
No 1
4
Selling eggs (from own chickens) Yes
No 1 2 3 4
Beekeeping (selling honey, bees-wax, or bee-hives) Yes
No 1
No 1
2
4
CROP FARMING
Cereal food crops (sorghum, maize, wheat, barley) Yes
No 1 2 3
4
Root crops (Irish potato, sweet potato) Yes
No 1 2 3 4
Pulses (beans, cow-pea, chick-pea) Yes
No 1 2 3 4
Oil crops (sesame, sunflower)
Yes
No
1 2 3 4
Fruits (mango, papaya, banana, orange, lemon, etc.) Yes
No 1 2 3
4
Vegetables (onion, tomato, cabbage, pumpkin, etc.) Yes
No 1 2 3
4
Khat
Yes
No 1 2 3 4
SALE OF NATURAL PRODUCTS
Charcoal
Yes
No 1 2 3 4
Firewood
Yes
No 1 2 3 4
Water
Yes
No 1 2 3 4
Precious stones (gold, gem-stones) Yes
No 1 2 3 4
Incense, frankincense, and natural gum (myrrh) Yes
No 1 2 3 4
Wild fruits (gob, yicib, etc.)
Yes
No 1
2 3 4
Animal feed (grass, fodder, forage, salty sand, etc.) Yes
No 1 2 3
4
Construction materials (sand, grass, wooden poles, etc) Yes
No 1 2
3 4
EMPLOYMENT / LABOUR
Salaried job (specify:___________________________) Yes
No 1 2
3 4
Daily labourer
Yes
No 1 2 3 4
Farm worker
Yes
No 1 2 3 4
Animal herder
Yes
No 1 2 3 4
House-maid (domestic servant)
Yes
No
1 2 3 4
Restaurant or hotel worker
Yes
No 1
2 3 4
Military service
Yes
No 1 2 3 4
Construction worker (masonry)
Yes
No
1 2 3 4
Carpenter, Furniture-maker, or Metal-worker Yes
No 1 2 3 4
TRADING (buying and selling)
179
Did anyone in
member
Which
your household
household
of
CurrA
ve
ra
ge
ency
Livelihood Activity
does it?
do this activity
income earned
(write
in the last year?
(Circle one only)
(Circle all that

per month
code)
apply)
(34) (35) (37) (38)
Livestock (cattle, sheep, goats, camels) Yes

No 1 2 3 4
Livestock products (hides, butter, etc.) Yes
No 1 2 3 4
Food crops (grains, pulses, vegetables) Yes
No 1 2 3 4
Other foods (sugar, flour, coffee, etc.) Yes
No 1 2 3 4
Clothes and shoes
Yes
No 1 2 3 4
Building materials and hardware
Yes
No 1
3 4
Contraband
Yes
No 1 2 3 4
Khat
Yes
No 1 2 3 4
Other commodities
Yes
No 1 2 3 4
CRAFTS / SMALL INDUSTRY
Basket-making, Mat-making
Yes
No 1 2
4
Weaving, Knitting, Embroidery, Tailoring Yes
No 1 2 3 4
Making traditional utensils or farm tools Yes
No 1 2 3 4
Henna (decoration)
Yes
No 1 2 3 4
Pottery
Yes
No 1 2 3 4
Making jewellery
Yes
No 1 2 3 4
SERVICES
Water-carrier, Porter
Yes
No 1 2 3 4
Counsellor (disputes, marriage)
Yes
No 1
3 4
Barber or Hairdresser
Yes
No 1 2 3 4
Musician (drum-beater, singer, dancer) Yes
No 1 2 3 4
Religious teacher (Koranic)
Yes
No 1 2
4
Circumciser
Yes
No 1 2 3 4
Traditional healer
Yes
No 1 2 3
Traditional birth attendant (TBA)
Yes
2 3 4
RENTS
House rent
Yes
No 1 2 3 4
Farm-land rent or Sharecropping
Yes
3 4
Renting out animals (for transport or farming) Yes
No 1 2 3 4
Renting out wheelbarrow
Yes
No
4
Cart (donkey-cart, horse-cart, etc.)
Yes
2 3 4
FOOD & DRINK PROCESSING
Selling tea, coffee, cake, bread
Yes
3 4
Selling fruit juice (orange, etc) or soft drink (Fanta, etc) Yes
No 1 2
Selling cooked food
Yes
No 1 2
OTHER (write name of activity)
Yes
2 3 4
Begging
Yes
No 1 2 3 4
Other:
Yes
No 1 2 3 4
Other:
Yes
No 1 2 3 4
180
4
No 1
No 1
No 1
No 1
3
3
4
4
No 1
IDS
RESEA
RCH
REPOR
T 57
C. CROP FARMING
(38) Do you (or any other member of your household) own any land?
(circle
one) Yes: 1 No: 2
(39) Did you (or any member of your household) farm during the last farming
season? (circle one) Yes: 1 No: 2
If NO, go to page 8.
(40) If YES, please tell us about the land you used for farming, and land you
rented out or sharecropped to others.
Access to Land
unit]
Yes No If Yes, how many Qodi? [or other

(count land for all
household members)
Farmed own land 1 2

Sharecropped in land (Shirkad) 1 2
Rented in land (Kiro) 1 2
Free access to someones land 1 2
Sharecropped out land 1 2
Rented out land 1 2
Gave land to someone for free 1 2
(41)
Is your farm situated in a favourable site (e.g. at a valley bottom)?
(circle one) Yes: 1 No: 2
(42)
Do you use mangat or moos to improve the farms potential?
(43)
Do you have permanent water for farming purposes in the vicinity?
(44)
Do
you
use
birkas
to
irrigate
your
crops?
(45)
Do you use any water works or wells to irrigate your crops?
(46) Do you use fertiliser on your crops to improve your farms productivity?
(47) Do you use animal manure on your crops to improve your farms productivity?
(48)
Do
you
have
enough
land
for
farming?
(49) For each crop grown, write the amount harvested in the most recent season,
and what they did with the harvest.
Total
How
many
sacks
were:
For each sack
sold:
Kilograms
Crop
Sacks
harvested
kilograms
Given
per sack
harvested Eaten at
home
away
Sold
Price
per
sack Currency
Maize
Sorghum
Beans
Sesame
Chat
Tomatoes
Onions
Other: ____________
(51) How does this years harvest compare to previous years? (circle one
per column)
Harvest Condition 2004 2003 2002 2001 2000 10 years ago If recent
harvests are worse than
Good Harvest 1 1 1 1 1 1 harvests in past years, why?
Normal Harvest 2 2 2 2 2 2
Poor Harvest 3 3 3 3 3 3
Crop Failure 4 4 4 4 4 4
Did not farm at that time 5 5 5 5 5 5
181
D. LIVESTOCK
I want to ask you about the conditions of pasture and water for your animals, also access to
livestock mark ets and
veterinary services. For each issue, state whether the situation during the past year has been
excellent, good,
adequate, poor or very poor. If you have no opinion or no information about this issue, say
dont know.
[Try to get the respondent to give a short answer. If he or she gives a long explanation, listen
carefully, and then
summarise their ex planation in one of the answers in the table below. Ask the respondent if this is
the correct
summary of their explanation, then circle the appropriate number. Then ask them to compare the
situation now
with the situation at the same time last year.]
(52) (53)
Livestock
Issues
now compared to the
What
is
How is the
situation
the
situation
like right now?

same time last
year?
(circle one only)
(circle
one
only)
How is the access to pasture for animals? 1 2 3 4 5 6 1
2
3
4
How is the quality of pasture for animals? 1 2 3 4 5 6 1
2
3
4
How is the availability of water for animals? 1 2 3 4 5 6 1
2
3
4
How is the access to veterinary services? 1 2 3 4 5 6 1
2
3
4
How is the access to drugs for livestock? 1 2 3 4 5 6 1
2
3
4
1= Excellent
1= Getting better
2= Good
2= Getting worse
3= Adequate
3= No change
4= Poor
4= Dont know
5= Very poor
6= Dont know
(54) As of today, how many of each different kind of livestock does your household own?
Please include any animals that belong to you, but are being raised by other
households.
Number
owned
by
your
household today
Livestock Type More
than 100 51 100 21 50 11 20 6 10 1 5 0
Camels: Female (Hal) 1 2 3 4 5 6 7
Camels: Male (Awr) 1 2 3 4 5 6 7
Camels: Young Female (Nirga) 1 2 3 4 5 6 7
Camels: Young Male (Qurba) 1 2 3 4 5 6 7
Cattle: Mature Female (Sac) 1 2 3 4 5 6 7
Cattle: Male bulls (Dibi) 1 2 3 4 5 6 7
Cattle: Female (Qalmo) 1 2 3 4 5 6 7
Cattle: under 1 year (Weilo) 1 2 3 4 5 6 7
Goats (Riyo)
1 2 3 4 5 6 7
Sheep (Ido)
1 2 3 4 5 6 7
Donkeys (Damer) 1 2 3 4 5 6 7
Mules (Baqal)
1 2 3 4 5 6 7
Horses (Faras)
1 2 3 4 5 6 7
Poultry (Digag)
1 2 3 4 5 6 7
182
IDS
RESEA
RCH
REPOR
T 57
(55) How has the number of livestock owned by your household changed during the past
10 years? (write number owned for each)
Change in livestock numbers Camels Cattle Sheep Goats Donkey

Mules Horses
One year ago (2004)
Two years ago (2003)
Five years ago (after the 2000 drought)
Ten years ago (around 1995)
(56) If the number owned has decreased during the last ten years (since 1995), what are
the reasons? (circle all that apply)
Reasons for decrease Camels Cattle Sheep Goats Donkey Mules

Horses
Conflict *
1 1 1 1 1 1 1
Raiding *
2 2 2 2 2 2 2
Died in drought
2 2 2 2 2 2 2
Disease **
3 3 3 3 3 3 3
Poisoned
4 4 4 4 4 4 4
Sold for food
5 5 5 5 5 5 5
Sold (not for food)
6 6 6 6 6 6 6
Ate at home
7 7 7 7 7 7 7
Lent out to somebody
8 8 8 8 8 8 8
Stolen
9 9 9 9 9 9 9
Lost
10 10 10 10 10 10 10
Eaten by wild animal (e.g. hyena) 11 11 11 11 11 11 11
Charity (Zakaat)
12 12 12 12 12 12 12
Contribution (Qaadhaan) 13 13 13 13 13 13 13
Dowry payment (Yarad) 14 14 14 14 14 14 14
Offering (Allah bari)
15 15 15 15 15 15 15
Compensation payment (mag) 16 16 16 16 16 16 16
Other ***
17 17 17 17 17 17 17
* If Conflict or Raiding,
what happened?
** If Disease, what is the name of the
disease, or the symptoms?
*** If Other, please be specific:
183
E. WATER
What is your main source of water for drinking and cooking, by season? (circle one per column)
What is your main source of water for livestock, by season? (circle one per column)
What does it cost you to use this water, in birr?
Rainy
Dry season
season
Source of water People

Livestock
People
Livestock
(drinking,
(drinking)
(drinking,
(drinking)
cooking)
cooking)
(57) (58) (59) (60)
River
1 1 1 1
Pond or lake (open access)
2 2 2 2
Pond or lake (fenced)
2 2 2 2
Borehole
3 3 3 3
Birkas
4 4 4 4
Shallow well
5 5 5 5
Rainwater harvesting
5 5 5 5
Tap
6 6 6 6
Other (specify): _________________________
______________________________________ 7 7 7 7
(61) How much do you pay for each source of
water each time you use it? (Write the amount in

birr. If no payment is needed, write 0.)
Unit for purchased water (e.g. 200-litre barrel)
How many barrels of water do you normally buy in one week during the dry season?
F. HEALTH & EDUCATION
(61) Is there a health clinic in your community?
If there is no health clinic in your community, how far is it to

the nearest health clinic? (write distance in kilometres)
(62) Is there a Quranic school in your community?
Is there a formal school in your community?
If there is no formal school in your community, how far is it to

the nearest formal school? (write distance in kilometres)
Are any of your children attending school?
If
NO,
why
not?
(1)______________________________________________________________________
_______
(2)________________________________________________________________
_____________
(3)________________________________________________________________
_____________
184
IDS
RESEA
RCH
REPOR
T 57
G.1. INFORMAL TRANSFERS
In the last 12 months (between now and the same month last year), has your
household received any of the following
types of assistance from anyone outside the household?
If YES, who gave you this help a relative, friend or neighbour, or someone else?
Where does the person live in your community, or somewhere else?
Where
do they
live?
Type of Transfer Yes
No From whom?
(circle as many as
apply)
(circle as many as
apply)
(63)
(66)
(64)
Remittances (from relative living elsewhere)

4 5 6 7 8 9
2 1
(65)
Zakaat
Other cash gift (e.g. Sadaqa)
7 8 9
2 1
3
2
1
4
3
5
4
2 1
6
5
2
7
1
3
8
2
4
9
3
1
4
2
7
3
8
4
9
2 1
6
1
4
2
7
3
8
4
9
2 1
6
1
4
2
7
3
8
4
9
2 1
6
1
4
2
7
3
8
4
9
2 1
6
1
4
2
7
3
8
4
9
2 1
6
1
2 1
4 5 6
4 5 1
2
7
2
3
8
3
4
9
4
3
3
Cash loan (no interest)

Food or grain gift
Grain loan (no interest)
Seed gift
Seed loan
Free labour
Free use of oxen or plough
1
2 1 2
8 9
Free use of pack animals (camels or donkeys)
3 4 5 6 7 8 9
2 1
Xoolo goyn (restocking of poorer relatives)

4 5 6 7 8 9
2 1
Ciyi (distribution of meat to neighbours after a

3
1
4
2 1
6
2
7
3
8
4
9
2 1
slaughter takes place)

Alabari / Sab (sacrifice made to feed the poor)
3 4 5 6 7 8 9
Dhibaad (cash or kind gift to married daughters

when they visit parents or relatives)
1
6 7 8 9
Kaalo (gift of livestock to newly-weds)
5 6 7 8 9
2 1
animals to a relative or friend)

1
2 1 2 3 4 5 1 2 3
7 8 9
Yarad (dowry given to brides parents)
1
2 1 2 3 4 5 1
5 6 7 8 9
2 1
1
5
Irmaansi / Maal (donation or loan of milking
Qaadhan (donations of cash or animals to

3
1
4
2 1
6
2
7
3
8
4
9
disaster stricken people)

Diiya (blood money compensation)
5 6 7 8 9
2 1
Other:
1
4
2
7
3
8
4
9
2 1
6
1
4
2
7
3
8
4
9
2 1
6
Other:
Codes: From whom?

Codes: Where do they
live?
1
=
relative
1 = in the same
community
2 = friend or neighbour
2 = elsewhere in rural
Somali Region
3
=
trader
3 = elsewhere in rural
Ethiopia
4
=
moneylender
4 = Jijiga town
5 = other
5
= other urban centres in
Ethiopia
6
=
n
e
i
g
h
b
o
u
r
i
n
g
S
o
m
a
li
c
o
u
n
t
r
i
e
s
7
=
t
h
e
M
i
d
d
l
e
E
a
s
t
8
=
t
h
e
W
e
s
t
(
e
.
g
.
U
K
o
r
U
S
)
9
=
o
t
h
e
r
:
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
_
185
G.2. ACCESS TO SUPPORT NETWORKS

(67) If your household had a problem and needed money or food urgently, would you be
able
to get it from people in your community or from relatives?

Yes: 1 No: 2
(circle one)
(68) How many people do you think you could ask for this kind of help (money or food)?
(write number, or 00 if none)

(69) If someone in your household fell ill or was injured, and you needed help with work,
would
you be able to get it from people in your community or from relatives?

(70) How many people do you think you could ask for this kind of help (with work)?
(write number, or 00 if none)
H. FORMAL TRANSFERS
In the past 12 months, which types of assistance did your household receive from
government or aid agencies?
Faffa
Cash-for-Work
Credit
Food aid Food-for-Work
project
project
(special food) Free
cash Seeds &
Tools
from NGO Livestock
(71) (72) (73) (74) (75) (76) (77) (78)
Yes
Yes
Yes
Yes
1
1
1
1
No
No
No
No
2
2
2
2
Yes
No
Yes
Yes
Yes
1
1
1
No
No
No
2
2
2
How many times in the last three years has your household received food aid or gone for food-forwork?
How much food aid did you receive each time? What did you do with this food?
How many times
How much food
did you receive
What did you do

with the
did you receive
food aid?
Year
food aid?
each time?
(write code)
(write number)
(write amount)
(79) (80) (81)
This year (2005)

Last year (2004)
Two years ago (2003)
Note: If no food
Note: Specify the

1 = Sold all food aid for
cash
aid was received
units (e.g. 10 kg).

2 = Sold some food aid
and
by this household
in any year, write
aid away to
0 next to each
it more
year.
consumed the rest

3 = Gave all food
others who needed
4 = Gave some food aid
away and
consumed the rest
5 = Gave all food aid to
others as
a
payment
for
something
6 = Gave some food aid
as a
payment, consumed
the rest
7 = Consumed all food
aid
8 = Gave it to livestock
for feed
9 = Other (specify):
__________
____________________
_______
186
IDS
RESEA
RCH
REPOR
T 57
I. COPING STRATEGIES
(82) During drought and other livelihood shocks, how does your household survive?
[Note: This question is about unusual behaviour, not what the household
normally does to get its food in a good year.]
Coping strategy Yes No Coping strategy Yes No
Eat less food (smaller portions) 1 2 Fishing for food 1 2
Collect bush products to sell to buy food 1 2 Rent out land 1 2
Collect wild fruits for food
1
2
Sell land
1 2
Hunting (gazelle, dik-dik, etc.) 1 2 Sell other assets to buy food 1 2
Migrate to urban areas to find work 1 2 Save fodder for sale 1 2
Go to IDP camp in Somali Region 1 2 Rely on handicraft 1 2
Flee to neighbouring countries 1 2 Sell firewood 1 2
Borrow food or cash to purchase food 1 2 Sell charcoal 1 2
Rely on help from relatives and friends
1
2
Begging
1 2
Reduce the number of meals per day 1 2 Smuggling / contraband 1 2
Rent out animals (donkey, camel) 1 2 Send children to work 1 2
Reduce spending on non-food items 1 2 Other (specify):
G.3. ZAKAAT
(83) When was the last time you paid zakaat?
(84) How much did you pay?
How
much
did you pay?
Year (circle
one only) Cash Camels Cattle Sheep & Goats
Crops (Rice,
Maize, etc.)
Last year (2004) 1
Two years ago (2003) 2
Three years ago (2002) 3
Four years ago (2001) 4
More than four years ago 5
Never paid any zakaat 6
(85) When was the last time you received zakaat?
receive?
(86)
How much did you
How much did

you receive?
Year (circle
one only) Cash Camels Cattle Sheep & Goats
Crops (Rice,
Maize, etc.)
Last year (2004) 1

Two years ago (2003) 2
Three years ago (2002) 3
Four years ago (2001) 4
More than four years ago 5
Never received any zakaat 6
187
J. IMPACTS OF CONFLICT
(87) Since the fall of Mengistu in
(88) Has any member of your family been
1991, how has conflict affected your

familys livelihood? (circle one only)
Very seriously 1
Quite seriously 2
Not very seriously 3
Not at all 4
injured, or lost their life, due to conflict?

(circle all that apply)
Someone was injured slightly 1
Someone was seriously injured 2
Someone lost their life 3
No injury or loss of life 4
(89) Have you lost or recovered any livestock,
(90) Have you lost or gained any access to
granary stocks or property to raids or conflict?

water, grazing, or farmland because of raids
or conflict? (circle all that apply)
Lost Eshanay
Lost
Eshanay
1-10 animals 1 1 Grazing land 1 1
11-50 animals 2 2 Farmland 2 2
More than 50 animals 3 3 Birka 3 3
Granary stocks 4 4 Wells 4 4
Movable property 5 5 Other water sources 5 5
None of the above 6 6 None of the above 6 6
(91) The last time there was any conflict that affected
(92) Who were the main beneficiaries
your community, who was involved? (circle all that apply)

apply)
Within
sections
of
the
clan
1
Within the woreda
2
Between other clans 3 Politicians
Between other woredas 4 Government officials 4
Between
our
clan
and
another
clan
5
Between
government
and
clans
6
6
Between militias and clans 7 Other (specify): 7
Bandits
8
No conflict
9
(93) Who do you turn to in cases of conflict or trouble,
neighbouring
to solve the problem? (circle all that apply)

Family members
1
Elders
2
Amakaris (Gurti)
3
Woreda administration 4 Conflict
Police
5
Militia
6
Regional representatives 7
Other (specify):
8
of that conflict?
(circle all that
Militias
2
Bandits
3
Clan
No
(99)
elders
conflict
What are your relationships with
groups at this time?

1
Good
2
Stable
3
Tense
4
________________________________
(94) Have you paid any compensation due to conflict with other families or clans? Yes: 1 No: 2
(95) If YES, what was decided to be your households
share (hagaag) of the compensation to pay?

(96) Have you received any compensation due to conflict with other families or clans? Yes: 1 No: 2
188
IDS
RESEA
RCH
REPOR
T 57
(97) If YES, what compensation did you receive?
(98b) Is there any

water point that
(98a) Is there any land that you used to
use for grazing or farming, but you dont

you
used
to
use,
Yes 1 No 2
use it any more?
but
you
Yes 1 No 2
dont
use
it any more?
If YES, why dont you
use that land any more?
If YES, why dont you
use that water point any
more?
(100) Do you feel that you are fairly and adequately

effective in
represented in local, regional, and federal

federal
governments? (circle one for each level)
Level of
Government Yes No Dont
Government
Yes
know
Local 1 2 3 Local 1 2 3
Regional 1 2 3 Regional 1 2 3
Federal 1 2 3 Federal 1 2 3
(101)
Are your representatives
representing you at local, regional and

levels? (circle one for each level)
know
No
Level of
Dont
K. FOOD SECURITY
(102) During the last Jilaal season, did your household suffer any shortage of food to eat?
(circle one) Yes 1 No 2
(103) During the worst month last year, how many times a day did the adults and children in
your household eat?
Number of meals per day (circle one for each row)
Adults
0
Children (= school-age / working, not infants) 0
Code:
1
1
2
2
3
3
4
4
0 = sometimes passed a whole day without eating
anything
L. CLOTHING & BASIC ITEMS
(103) During the past 3 years, how many times has your household
following items
bought clothes for adults and for children?

now? (circle one per row)
(104) Do you have the
(circle one per row)
in your house
New clothes Number of times Items Yes No

New clothes for adults 0 1 2 3 4 or more Salt (Cusbo or Milix) 1 2
New clothes for children 0 1 2 3 4 or more Sugar (Sonkor) 1 2
Second-hand clothes for adults 0 1 2 3 4 or more Tea-leaves (Caleen shah)
1 2
Second-hand clothes for children 0 1 2 3 4 or more Kerosene (Gaas) 1 2
189
M. SELF-ASSESSMENT OF HOUSEHOLD SITUATION
Considering all the questions discussed in this interview, how would you describe the
situation of your household now?
Are you able to meet your household needs by your own efforts?
Are you making any extra for stores, savings or investments (e.g. buying livestock
or other assets, improving your land)?
Do you sometimes need help from your community, or from government or other
agencies?
Are you dependent on this help? (Could you survive without it?)
Choose the category in column 97 that best fits the respondents answers. Then say,
So would you agree that at this time your
household is (read category description)? If they do not agree, discuss
further and identify the category they
agree with. When they agree, circle the corresponding code for Now (column 98).
Then ask: At the same time (same month) last year, was your household situation
better, the same, or worse? Repeat the
questions if necessary, read the category that best fits the respondents description of
their situation a year ago, and when
they agree circle the code in column 99.
Repeat for 2 years ago (column 100), 4-5 years ago (column 101), and 10 years ago
(column 102).
Now
The same
years
10 years ago
time last year
two
(1994)
(2005)
ago
Self-Assessment
(During
(After the fall of
(2004)
The same
4-5
ago
time
(1999 / 2000)
years
Categories
drought
(2003)
em
erg
en
cy)
Me
ngi
stu
in
19
91)
(circle one)
(circle one)
one)
(circle one)
(105)
(circle one)
(circle
(106) (107) (108) (109) (110)
was not yet formed at that time
0 0 0 0
ble to meet household needs by your own efforts,

1 1 1 1 1
making some extra for stores, savings and
2 2 2 2 2
ble to meet household needs, but with nothing
to save or invest
anaging to meet household needs, but only by
3 3 3 3 3
and / or sometimes
4 4 4 4 4
dependent on support from community or
N. Livelihood Problems
(111) What are the most serious problems that your household faces in making a
living?
(Probe until they mention three problems. Write whatever they say, with direct
quotations if possible.)
1
2
3
190
IDS
RESEA
RCH
REPOR
T 57
O. Food Consumption
(112) Have you eaten these foods at home in the last week?
Yes No If YES, where did you get this food? (circle all that
apply)
Food (circle one
Own Production
only)
or Own Livestock Purchased Food Aid Gift
(from
Relative or
Friend)
Sorghum 1 2 3 4 5 6
Maize 1 2 3 4 5 6
Rice 1 2 3 4 5 6
Barley 1 2 3 4 5 6
Pasta 1 2 3 4 5 6
Wheat 1 2 3 4 5 6
Bread 1 2 3 4 5 6
Beans 1 2 3 4 5 6
Meat 1 2 3 4 5 6
Milk 1 2 3 4 5 6
Eggs 1 2 3 4 5 6
Fish 1 2 3 4 5 6
Vegetables 1 2 3 4 5 6
Fruit 1 2 3 4 5 6
Other: ________ 1 2 3 4 5 6
P. Changes in Livelihood Activities

(113) Did you or any member of your household ever do any livelihood activity in the
past
Ye
s:
1
No
:
2
that you do not do anymore? (circle one)
(114) If YES, what was that livelihood activity?
______________________________________________________________________
____________________________
(115) Why did you or your relative stop doing it? [Only answer if question 113 was
YES.]
______________________________________________________________________
____________________________
______________________________________________________________________
____________________________
______________________________________________________________________
____________________________
______________________________________________________________________
____________________________
(116) Did your father/mother do the same livelihood activity that you are doing now?
Yes: 1 No: 2
(117) If NO, what different livelihood activity did he do?
______________________________________________________________________
____________________________
(118) Why do you not follow the same livelihood activity as your father? [Only
answer if question 116 was NO.]
______________________________________________________________________
____________________________
______________________________________________________________________
____________________________
______________________________________________________________________
____________________________
______________________________________________________________________
____________________________
Thank the respondent for his or her time. Then fill in the End
time on page 1.
191
Glossary
aabsi fear and tension caused by general insecurity and the
possibility of conflict erupting
between specific different groups or (sub-)clans
aw a type of grass used to make mats and for roofing huts
birka constructed water reservoirs (plural is berkad)
Birr Ethiopian currency
faffa supplementary food for young children
hagaag compensation payment
Haud or Hawd dry bush in the plains areas; refers to
districts to the south of the Ethiopia
Somaliland border (including Harshin, Aware, Gashamo) that
were called the reserve area
during the postWorld War II British administration
iskutab sharecropping
kebele administrative sub-district in Ethiopia (Amharic)
khat a mild stimulant (catha edulis) commonly chewed in the
Horn of Africa
moos furrow for channelling water to a farmed patch of land
(mangat in Amharic)
qamadi wheat has become a popular but derogatory term
for food aid wheat
qharan contribution of food or animals to relatives)
qodi unit of land in Somali Region (8 qodi equals one
hectare)
rer a pastoralist settlement or encampment in Somali Region
sadaqa cash gift
shoats sheep and goats
woreda administrative district in Ethiopia (Amharic)
Seasons in Somali Region

Deyr short rains (OctoberDecember)
Gu long rains (AprilJune)
Hagaa or Hagar dry season (JulySeptember)
Jilaal hot dry season (JanuaryMarch)
Karan third rainy season (AugustSeptember) in northern
Somali Region
Droughts and famines in Somali Region

Abarti daba deer long-tailed a drought that lasted for
two years (19745)
Barbar inflammation or swelling those affected had
swollen bellies and necks (1992)
Sima the equaliser a term applied to the Gode droughtfamine (1999/2000)
Tuur ku Qaat carry on your back with even pack
animals such as camels dying, people

had to transport things themselves (2004)
192
IDS
RESEA
RCH
REPOR
T 57
Community coping strategies

Allah bari or sab sacrifice made to feed the poor
awino food cooked for the poor
ayitu charitable contributions
ciyi distribution of slaughtered meat to neighbours, usually
immediately after a slaughter
takes place
dhibaad cash or kind to married daughters when they visit
parents/relatives
goob free labour provided to relatives or friends
guus Assistance provided in farming communities, taking the
form of a loan of oxen for
ploughing, farming instruments, or casual labour. The
beneficiary provides lunch/tea, or khat
to those assisting.
hagaag share of compensation payment
irmaansi animal loaned to a relative for milking for a certain
period
ischar women's contributions to another woman in need
jiisin traditional restocking mechanism, found especially in
Shinile District
kaalo gift of livestock to newly married couples
maal donation or loan of milking animals to next of kin,
neighbour or friend
nisaab the minimum level of income or wealth before zakaat
is payable
qaadhaan donations made in cash or animals to disaster
stricken people
xoolo goyn restocking of poorer relatives
yarad dowry given to the parents of the bride
zakaat Third Pillar of Islam, an obligatory redistribution
system whereby the better-off
provide support to the poor in cash or in kind (livestock or
crops)
Somali informal institutions

Ayuto informal savings and credit groups
Dia the Dia paying group is an important institution that all
Somalis must belong to. It acts
as a form of social insurance, as the group accepts
responsibility for the action of its
members who are collectively bound to pay blood money.
Access to water, wells and
pasture is availed freely to all members of the group. Dia is
Arabic, but widely used; Mag has
the same meaning, in Somali.

Gurti Elders Advisory Council in Somali Region; also the
Elders Council in Somaliland that
forms the Upper Chamber of Parliament. Amakari, an
Amharic word for the same council,
is sometimes used in Somali Region
Xeer Somali customary law
Note on spelling
Somali place names have several spellings. Often k and q
are interchangeable, as in Fiq or
Fik, or Kebribayah or Qabribeyah (and many other
variations). The capital of Somali Region is
sometimes spelt as Jigjiga and sometimes as Jijiga. In this
report, Somali spellings will be
preferred; for example, khat is Somali, while chat is Amharic
and qat is Arabic.
193
Bibliography
Baars, R.M.T. and Aptidon, S.M. (2002) Pastoralists
Perceptions of Rangeland Degradation
in Eastern Ethiopia, Nomadic Peoples 6.1: 14457
Barrett, C. and Maxwell, D. (2005) Food Aid After Fifty
Years: Recasting Its Role, London:
Routledge
Baulch, B. and Umar, A. (2005) Trading in an Uncertain
Environment: Livestock Marketing
and Price Behaviour in the Somali Region of Ethiopia',
Report for Pastoral Communication
Initiative, Brighton: Institute of Development Studies
Bradbury, M. (2000) Normalizing the Crisis in Africa,
Journal of Humanitarian Assistance,
www.jha.ac/articles/a043.htm, (accessed: 1 April 2006)
Burton, R. (1856) First Footsteps in East Africa, London:
Tylston and Edwards
Chabal, P. and Daloz, J.-P. (1999) Africa Works: Disorder as
Political Instrument, London: Zed Press
Chambers, R., Pacey, A. and Longhurst, R. (eds) (1981)
Seasonal Dimensions to Rural Poverty,
London: Frances Pinter
Chambers, R. (1989) Vulnerability, Coping and Policy, IDS
Bulletin 20.2
Clay, D., Molla, D. and Habtewold, D. (1999) Food Aid
Targeting in Ethiopia; A Study of
Who Needs It and Who Gets It, Food Policy 24: 391409
Deng, Luka Biong (2003) Confronting Civil War: A
Comparative Study of Household
Livelihood Strategies in Southern Sudan, D.Phil thesis,
Brighton: Institute of Development
Studies
Devereux, S. (1993) Goats Before Ploughs: Dilemmas of
Household Response Sequencing
During Food Shortages, IDS Bulletin 24.4: 529
Devereux, S. (2004) Food Security Issues in Ethiopia:
Comparisons and Contrasts Between
Lowland and Highland Areas, paper for a seminar organised
by the Pastoralist
Communication Initiative, UNOCHA, Sheraton Hotel, Addis
Ababa, 16 February 2004
Devereux, S., Sharp, K. and Amare, Yared (2003) Destitution
in Wollo, Ethiopia, IDS Research
Report 55, Brighton: Institute of Development Studies
Dyson, T. and Grda, C. (eds) (2002) Famine
Demography: Perspectives from the Past and
Present, Oxford: Oxford University Press
Ellis, J. and Swift, D. (1988) Stability of African Pastoral

Ecosystems: Alternative Paradigms
and Implications for Development, Journal of Range
Management 41.6: 4509
European Union (2005) European Union Election
Observation Mission Ethiopia 2005
Preliminary
Statement,
www.eu.int
/comm/external_relations/human_rights/
eu_election_ass_observ/ethiopia/pre_stat_17-05-05.pdf
(accessed: 1 April 2006)
FAO (1998) Terminology for Integrated Resources Planning
and Management', Rome: FAO
Federal Democratic Republic of Ethiopia (FDRE) (1998) The
1994 Population and Housing
Census of Ethiopia: Results for Somali Region, Volume I
Statistical Report, Addis Ababa: Office
of Population and Housing Census Commission, Central
Statistical Authority
Federal Democratic Republic of Ethiopia (FDRE) (2001)
Statement on Pastoral Development
Policy, Addis Ababa: Ministry of Federal Affairs
Development and Poverty Profile of
Ethiopia for 2002, Addis Ababa: Welfare Monitoring Unit,
Ministry of Finance and Economic
Development
Productive Safety Net Programme:
Programme Implementation Manual, Addis Ababa: Ministry
of Agriculture and Rural
Development
194
IDS
RESEA
RCH
REPOR
T 57
Kebebew, Fassil, Tsegaye, Diress and Synnevg, Gry (2001)

Traditional Coping Strategies of the
Afar and Borana Pastoralists in Response to Drought,
Drylands Coordination Group Report 17,
s, Norway: Noragric
FEWS NET (2003a) Greater Horn of Africa (GHA) Food
Security Bulletin, no. 18, November
2003, Washington D.C.: FEWS NET
FEWS NET (2003b) Greater Horn of Africa (GHA) Food
Security Bulletin, no. 19, December
2003January 2004, Washington D.C.: FEWS NET
FEWS NET (2004a) Greater Horn of Africa (GHA) Food
Security Bulletin, no. 20, February
2004, Washington D.C.: FEWS NET
FEWS NET (2004b) Greater Horn of Africa (GHA) Food
Security Bulletin, no. 20, March 2004,
Washington D.C.: FEWS NET
FEWS NET (2004c) Greater Horn of Africa (GHA) Food
Security Bulletin, no. 22, April 2004,
Washington D.C.: FEWS NET
FEWS NET (2005) 2005 Outlook for Pastoral Areas in the
Horn of Africa, Washington D.C.:
FEWS NET
Food Security Analysis Unit (FSAU) (2004) 2004 Post Gu
Analysis, Technical Series Report
IV.2, Nairobi: Food Security Analysis Unit
Gebre-Mariam, A. (2005) The Critical Issue of Land
Ownership: Violent Conflict Between Abdalla
Tolomogge and Awlihan in Godey Zone, Somali Region of
Ethiopia, Working Paper 2. Bern:
NCCR North-South.
Grahn, R. (2001) Livelihood Diversification in Southern
Ethiopia: A Study of Gobicha
Settlement, MPhil dissertation in environment and
development, Cambridge: University of
Cambridge, Dept of Geography
Hagmann, T. (2005) Beyond Clannishness and Colonialism:
Understanding Political Disorder
in Ethiopias Somali Region, Development in Practice 15.2
Herren, U. (2000) The Commercial Sale of Camel Milk from
Pastoral Herds in the Mogadishu
Hinterland, Somalia, ODI Pastoralist Development Network
Paper 30a, London: Overseas
Development Institute
Hoddinott, J. and Yohannes, Yisehac (2002) Dietary Diversity
as a Food Security Indicator,

FCND Discussion Paper 136, Washington D.C.: International
Food Policy Research Institute
Institute of Development Studies (IDS) (2002) Learning the
Lessons? Famine in Ethiopia,
19992000 (video), London: Rockhopper Productions
Little, P. (2000) Cross-Border Livestock Trade and Food
Security in the Somalia and Northeastern
Kenya Borderlands, research report of the Broadening
Access to Markets and Input SystemsCollaborative Research Support Program (BASIS-CRSP) and
OSSREA Project on CrossBorder Trade and Food Security in the Horn of Africa,
Binghamton, New York: Institute for
Development Anthropology
Little, P., Smith, K., Cellarius, B., Coppock, D. and Barrett,
C. (2001) Avoiding Disaster:
Diversification and Risk Management among East African
Herders', Development and Change
32: 401-33
Little, P. (2002) The Global Dimensions of Cross-Border
Trade in the Somalia Borderlands
in A. Ghaffar and M. Ahmed (eds), Globalisation, Democracy
and Development in Africa: Future
Prospects, Ethiopia: Organisation for Social Science
Research in Eastern and Southern Africa
Luqman, W. and Danowski, T. (1976) The use of Qat (Catha
Edulis) in Yemen: Social and
Medical Observations, Annals of Internal Medicine, 85:
2469
Lybbert, T.J., Barrett, C.B., Desta, S. and Coppock, D.L.
(2004) Stochastic Wealth Dynamics
and Risk Management Among a Poor Population, Economic
Journal, 114: 75077
195
Ministry of Finance and Economic Development (MoFED)

(2002) Development and Poverty
Profile of Ethiopia for 2002, Addis Ababa: Welfare
Monitoring Unit (WMU)
Miz-Hazab Research Centre (2002) In-depth Study of the
Knowledge, Attitude, Behaviour and
Practice of IDPs in Ethiopia Towards HIV/AIDS and Their
Health Status and Medical Care
Assessment, Addis Ababa: UNAIDS
Narbeth, S. (2003) Livelihoods and Protection Study of IDPs
and Vulnerable Communities in
Kismaayo, Nairobi: United Nations Office for Coordination
of Humanitarian Affairs
(UNOCHA), Somalia
Narbeth, S. (2004) Emergency Cash Relief to Drought
Affected Families of Sool Plateau: Postdistribution Survey Findings, Nairobi: United Nations Office
for Coordination of Humanitarian
Affairs (UNOCHA), Somalia
National Election Board (NEB) (2004), Registration of
Voters and Summary of the Results by
Wereda of January [Tere] 16, 1996 (E.C.) Elections in
Somali Region, Addis Ababa: NEB
Salama, P., Assefa, F., Talley, L., Spiegel, P., van der Veen, A.
and Gotway, C. (2000)
Malnutrition, Measles, Mortality, and the Humanitarian
Response During a Famine in
Ethiopia, Journal of the American Medical Association
(JAMA), 286.5: 56371
Save the Children UK (SC-UK), DPPB and partners (2001a)
HFE Baseline Study:
Lowland/Hawd Pastoral Food Economy Zone, Jigjiga: Save
the Children UK
Save the Children UK (SC-UK), DPPB and partners (2001b)
HEA Baseline Study: Jigjiga
Sedentary Farming Food Economy Zone, Jigjiga: Save the
Children UK
Save the Children UK (SC-UK) (2004) Emergency Nutrition
Assessment, London: Save the
Children
Save the Children UK (SC-UK) (2005) Preliminary Results
of Hart Sheik and Fafen IDP Camps:
Summary of the Main Findings (Anthropometry & Mortality),
Addis Ababa: Save the Children UK
Scoones, I. (ed.) (1995) Living with Uncertainty: New
Directions in Pastoral Development in Africa,
London: Intermediate Technology Publications
Sen, A. (1981) Poverty and Famines, Oxford: Clarendon
Press
Shuaib, T. (1991) Essentials of Ramadan, The Fasting
Month, Los Angeles: Daawah Enterprises;
www.usc.edu/dept/MSC/fundamentals/pillars/fasting
(accessed: 20 October 2004)
Swift, J. (1989) Planning Against Drought and Famine in
Turkana: A District Famine
Contingency Plan, in T. Downing, K. Gitu and C. Kamau
(eds), Coping with Drought in Kenya:
National and Local Strategies, Boulder and London: Lynne
Rienner
Tezera Getahun and Kassa Belay (2002) Camel Husbandry
Practices in Eastern Ethiopia: The
Case of Jijiga and Shinile Zones, Nomadic Peoples 6.1:
15879
Umar, A. (2000) The Organisation of Livestock Marketing in
Northern Kenya and Southern
Somalia, and Its Potential for Pandemic Control, Report of
Study for the Terra Nuova and the
OAU/IBAR/PARC, Nairobi: Kenya Pastoralist Forum
Umar, A., Ahmed, H. and Adan, T. (2002) Herders with no
herds: Displaced Pastoralists in the
Fik Zone, Somali Regional National State, Ethiopia, Jigjiga:
Save the Children (UK)
UNOCHA (2004) Updated Statistics: Internally Displaced
Persons (IDPs) in Ethiopia, Addis
Ababa: UNOCHA
van de Walle, N. (2001) African Economies and the Politics
of Permanent Crisis, 19791999,
Cambridge: Cambridge University Press
World Bank (2001) Ethiopia Woreda Studies, Addis Ababa:
World Bank Country Office in
Ethiopia
World Food Programme (2000) WFP Food and Nutrition
Handbook, Rome: World Food
ProgrammeWorld Food Programme (2005) Food Security
and Vulnerability Analysis Report:
Haiti, Rome: World Food Programme VAM Branch
196
Publications
IDS Online Bookshop
You can browse our catalogue and order publications securely online at
www.ids.ac.uk/ids /bookshop
Recent IDS Series Papers
The Political Economy of the Resource Curse: A Literature Survey,
Andrew Rosser, IDS Working Paper 268, April
2006
New Democratic Trends in China: Reforming The All-China Federation of
Trade Unions, Jude Howell, IDS Working
Paper 263, March 2006
Polio Vaccines Difficult to Swallow. The Story of a Controversy in
Northern Nigeria, Maryam Yahya, IDSWorking
Paper 261, March 2006
Microfinance Impact and The MDGs: The Challenge of Scaling-up,
Martin Greeley, IDS Working Paper 255, March
2006
Unpacking Rights and Wrongs: Do Human Rights Make a Difference? The
Case of Water Rights in India and South
Africa, Lyla Mehta, IDS Working Paper 260, November 2005
IDS Publications Series
IDS Bulletin
Brings together the latest thinking and research from programmes and
event s at IDS and present s them in an
accessible way for development practitioners, policymakers and
researchers.
IDS Working Papers
Present work in progress and interim results from a range of research
projects.
IDS Discussion Papers
Present new ideas and research material at an advanced stage for
circulation and comment.
IDS Research Reports
Final reports of major research projects conducted through IDS.
I D S_ M a ste r L og o Bl a ck
IDS Development Bibliographies

Present annotated literature guides and an overview essay on specific
development themes.
All IDS academic papers are peer reviewed by one internal and one
external reviewer.
For mail order sales and

subscription details please
contact:
I D S _M a ste r L o go _ M i n i
m Si z e
mu
IDS Communications Unit

Institute of Development
Studies
X

Mi n i mu m S i z e
ni
m u mS i z
X: 1 5 m m
X:1
M i
e
Tel: +44 (0) 1273 678269
5mm
Fax: +44 (0) 1273 621202

Web:
www.ids.ac.uk/ids/bookshop
ID S_ M a ster L ogo _ M ini m um Si ze
Mi ni mu m S i ze
um S i
X : 15m m

Brighton BN1 9RE UK
T +44 (0) 1273 606261
F +44 (0) 1273 621202
E ids@ids.ac.uk
www.ids.ac.uk
Mi ni m
ze
X : 15m m

Somali Region

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Somali Region

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Somali Region

Uploaded by

Copyright:

Available Formats

RESEARCH

ID S_M a ster L ogo

For further information on IDS publications and for a free catalogue,

IDS RESEARCH REPORT 57

IDS RESEARCH REPORT 57

Vulnerable Livelihoods in Somali Region, Ethiopia

Population and demography in Somali Region

Government policy towards the IDP camps

Livelihood vulnerability and mortality outcomes

IDS RESEARCH REPORT 57

Section 4 Responding to vulnerability

Coping strategies adopted in Somali Region

Informal transfer mechanisms in Somali Region

Section 5 Conclusions and implications for policy

IDS RESEARCH REPORT 57

Box 13.1 Case study: household drought coping strategies

IDS RESEARCH REPORT 57

Pastoral and Rural Development

IDS RESEARCH REPORT 57

analysing the data and writing up

relatively wealthy in terms of

IDS RESEARCH REPORT 57

which might partly explain why the people of Shinile are

selling baskets and mats made

Livestock marketing, mainly across the border with Somalia

the supply and quality of animals offered, and export routes

numbers fluctuate. Fieldwork for

kerosene) in their home.

IDS RESEARCH REPORT 57

on the livelihood of others. Pastoralists negotiate access to

people claim to be struggling today than ten years ago, which

arid areas, and local livelihoods are sensitively adapted to the

Closely related to drought is access to water, and this study

disputes can escalate into violence.

khat (a mild stimulant), leaving their wives to find income

IDS RESEARCH REPORT 57

American wheat, tons of food aid is also sold on local

from informal Koranic teachers) in rural areas. Our

Some parents see little need for pastoralists to be educated,

Addressing vulnerability also requires recognising the

IDS RESEARCH REPORT 57

1.1 Research issues

prone region, have become relatively wealthy.

IDS RESEARCH REPORT 57

most extensive to be undertaken in the region for many

1.2 Structure of this report

vulnerability into formal and informal responses. Chapter 13

1.3.1 Research methods

The quantitative and qualitative research instruments were

IDS RESEARCH REPORT 57

Household questionnaire survey

wealth ranking, seasonal calendars

Trade and marketing survey

population into representative subsamples. Unfortunately, both these prerequisites are

difficulties of accessing several districts within Somali

It follows that it is not possible to implement a sample survey

IDS RESEARCH REPORT 57

Table 1.2 Sampling frame for the household survey in Somali

zones, six as agro-pastoralist zones, and three as crop farming