DETERMINANTS AND DIMENSIONS OF HOUSEHOLD FOOD

INSECURITY IN DIRE DAWA CITY, ETHIOPIA

A Thesis Submitted to the

Department of Agricultural Economics, School of Graduate Studies
ALEMAYA UNIVERSITY

In Partial Fulfillment of the Requirements for the Degree of

MASTER OF SCIENCE IN AGRICULTURAL ECONOMICS

By

ASCHALEW FELEKE ASFAW

January 2006
Alemaya University
 SCHOOL OF GRADUATE STUDIES
ALEMAYA UNIVERSITY

As members of the Examining Board of the Final M. Sc. Open Defense, we certify that we
have read and evaluated the thesis prepared by: Aschalew Feleke Entitled: Determinants and
Dimensions of Food Insecurity in Dire Dawa City, Ethiopia and recommended that it be
accepted as fulfilling the thesis requirement for the degree of: Master of Science in
Agricultural Economics

________________________________ ______________ ________________

Name of Chairman Signature Date

________________________________ ______________ ________________

Name of Internal Examiner Signature Date

_________________________________ ______________ ________________

Name of External Examiner Signature Date

Final approval and acceptance of the thesis is contingent upon the submission of the final copy
of the thesis to the Council of Graduate Studies (CGS) through the Departmental Graduate
Committee (DGC) of the candidate’s major department.

I hereby certify that I have read this thesis prepared under my direction and recommend that it
be accepted as fulfilling the thesis requirement.

________________________________ _____________ ______________

Name of Thesis Advisor Signature Date

________________________________ ______________ ______________

Name of Thesis Co-Advisor Signature Date

ii
 DEDICATION

I dedicate this thesis manuscript to my mother BIZUNESH KEBEDE and my late father
FELEKE ASFAW for nursing me with affection and love and for their dedicated partnership
in the success of my life.

iii
 STATEMENT OF AUTHOR

First, I declare that this thesis is my bonafide work and that all sources of materials used for
this thesis have been duly acknowledged. This thesis has been submitted in partial fulfillment
of the requirements for an advanced MSc degree at the Alemaya University and is deposited at
the University Library to be made available to borrowers under rules of the Library. I
solemnly declare that this thesis is not submitted to any other institution anywhere for the
award of any academic degree, diploma, or certificate.

Brief quotations from this thesis are allowable without special permission provided that
accurate acknowledgement of source is made. Requests for permission for extended quotation
from or reproduction of this manuscript in whole or in part may be granted by the head of the
major department or the Dean of the School of Graduate Studies when in his or her judgment
the proposed use of the material is in the interests of scholarship. In all other instances,
however, permission must be obtained from the author.

Name: Aschalew Feleke Signature:______________

Place: Alemaya University, Alemaya
Date of Submission: January 2006

iv
 BIOGRAPHICAL SKETCH

The author was born in 1969 at Dogu town, Girawa Woreda, East Hararghe Zone of Oromiya
Regional State. He attended his elementary and secondary school education at Dogu
elementary and Girawa junior and senior secondary schools.

After successfully passed the Ethiopian School Leaving Certificate Examination, he joined the
then Alemaya University of Agriculture in 1986/87 academic year and graduated with B.Sc.
Degree in Agricultural Economics in 1990.

The author has more than fourteen years of progressive working experience both in
Government and Non-Government Organizations.

He joined the School of Graduate Studies of the Alemaya University in the 2003/2004
academic year as a graduate student for the Master of Science Degree in Agricultural
Economics.

v
 ACKNOWLEDGEMENTS

Above all I would like to thank the Almighty God for his unreserved gift.

I would like to express my deep gratitude to my research advisor, Dr. Ayalneh Bogale for his
continuous advice and professional guidance towards realization of this work. My heartfelt
appreciation and great thanks go to my co – advisor, Dr. Abebe Haile Gabriel for critically
reviewing the manuscript and for his continuous professional guide on matters pertinent to this
thesis.

My special thanks go to my wife, Aynalem Tesfaye, as well as my brothers Dr. Sisay Feleke
and Fasil Alemayehu for their continuous moral support during the study. I would like to
extend my special thanks to my uncle: Kasu Kebede and his family for their key and principal
role in the success of my life. I am proud of you.

I would like to forward my warm appreciation and great thanks to my friends Abebaw
Shimeles and Helina Mikre for their valuable support during this work. My special
appreciation and warm thanks go to Ahemed Rage, Petros Birhane, Zerihun Bekele, Helina
Fresenbet, Zerihun Ashebir, Teferi Mola, Anteneh Mengiste, Mesfin Eshetu, Tesfaye Bekele,
Simegn Tafesse, Sileshi Deme, Frederick Asamoa, Kidest Kebebd and Konjit Engida for their
unreserved support and encouragement during the study. I am also indebted to my classmates:
Asefa and Daniel for their positive thinking, attitude and interesting period we spent together.

vi
 ACRONYMS AND ABBREVATIONS

ACCORD Agency for Cooperation & Research in Development

AE Adult Equivalent
CSA Central Statistical Authority
C Contingency Coefficient
EHNRI Ethiopian Health & Nutrition Research Institute
FAO Food and Agricultural Organization
FDRE Federal Democratic Republic of Ethiopia
FGT Foster, Greer and Thorbeck
FS Food Secure
FI Food Insecure
GER Gross Enrollment Ratio
HH Household
HICE Household Income, Consumption & Expenditure
IFAD International Fund for Agricultural Development
ILO International Labor Organization
Kcal Kilo calorie
masl meter above sea level
MEDaC Ministry of Economic Development & Cooperation
MoFED Ministry of Finance and Economic Development
MoPED Ministry of Planning & Economic Development
SD Standard Deviation
UN United Nations
UNCHS United Nations Center for Human Settlement
UNICEF United Nations Children Fund
UNDP United Nations Development Programme
VIF Variance Inflation Factor
WM Welfare Monitoring

vii
 TABLE OF CONTENTS

STATEMENT OF AUTHOR iv
BIOGRAPHICAL SKETCH v
ACKNOWLEDGEMENTS vi
ACRONYMS AND ABBREVATIONS vii
1. INTRODUCTION 1
1.1. Background 1
1.2. Statement of the Research Problem 4
1.3. Objectives of the Study 6
1.4. Significance of the Study 7
1.5. Scope and Limitation of the Study 7
1.6. Organization of the Thesis 8
2. REVIEW OF LITERATURE 9
2.1. Concepts and Definition of Food Security 9
2.2. Measurement and Indicator of Food Security 11
2.3. Determinants of Food Security 17
2.4. Urban Livelihood and Food Security Shocks 22
2.5. Empirical Evidences 23
3. DESCRIPTION OF THE STUDY AREA 26
4. RESEARCH METHODOLGY 29
4.1. Sampling and Data Collection 29
4.2. Method of Data Analysis 31
4.3. Definition of Variables and Hypothesis 37
5. RESULTS AND DISCUSSION 41
5.1. Demographic and Socio-Economic Characteristics 41
5.1.1. Age and sex composition 41
5.1.2. Characteristics of household by headship 42
5.1.3. Household food security status and family size 44
5.1.4. Food security status by kebeles 45

viii
 TABLE OF CONTENTS (Continued)

5.1.5. Household food security status and headship 46

5.1.6. Household food security and marital status 47
5.1.7. Household food security and education of household head 48
5.1.8. Household food security and urban agriculture 49
5.1.9. Household food security and occupation 50
5.2. Household Income Source and Food Security 51
5.2.1. Household income sources and level 51
5.2.2. Household food security status by number of income source 53
5.2.3. Household food security and daily income per adult equivalent 54
5.3. Household Food Security and Expenditure 55
5.3.1. Household food security and daily food expenditure per adult equivalent 58
5.3.2. Household food consumption 59
5.4. Determinants of Food Insecurity 59
5.4.1. Explanation of significant independent variables 63
5.5. Extent and Incidence of Food Insecurity 66
5.5.1. Extent and severity of food insecurity 67
5.5.2. Incidence of food insecurity and household characteristics 68
6. SUMMARY AND POLICY RECOMMENDATIONS 70
6.1. Summary 70
6.2. Policy Recommendations 72
7. BIBILIOGRAPHY 74
8. APPENDICES 79

ix
 LIST OF TABLES
Table Page

1 Factors contributing to urban food insecurity --------------------------------------- 21

2 Classification of urban food security shocks ---------------------------------------- 23
3 Distribution of sampled kebeles and households ----------------------------------- 30
4 Characteristics of households by age and sex --------------------------------------- 42
5 Descriptive characteristics of household by headship------------------------------ 43
6 Household food security status by family size--------------------------------------- 44
7 Household food security status by adult equivalent -------------------------------- 45
8 Household food security status by kebeles ------------------------------------------ 46
9 Household food security by sex of household head -------------------------------- 47
10 Household food security and marital status of head ------------------------------- 47
11 Household food security by educational status of household head --------------- 49
12 Household food security and urban agriculture ------------------------------------- 50
13 Household food security status by occupation ------------------------------------- 50
14 Household income source and level ------------------------------------------------- 52
15 Mean monthly household income by income group -------------------------------- 52
16 Household food security and number of income sources ------------------------- 53
17 Household food security and daily income per adult equivalent------------------ 54
18 Monthly household food and non- food expenditure ------------------------------ 55
19 Monthly household income and expenditure proportion -------------------------- 57
20 Household food security status by daily food expenditure ------------------------ 58
21 Contingency coefficient value of dummy variables -------------------------------- 59
22 Variable inflation factor of continuous variables ----------------------------------- 60
23 Types, codes and definition of variables in the model ----------------------------- 61
24 The maximum likelihood estimates of the logit model ---------------------------- 62
25 Incidence of food insecurity ----------------------------------------------------------- 69

x
 LIST OF TABLES IN THE APPENDEIX

Appendix Table Page

1. Summary of survey questionnaire ----------------------------------------------80

2. Caloric value of food consumed by sample households -------------------------- 81
3. Conversion factor used to calculate adult equivalent ------------------------------82

xi
 DETERMINANTS AND DIMENSIONS OF HOUSEHOLD FOOD
INSECURITY IN DIRE DAWA CITY, ETHIOPIA

ABSTRACT

The problem of urban food insecurity in Ethiopia has received little attention despite its
increasing importance and far reaching ramifications. High rates of unemployment resulted
from sluggish economic growth and low level of investment in urban centers combined with
high population growth has negatively affected income earning potentials of the residents.
Urban population will continue to grow with its associated problems unless meaningful
development interventions backed by research findings are undertaken.

The present study attempts to examine the food insecurity situation, estimate food insecurity
gap and severity and identify the determinants of food insecurity in Dire Dawa town at
household level. The primary data source for this work was the Dire Dawa urban household
socio-economic data collected by undertaking a survey on 200 households. Descriptive
statistics and econometric methods were employed for analysis purpose to meet the stated
objectives.

The descriptive statistics revealed a significant mean difference at an acceptable significance

level between the food secure and food insecure households in terms of household size, adult
equivalent, number of income sources, daily income per adult equivalent and daily food
expenditure per adult equivalent. Education and occupation of household head were also
statistically significant at 5 and 1 percent significant level respectively.

A binary logit econometric model has identified seven out of nine variables included in the
model as significant. Household size, daily income per adult equivalent, proportion of food
expenditure, household head education, sex of household head, access to credit and marital
status of the household head were found to be significant determinants of food insecurity in the
study area. The head count ratio computed by FGT model revealed that 43 percent of sample

xii
households were food insecure. The food insecurity gap and severity were 13 and 5.9 percent
respectively.

The findings suggest the following set of policy recommendation. Action based awareness
creation on the impacts of population growth at family, community and national level,
provision of technical skill training to the unemployed that enhance the job creativity of the
participants should be promoted extensively. Access of credit to the needy and trained people
needs to be provided with proper targeting criterion. Education, which is the basis for any
development, should be strengthened with particular focus on vocational training.

xiii
xiv
 1. INTRODUCTION

1.1. Background

The dramatic growth of cities in the developing world has brought with it a new challenge –
widespread and increasing urban poverty. However efforts to address the unique problems of
urban poverty lag far behind the growth of the problems themselves. Antipoverty initiatives
have traditionally targeted rural areas, which were presumed to have been worse off than
urban areas. But the problems of poor city dwellers have become more pressing including the
issues of how the urban poor earn their livelihood and the ways in which this affects key
indicators of human welfare, such as food security and nutrition (Maxwell et al, 2000).

No developing country can afford to ignore the phenomena of urbanization, which will be one
of the strongest social forces in coming years, especially in developing countries. Within the
next 20 years, more poor and undernourished people in developing countries will live in the
cities than rural areas. High rates of urbanization mean that urban food insecurity and
malnutrition are concerns even for regions like Africa and Asia, where current levels of
urbanization are relatively low (Engle et al, 1998).

Sixty-six percent of the world’s urban population lives in developing countries, and this will
increase to 80 percent by the year 2030. Within developing countries, 38 percent of population
now lives in urban areas. The urban population in developing countries is growing three times
faster than the rural population (UN, 1998)

Urban populations in Africa are growing rapidly, and inequality is increasing. The major urban
food problems of the 1970s and 1980s, food shortage and price shocks, have apparently been
largely resolved at least in the short to medium term. Because of this urban food security
having long been defined as the issue of feeding the cities has dropped off the political agenda
of urban planners and managers; indeed specifically urban food security problems receives
relatively little attention from the national food or nutritional policy planners. Urban food
security problems in Africa receive little attention partly because it tends not to be linked to
seasonal or community wide process and partly because of a long – held belief that urban
populations are better off but urban food insecurity is directly linked to urban poverty and
inequality and for this reason research on urban food security must focus on the question of
access to food (Maxwell, Undated).

Although Ethiopia is one of the most populous countries in Africa, it is also one of the least
urbanized African countries. The level of urbanization reached 17 percent in 2002. However,
this level is expected to reach nearly 30.1 percent by the year 2020, as the urban areas are
currently growing at around 6 percent per year. The effects of natural population growth and
growing rural - urban migration have been felt in terms of poor urban management, lack of
infrastructure and inadequate service delivery. Drought and war have also contributed to the
high population influx in to the cities and towns, which in turn contributed to the deterioration
of infrastructure and service. Slow economic growth and the low level of investments in urban
centers combined with high population growth, have resulted in high rates of unemployment
and the inaccessibility and inadequacy of existing services for low income groups, which
further exacerbated urban poverty (FDRE, 2002).

A combination of factors has resulted in serious and growing problem of food insecurity in
Ethiopia. Adverse climatic changes (drought) combined with high human population pressure,
environmental degradation, technological and institutional factors have led to a decline in the
size of per capita land holding in rural areas. This was exacerbated by policy induced
stagnations of agriculture and internal conflict and instability in the past resulting in to the
widening of the food gap for more than two decades, which had to be bridged by food aid. On
average at any year 4 million people need food aid (FDRE, 2002). In the last twenty years
(1981 – 2001), on average 5.3 million people were affected by drought (FDRE, 2001).

Ayalneh (2002) noted that food availability in Ethiopia is to a great extent determined by
domestic staple food production by subsistence agriculture. There is growing evidence
confirming that even in the ‘normal’ years, Ethiopia has failed to domestically produce enough
food to meet the subsistence needs of its population. The path towards food security is often

2
constrained by major obstacles of land degradation and natural hazards, armed conflict,
poverty, population growth and failure in governance to mention few. He further explained
that during the last two decades population has continued to grow with an average rate of 2.9
percent, whereas food crop production has grown at an annual rate of 2.5 percent. The
immediate effect of this is reflected in a declining trend of per capita agricultural production
index.

A report by Ethiopian Economic Association (2004) noted that the total food production
during the subsequent four years of 2000s with the exception of 2002/03 drought year had
been reasonably good when compared to the last three years of 1990s. However, agricultural
growth in agricultural production is still lagging behind population growth. In terms of per
capita food production average national food crops production between 2000/01 and 2003/04
was only 145.5 kilogram of grain per person, while it fluctuated between 106 and 165
kilogram during these four years.

Urban food insecurity in terms of access at the household or individual level has increased as
urban poverty and inequality have increased. Under circumstances where low-income urban
populations are spending up to three quarters of their total income on food, the issue of income
and livelihood are directly linked to food security. In the earlier era, urban food security in Sub
- Saharan Africa was characterized by short term, acute crises, but has become a chronic
problem experienced mainly by the poor (Maxwell, 1998).

A significant proportion of household expenditure in Ethiopia goes to spending on food. Food

expenditure on average accounted for about 61 percent of household’s budget. Rural
households spend a little over 68 percent of their budget on food while spending by their urban
counterparts stood at about 55 percent. Rural households spend about 33 percent of their food
budget on cereals while their urban counterparts spend only 20 percent. With regard to non –
food category house rent, construction materials, water, fuel and power accounted for 16 and
19 percent of expenditure in rural and urban areas respectively (MoFED, 2002).

3
The issue of household food security needs the participation of all development partners
including the involvement of food insecure households. The involvements of these
development partners demand area specific information on the determinants and dimensions of
food insecurity.

1.2. Statement of the Research Problem

Food insecurity in urban areas takes place in an environment of diverse condition, problems
and actors. To meet the rising challenge of urban food insecurity and malnutrition
governments, communities and aid agencies must work together. But improved responses to
these problems require more information about their causes and better analysis of what
programs and policies are most effective (IFPRI, 2002).

Edilegnaw (1997) explained the paramount importance of information about the relationship
between food consumption and different household characteristics in the design of food
policies.

The report by UNICEF represents one of the few attempts to evaluate program experiences in
urban areas and draw cross- country conclusion, but much more can be done to understand the
nature of urban poverty, food insecurity and malnutrition, their determinants, and the
successful and unsuccessful responses to it (UNICEF, 1994; cited in Haddad et al, 1999).

An increasing trend of the share of the poor population in urban areas, high inequalities in
income distribution, high population growth rates, high food prices and low job creation all
combine to threaten the food security of the urban poor. This presents a new challenge to
policy makers of balancing their efforts to address not only rural poverty but also urban
poverty. More importantly the issues of urban poverty such as food and nutrition insecurity
need attention (Ssewanyana, 2003).

Policy makers and programme administrators cannot simply transfer their programme from
rural to urban areas because the levels if not the magnitude of the effects, of the determinants

4
are different in each location. They must understand specific community level conditions so
that they can identify which of the key variables programs must address (Garrett et al, 1999).

The 1999/00 HICE and WM based poverty analysis showed that overall consumption poverty
measured by the head count ratio has witnessed a 1.3 percentage point decline between
1995/96 and 1999/00. With regard to rural areas consumption poverty has declined by 4.2
percent while urban areas witnessed 11 percent increases during the same period. Depth and
severity of consumption poverty has also shown improvement in rural areas while a slight
deterioration has been observed in urban areas. In urban areas female headed households have
been found to have higher poverty incidence, depth and severity than their male counterparts.
(MoFED, 2002).

The incidence of poverty in Addis Ababa indicates that in 1997, 49.5% of the population is
below the food poverty line, which means that they did not have sufficient income to purchase
the minimum food necessary to provide essential energy required for a healthy life. About
51.4 percent of the population lies below total poverty line that indicate they can not afford to
buy the minimum food basket and basic non – food items. The incidence of poverty has
declined slightly from 1995 where it was 61 percent (Mekonnen, 1997; cited in UNCHS,
2000).

The report on household food security study of four towns of Ethiopia has thrown light on the
magnitude of urban food insecurity. Considering an intake of 2100 calories per person per day
and the minimum cost prevailing in each location, households with insufficient purchasing
power in Dire Dawa were identified to be 47 % (MoPED, 1994).

Study result of MoFED (2002) Welfare Monitoring Unit showed that for the period 1995/96
and 1999/00 the calorie intake per adult per day in urban Dire Dawa is estimated to be 1831
and 1929 kcal respectively which is below the minimum requirement of 2100kcal. All urban
households (rich and poor) are averaged out to provide one single estimate of the caloric
intake.

5
The Central Statistics Authority (1999) report indicated that population growth rate in urban
Dire Dawa with the medium variant was estimated to be about 4.1 for the period between 2005
– 2010. With regard to urban unemployment CSA (2004) report revealed the unemployment
rate of urban Ethiopia in 2004 was 22.9 percent and the rate was highest in urban areas of Dire
Dawa (33.5%) followed by Addis Ababa (22.1%).

This increase in urban population outstrips the capacity of the city to provide employment
opportunity because it can not absorb all the additional supply of labor coming from other
areas. Although seventy four percent of regional population resides in urban Dire Dawa ear-
tagged food security budget which was allocated for the Dire Dawa Administration Council
has not ever been allocated for the urban areas employment generation activity. The perception
of policy makers and other development agents that rural conditions are much worse than
urban ones does not mean that resources should not be directed to the urban poor.

Preparation and implementation of different policies to improve the livelihood and food
security situation based on only past research findings might ultimately lead to erroneous
outcomes. This is mainly because of the fact that socio-cultural, political and economic
features might be changed overtime especially in the urban setting. Therefore this study was
envisaged to narrow the existing information gap and capitalize on the existing ones so that
properly advised policies could be sought.

1.3. Objectives of the Study

This study has the following objectives.

1) To examine the food insecurity situation & to estimate the food insecurity gap and its
severity; and
2) To identify the determinants of food insecurity of the Dire Dawa town

6
1.4. Significance of the Study

Urban population will continue to grow, and problems will also grow with them unless action
is started now. Meaningful development interventions should be based on informed decisions
backed by research findings.

Understanding the level and incidence of food insecurity can enable policy makers or decision
makers and development actors to make informed decisions and actions. Area specific
identification of major determinants of the food insecurity will ease the implementation of
different development projects in the Dire Dawa town.

Though both the problem of urban food insecurity and the affected population have been
increasing progressively, urban food insecurity studies were not promoted in the country.
Research findings and information in general on factors affecting food insecurity in urban Dire
Dawa is scarce; currently most programmes and projects are designed to alleviate food
insecurity in rural areas.

In situation where urban food security is emerging as an important area of concern, this study
will make a modest contribution to attract attentions of development partners and to make
targeted intervention especially by city administration.

This study attempts to address those research questions related to food insecurity, and aims to
highlight on the findings for improving food insecurity in Dire Dawa town. It also aims to
provide insight in to the factors that affect food security, in ways that can assist policy makers
and programme administrators to act in a more effective manner so as to reduce food
insecurity in urban areas of Dire Dawa.

1.5. Scope and Limitation of the Study

The study specifically focused on identifying major determinants of food insecurity at

household level by comparing direct calorie consumption per adult equivalent with the

7
minimum requirement by classifying sample households as food secure and insecure and then
looking at the extent of food insecurity in Dire Dawa town.

The major constraint has been shortage of financial resources (it was undertaken through
researcher’s personal means) and shortage of time; as a result, it could not venture to
investigate the wider social and environmental dimensions of food insecurity. Only
dimensions of food insecurity in terms of incidence and severity were investigated. It has not
also been possible to delve in to variations among households in terms of variables reflecting
quality differences. Moreover, the survey sample did not include urban residents who are
classified as street children and homeless population.

1.6. Organization of the Thesis

The rest part of this thesis is organized in to five parts. The second part deals with review of
literature that includes theoretical framework of food security and empirical studies made in
the country and elsewhere in the world. The third part touches the brief description of the
study area. The fourth part discuss about the methodology employed in data collection and
analysis method. Part five goes on dealing with the results and discussions of descriptive and
model results of the research and finally part six presents summary and recommendations
based on the findings of the research.

8
 2. REVIEW OF LITERATURE

2.1. Concepts and Definition of Food Security

The development of the concept of food security was launched at the time with a relatively
clear focus on national and international food supply. In the 1970s, food security was mostly
concerned with national and global food supply and then in the 1980s the focus shifted to
questions of access to food at household and individual levels (Maxwell and Smith, 1992;
Wiebe and Maxwell, 1998).

Available evidence indicates that during the last two decades, there has been an increasing
trend in per capita food output in the world. In contrast, a significant proportion of the
population, particularly in the developing world, has been suffering from hunger and
malnutrition. In 1990, for example the calorie supply at the global level was more than 10
percent compared to the total requirement. However during the same period, more than 100
million people were affected by famine and more than a quarter of the population was short of
enough food (UNDP, 1992; cited in Debebe, 1995). These facts show that availability at
international level does not guarantee acquisition of food at national or household level.

The food security concept addresses people’s risk of not having access to needed food. These
risks can arise from income or food production. Even in normal situation with no crises such
as war or shocks such as sudden price changes, these risks typically are higher the closer a
household is to inadequate dietary intake. Thus, at the household level, food security is the
ability of the household to secure enough food to ensure adequate dietary intake for its entire
member (Von Braun et al, 1992).

World Bank (1986) defined food security as access by all people at all times to enough food
for an active, healthy life. Its essential elements are the availability of food and the ability to
acquire it. Food insecurity is the lack of access to enough food.

9
The same source classified food security as chronic and transitory. Chronic food insecurity is a
continuously inadequate diet caused by the inability to acquire food. Chronic food insecurity
affects households that persistently lack the ability either to buy enough food or to produce
their own. On the other hand transitory food insecurity is a temporary decline in a household’s
access to enough food. It results from instability in food prices, food production or household
incomes.

Von Braun et al (1992) defined food security as access by all people at all times to the food
required for a healthy life. Access to the needed food is a necessary but not sufficient
condition for a healthy life. A number of other factors, such as the health and sanitation
environment and household or public capacity to care for vulnerable member of society also
come into play.

There are four core concepts, implicit in the notion of “secure access to enough food all the
time.” These are (a) access to enough food, defined by entitlement to produce, purchase or
exchange food or receive it as a gift; (b) sufficiency of food, defined mainly as the calories
needed for an active, healthy life; (c) security, defined by the balance between vulnerability,
risk and insurance; and (d) time, where food insecurity can be chronic, transitory or cyclical
(Smith et al, 1992).

The World Bank definition of food security contains two features that help to sharpen focus on
access to food. First, it requires that access be sufficient for activity and health. Sufficiency is
usually measured in terms of caloric intake relative to physiological requirements for a
specified period of time. Requirements vary with individual characteristics such as age, sex,
and level of physical activity and with environmental characteristics such as climate and
quality of water and health to which the household has access. Secondly, it must be
sustainable. A household can hardly be considered food secure if it is able to meet its current
nutritional requirement only by selling or depleting its endowment of resources. The other way
sufficient access at all times can be interpreted is that access to food be sufficient under all
possible circumstances within any particular period of time, which bring us to the notion of
vulnerability (Wiebe and Maxwell, 1998).

10
The risk profile of individual households and communities will be determined by the channels
through which their access to food is normally mediated and by the assets which are available
to them as buffers. The most food insecure households will be those facing the greatest
probability of an entitlement failure with the least assets. If the risk should materialise, these
households will have no choice but to bear the costs of an entitlement failure in the form of
reduced dietary intake, either in the current time period or in the future. Even where asset
holdings are larger, households may be reluctant to dispose of productive assets to safeguard
current food consumption, because of the opportunity cost in terms of future food access.
However, there will come a point when it is no longer rational to protect future entitlement by
under consumption if the household will not survive the current period by doing so (Smith et
al, 1992).

Generally, available literature on food security revealed the definition of food security in terms
of access by all household members at all times as central core being other condition also
deemed necessary for healthy life. Food security can be explained as access of food by all
people to the required dietary intake through various means. It touches the supply in terms of
availability and capacity of people to obtain sufficient amount through their own ways.

2.2. Measurement and Indicator of Food Security

Measurement is necessary at the outset of any development projects to identify the food
insecure, to assess the severity of their food shortfall, and to characterize the nature of their
insecurity (Hoddinot, 2001).

There is no single indicator that best measure household food security, so several indicators
are used. One common indicator is caloric adequacy (Payne, 1990; Habicht and Pelletier,
1990; Maxwell and Frankberger, 1992; Kenedy et al, 1994; Maxwell, 1996; Chung et al,
1997; cited in Maxwell et al, 2000). This measure captures food sufficiency in terms of
quantity but does not address the quality of the diet or issue of vulnerability or sustainable
access. The traditional approach of measuring food security using dietary intakes has been to
select an optimal caloric intake based on a recommended daily allowance for the equivalent of

11
a moderately active adult and compare it to observed household caloric intake per adult
equivalent (Maxwell et al, 2000).

According to Frankenberger (1992) household food security indicators are divided in to

process, access and outcome indicators. He further explained each indicator as:

a) Process indicators reflect food supply/availability that includes inputs and measures of
agricultural resources, institutional development and market infrastructures and
exposure to regional conflict and its consequences.
b) Access indicators are various means or strategies used by households to meet their
household food security needs.
c) Outcome indicators can be grouped in to direct and indirect indicators. Direct
indicators of food consumption include those indicators, which are closest to actual
food consumption rather than marketing channel information or medical status.
Indirect indicators are related to nutritional status assessment and generally used when
direct indicators are either unavailable or too costly (in terms of money and time) to
collect.

However, Frankenberger (1992) ultimately classified these indicators in to two main

categories, process and outcome indicators. The former provides an estimate of food supply
and food access situation and the latter serves as proxies for food consumption.

The same source after making distinction between ‘’process indicators’’ and ‘’outcome
indicators’’ explained each accordingly where process indicators were grouped in to indicators
that reflect food supply such as meteorological data, information on natural resources,
agricultural production data, food balance sheet etc and those reflect food access. On the other
hand, outcome indicators can be grouped in to direct indicators, which are closest to actual
food consumption, rather than indirect indicators focusing on storage estimates, subsistence
potential ratio and nutritional status assessment.

12
The level of food insecurity can be measured by employing various methods that depends on
the availability of resources, type of research and its purpose. Regarding measurements of
food insecurity Debebe (1995) noted that despite the shortcomings in the measurements,
qualitative and quantitative approaches or a combination of both have been used to identify
and apply food security indicators. There is no fixed rule as to which method to employ due to
the diversified characteristics of food insecurity and the different level of consideration. The
decision to rely on a particular method usually depends on resource and time constraints,
objectives of the study, availability of data, type of users and degree of accuracy required.

The measurement of food insecurity at different levels was described by Von Braun et al
(1992) as follows:

Country level: Food security at the country level can be monitored in terms of demand and
supply indicators; that is, the quantity of available food versus needs, and net imports needed
versus import capacity.

Household level: Food security at household level is best measured by direct surveys of
dietary intake in comparison with appropriate adequacy norm. However, it measures existing
situation and not the down side risks that may occur. The level of, and changes in, socio
economic and demographic variables such as real wage rates, employment, price ratio, and
migration properly analyzed, can serve as proxies to indicate the status of and changes in food
security. Indicators and their risk patterns needed to be continually measured and interpreted
to monitor food security at the household level.

Individual level: Anthropometric information can be a useful complement because

measurements are taken at the individual level. Yet such information is the outcome of change
in the above indicators and of the health and sanitation environment and other factors. Most
important, this information indicates food insecurity after the fact.

There is little correlation between a large set of process indicators and measures of food
security outcomes. The finding echoes the conclusion of some development agencies, namely

13
that there is little correlation between area level food production and household food security
(IFAD, 1997; Chung et al, 1997; cited in Hoddinot, 2001). When the two indicators are
compared for food security measurement at household level, direct outcome indicators are
preferable that indicate actual consumption by the household.

Direct outcome indicators are preferred than area level production due to the fact that the latter
showed only availability or supply in the area which doesn’t guarantee acquisition of food at
the household level. If a household has no purchasing or getting through other means
accessibility by the household was denied which is central in household food security
definitions.

Area level food production can not guarantee the availability at the household level due to lack
of access by the household. Area level production ensures availability at the households only
when they have purchasing power at the prevailing price, get through exchange or other
means, as a result outcome indicators are preferable than process indicators for food security
assessment.

Food security is such a complex notion that is virtually impossible to measure directly, and a
variety of proxy measures have been suggested. The most frequently used measures include
consumption and expenditure, nutritional status, coping strategies, and resource related
correlates. Anthropometric measures of nutritional status are used as food security indicators.
However, because various factors in addition to food security are determinants of nutritional
status, its use as a food security indicator requires that these factors be taken in to
consideration, including health, intrahousehold distribution of food, and maternal care and
time allocation (Wiebe and Maxwell, 1998).

Income or consumption has been traditionally used as measures of material deprivation.

Consumption is typically preferred over income as the former better captures long run welfare.
Consumption may also better reflect households ability to meet their basic needs. Income is
one of the factors that enable consumption, though consumption also reflects a household’s
access to credit and saving at times when their income is too low. Hence, consumption is a

14
better measure of a household’s welfare than income. Moreover, in a developing country
setting, households are likely to underreport their income level more than they do with their
consumption (MoFED, 2002).

Several methods have been used to measure food consumption; the first method is to estimate
both total household production and purchase over a period of time and growth or depletion of
food stocks held over that period of time. Food entering the household possession and
disappearing is presumed to have been consumed. The second estimate is to undertake 24 –
hour recalls of food consumption for individual members of a household and analyze each
type of food mentioned for caloric content (Bouis, 1993; cited in Wiebe and Maxwell, 1998).

Comparison of methods to measure the state of food insecurity at household level in terms of
costs, time requirement, skill and susceptibility to misreporting among four outcome indicators
namely individual intake, household calorie requisition, dietary diversity and index of coping
strategies, household caloric acquisition are found to be better measurement (Hoddinot,
2001). Hoddinot briefly outlines the four outcome measures of household food security as
follows:

Individual intake: This is a measure of the amount of calories or nutrients consumed by

individual in a given time period usually 24 hours. Methods of generating data with this
method are that an enumerator resides in the household through out the entire day, measuring
amount of food served to each person. The enumerator also notes the type and quantity of food
consumed outside the household. The second method is recall of the previous 24 hour
consumption for each household member. When implemented correctly, it produces the most
accurate measures of individual caloric intake and it is possible to determine that sufficient
calories are being consumed within the household. Against these advantages, it needs to be
made repeatedly ideally for seven nonconsecutive days and require highly skilled enumerators
who can observe and measure quantities repeatedly and quickly and in a fashion that does not
cause households to alter typical level of food consumption and distribution within the
households. The recall method requires interviewing carefully every household member which
obviously is an extremely difficult task.

15
Household caloric acquisition: Here the person responsible for preparing meals is asked how
much food was prepared for consumption over a period of time. The most knowledgeable
person in the household is asked a set of questions regarding food prepared for meals over
specific period of time usually 7 or 14 days. It requires listing out food types on questionnaire
and distinguishing unambiguously between the amounts of food purchased, prepared for
consumption and the amount food served. This measure produces a crude estimate of number
of calories available for consumption in the household. Because the questions are retrospective
rather than prospective, the possibility that individuals will change their behavior as a
consequence of being observed is lessened. The level of skill required by enumerators is less
than that needed to obtain information on individual intakes.

Dietary diversity: One or more persons within the household are asked about different items
they have consumed in a specified period. Where it is suspected that there may be differences
in food consumption among household members, these questions can be asked of different
household members. Calculating a simple sum of the number of different foods eaten by
that person over the specified period of time or calculating a weighted sum where the weights
reflect the frequency of consumption are used. The disadvantage of this measure is that simple
form does not record quantities. If it is not possible to ask about frequency of consumption of
particular quantities, it is not possible to estimate the extent to which diets are inadequate in
terms of caloric availability.

Indices of household coping strategy: This is an index based on how households adapt to the
presence or threat of food shortages. The person within the household who has primary
responsibility of preparing and serving meal is asked a serious of questions regarding how
households are responding to food shortages. Among several ways of summarizing the result,
counting the number of different coping strategies used by the household is one. The higher
the sum, the more food-insecure the household. Calculating the weighted sum of these
different coping strategies where the weights reflect the frequency of use by the household is
another method. Merits of this measure are that it is easy to implement and it captures the

16
notion of adequacy and vulnerability. As it is subjective measure, comparison across
household or localities is problematic.

Among the four types of food security measurement indicated above the last two methods
didn’t capture consumption of particular quantities and as a result it is not possible to estimate
kilocalorie consumption per household. For household caloric acquisition method, conversion
of gross household food consumption into calories, and dividing the calories figure by the
number of adult equivalents in the household and the number of days in the recall period
results in a concise figure for average calories consumed per adult equivalent per day, which is
then compared with an estimate of caloric requirement. Since focus of unit of analysis for this
study was household level, employing household caloric acquisition has been preferred than
the rest three methods.

2.3. Determinants of Food Security

Food security is composed of availability of food, access to food, and risks related to either
availability or access. National, regional or local availability of food is determined primarily
by food production, stockholding, and trade at any of these levels. Variation in any of these
can contribute to food insecurity. Fluctuations in a country’s capacity to import food also
contribute to food insecurity by affecting the local availability of food, although it may be
available globally (Von Braun et al, 1992).

The access that a household has to food depends on whether the household has enough income
to purchase food at prevailing prices or has sufficient land and other resources to grow its own
food. It can also receive assistance from formal programmes or informal networks to
compensate any shortfall. Factors other than income and prices can also affect household
calorie availability. These factors include household demographic structure, educational level
of household members and location (Garrett and Ruel, 1999).

In urban areas, the main determinants of food availability at the household level are prices and
income, access to home production (urban agriculture), and access to formal and informal

17
transfers. Household behavior, influenced by cultural factors and knowledge, also affect
patterns of food demand and distribution within the household (Engle et al, 1998).

Involvement in cash economy is an inescapable fact of life in urban environments. Whereas

rural people can grow their own food, urban residents must buy most of what they eat. Food
prices are thus especially important to urban food security for an ample supply of food does
not guarantee that the urban poor will be able to afford it. Several factors conspire to raise
urban prices of food. Urban marketing systems in developing countries are often inefficient,
with retail markets frequently small and scattered. Poor people are often unable to lower their
food costs by buying in bulk because they simply do not have enough cash to do so. In
addition, many developing countries have eliminated food subsidies, increasing food prices for
urban dwellers (IFPRI, 2002).

The state of household food security is mainly conditioned by factors, which are related to the
process of food acquisition, household procurement strategies and socio economic conditions
of the society. It is generally affected by two major determinants: availability of food and
accessibility to it. The former is further influenced by the different sources of food and
handling patterns, which facilitate the time dimension of food availability in the household
(Debebe, 1995).

A household ability to achieve food security is derived from the household’s human, material,
and institutional resource base, which are often collectively referred in the literature as “food
security factors.” These factors (Table 1) include the educational and employment status,
household demographics, urban agriculture, assets, saving, formal social assistance or direct
transfer, informal social networks, access to clean water and sanitation and cost of living
(Bonnard, 2000).

The same source indicates that lower educational level and illiteracy are directly related to
food insecurity. This is usually measured just for the head of household, but the educational
status of other household members, especially income earners, is also important. The
possession of assets especially liquid assets is positively correlated with food security. More
urban households save, maintain bank account, and belong to saving and microfinance clubs

18
than rural households. Households with either form of saving tend to be more food secure than
those without convertible assets.

Bonnard (2000) further explained each of food security factors as follow:

Household demographics: The size of the household and the dependency ratio are negatively
associated with food security. Exception to this would include very small households
comprised of one or two elderly people or a solitary widow. Child headed households are
nearly always food insecure.

Educational level: Lower educational level and illiteracy are directly related to food
insecurity. This is usually measured just for the head of household, but the education status of
other household members, especially income earners, is also important.

Employment status: Urban households’ livelihoods are generally heavily dependent on

employment outside the home, and much more so than rural households. This is particularly
true of large industrialized or capital cities with constrained land access and more lucrative
non-farm employment opportunities. Source of income and employment are more diverse in
urban areas. Household members tend to have education and training and are, therefore, able
to collectively undertake a wider range of unskilled, semi-skilled and skilled jobs. The more
complex urban economy offers a greater variation in employment option. In general
households with more diversified income are more inclined to be food secure and able to
better cope with shocks. When unemployment of young adult men is high, young adult male
household members can be a drain on household resources. Besides constituting an extra
mouth to feed, unemployed males have a tendency to drink, take drugs, and become involved
in gangs, crime, and other deviant activities that cause stress within the household unit.

Urban agriculture: It is difficult to generalize about urban agriculture. The opportunities for,
and the role of, urban agriculture in urban food security varies across cities. For households an
expansion in production or rise in produce prices represent an increase in income and an
improvement in food security.

19
Assets: Urban dwellers tend to posses more consumer items such as televisions, radios,
refrigerators, bicycle and furniture than rural residents. The possession of assets particularly
liquid assets is positively correlated with food security. The greater the store of wealth and the
more diverse the store of wealth, the less vulnerable a household is to food insecurity.

Savings: More households save, maintain bank accounts, and belongs to saving and
microfinance clubs than rural households. This is not to say rural households do not save, but
rather that is more common for urban households in the majority of country scenarios
reviewed. While urban households save in the form of cash, rural households can also stock
grain and accumulate livestock as a store of wealth. Households with either form of savings
tend to be more food secure than those without convertible assets.

Formal social assistance: Urban households tend to have greater access to safety net
programmes or formal assistance such as school lunch programmes; supplementary feeding
for infants, children and pregnant and lactating women; health or school fee wavers; and cash
for work

Informal social networks or reciprocity networks: Community organizations, clubs, and

associations as well as family and community support are all part of the social or reciprocal
networks.

Access to clean water and adequate sanitation: The quality of water and sanitation affect food
security through food utilization. Disease inhibits the proper utilization of ingested foods and
results in inadequate nutrition.

Perception of crime: Crime is more of a community than household characteristics, but the
perception of crime alters a person’s perception of opportunities. In general, cities are
associated with more crime than rural areas. Crime tends to affect women more than men to
modify their day-to-day behavior such as where they work, when they commute to and from
work, and when they do their household chores based on their perception of the risk and type
of violence prevalent in their communities. These can constrain their ability to earn income.

20
Table 1. Factors contributing to urban food insecurity

Demographic factors Formal social assistance factors

High dependency ratio Limited or no access to formal assistance programmes
Low educational level of household head and mothers Informal social assistance and network factors
Female or child headed household Limited or no relation to social or reciprocity networks
Length of time in residency No informal transfers from rural relatives
Employment and income factors Health factors
Unemployment of working age members Inadequate access to health care
Single income source (Income not diversified) High level of persistent health hazards
Underemployment of working age members High incidence of illness and death
No urban agriculture Environmental factors
Wealth and asset factors High cost of living
Asset poor High incidence of crime
Lack of diversity in assets, especially liquid asset Concentration of refuges and internally displaced people
Limited or no saving

Source: Bonnard, 2000

21
2.4. Urban Livelihood and Food Security Shocks

Livelihood refers to the way people make a living, and analyzing livelihood systems is the
analysis of the factors involved in the way in which people make a living. ‘Making a living ’ is
largely about generating income, this is a means to an end which also includes aspects of food
security, providing home, health, security (reduced vulnerability to climatic, economic or
political shocks etc), sustainability (the ability to continue to make a satisfactory living) and
power (ICRA, Undated).

Livelihood in African cities no longer revolves around wage earning jobs in the formal or
government and parastatal sectors of the economy. Increasingly people earn their living from
informal wage labor or self – employment (ILO, 1990a; 1990b; cited in Maxwell et al, 2000)

MoFED (2002) indicated that income sources of urban Ethiopia comes from eight sources
namely; wage and salaries (41.4%), household enterprise other than agriculture (30.3%), gift
and remittance (8.6%), house rent and other rent (8%), saving or bank (0.46%), own
agricultural enterprise (4.5%), dividends, profit shares source (0.03%) and other receipt
sources account for (6.7%). When it comes to town specific, income for Dire Dawa is derived
from wage and salaries (44.5%), household enterprise other than agriculture (27%) gift and
remittance (10.5%), house rent and other rent (8.32%), own agricultural enterprise (1.5%),
saving or bank (0.34%), dividends and profit shares (0.05%) and other source of income( 7%).

Food security shock is an event that compromises a household’s food security. A shock can be
viewed as a sudden significant divergence from the norm of a given factor. While households
may have adapted to high food price levels or a series of gradual price increase, a sudden
upward surge in food prices is more difficult to absorb and seriously compromise household
income and hence food security. Food security shocks can be idiosyncratic or covariant (Table
2). The former is a shock that affects an individual household and the latter affects a
community or a whole set of households such as drought and inflation (Mutangadura and
Makaudze, 1999).

22
Table 2. Classification of urban food security shocks

Shocks Idiosyncratic Covariant

Inflation X
Food price increase X
Basic non-food price increase X X
Exchange rate devaluation X
Policies and regulation X
Unemployment X X
Crime X X
Illness/death X X
Social and economic decline X X
Conflict X
Natural disaster X
Source: Mutangadura and Makaudze, 1999

2.5. Empirical Evidences

Some studies were conducted to analyze food security in general in several countries. For
example in Ghana, the calorie gap (the difference between the recommended and reported
calorie intake) at household level is influenced by a set of demographic and household
characteristics (Keyreme and Thorbecke, 1991; cited in Tegegne et al, 1999). In particular, the
study showed that, ceterius paribus, the calorie gap declines with age of household head,
educational level of household head, income and the value assets owned by the family. Lower
calorie gaps are also associated with male-headed households and maturity index, a variable
that measures a contribution of older children relative to adults. On the other hand, calorie gap
widened with dependency ratio.

23
A study by Garrett and Ruel in Mozambique found that in urban and rural areas, much the
same factors determine calorie availability (Garret and Ruel, 1999). Only expenditure and
household size exerts significantly different effects, and even then the magnitude of difference
is not large. The most important determining factors that affect calorie availability are income
(proxied by expenditure), prices and demographics such as household size. Urban dwellers do
seem to be slightly more sensitive to changes in income than rural dwellers. This may reflect
urban residents' lack of natural resources "cushion" to absorb income or price shocks and also
their need to purchase, rather than grow, their own food.

The study of Urban Livelihoods and Food and Nutrition Security in Greater Accra, Ghana
indicated that household food availability is a function of food prices, household
demographics and household tastes and preferences (Maxwell et al, 2000). The result of this
study showed that increase in per capita income, proxied by per capita expenditure and all
other factors constant, are associated with higher household calorie availability. Households
with relatively high numbers of school aged children and men of prime working age have
lower food availability and calorie availability is higher in female-headed households.

The main factors that affect food security in urban Maputo, Mozambique, are poverty, low
family income, low availability of general alimentation at the family level, floods, family
crisis, high unemployment levels and low levels of schooling and training and the absence of a
social security system to alleviate the urban shocks (Mucavele, 2001).

Household food security in urban areas primarily depends on the level of income, the prices of
food and other consumer goods on which the households make outlay. Households with larger
family size, less literate and older household heads are likely to fall in poverty than those with
smaller family size, more literate and younger household heads (MoPED, 1994; MoFED,
2002).

Abbi and Mckey (2003) computed trend of poverty incidence using consumption data of
Ethiopian urban household survey data collected from 1994 up to 1997 and the results show
increasing urban poverty over this period particularly between 1994 and 1995 in the cities in

24
the north (Mekelle, Dessie and Bahirdar) and south (Dire Dawa, Awassa and Jimma). The
incidence of consumption poverty in 1994, 1995 and 1997 were 34.4, 40. 5 and 42.9 percent
respectively. Chronic poverty is strongly associated with households having high dependency
rate and female headed household. The heads of households that are never poor are much less
likely to have no schooling and low levels of education are another strong feature of chronic
poverty. Among the chronically poor households 27.5% of their head work as casual labourers
compared to only 7.7% for the never poor.

Dire Dawa is located in a food deficit region of Ethiopia, as the result of which prices of food
grains and other food items are high almost throughout the year. The gap between the low
income and other income groups with respect to percent of expenditure on food is narrow
probably reflecting the unsatisfied food demand by the urban community in Dire Dawa. The
relative prices of food items are higher as compared to other towns (Jimma, Bahirdar,
Awassa). It may therefore, be expected that the real income of households in Dire Dawa is
lower as a result of which the affordability by households to purchase food is reduced
(MoPED, 1994).

25
 3. DESCRIPTION OF THE STUDY AREA

Dire Dawa town, located in the eastern part of the country, has an estimated total land area of
39.54 Km2. It is located at 515 Km east of Addis Ababa between Addis Ababa and Djibouti.
Its altitude is about 1200 masl. Dire Dawa has bimodal rainfall pattern with small rains from
March to April and big rains from August to Mid September. Dire Dawa has an annual rainfall
of about 604mm and mean annual temperature of 24.8 0C.

The foundation of Dire Dawa town is directly related with the construction of the railway line
from Djibouti to Addis Ababa. It was established to serve mainly as railway station in 1902.
The railway administration and the Ethiopian government administered the town jointly until
1926. Since then it came fully under the administration of the Ethiopian government. As time
went on, the population of the town showed significant increase and a master plan, mainly
focused on spatial aspects, was prepared.

In the 1950s and 1960s, Dire Dawa was an industrial town but later on (1970s & 1980s) with
the industry lagging behind, the situation changed and commerce started to play a leading role
in the local as well as the national economy to some extent. It is worth noting at this juncture
that most of the commercial activities that ‘’boomed’ in Dire Dawa in those days were illegal
and macro economically harmful contraband trade.

Dire Dawa is considered as one of the industrial, commercial and service center of the country.
The town hosts a total of seven with three public owned factories and 171 small and micro
enterprises in 2002. The total industrial employment in the town in the same period was 4877
both temporary and permanent workers.

Owing to its strategic location and relatively better transport network, agricultural produce
such as coffee, livestock, fruits, vegetables and chat grown on the eastern part of the country
are some of the major export items to Djibouti, Somalia and Middle East passing through Dire
Dawa.

26
Dire Dawa town has an estimated population of 204,244 of which the female population
comprises the majority (52.3%). The average household size and sex ratio, defined as number
of males per 100 females, is 4.5 and 91.4 respectively. Other demographic indicators revealed
economic activity rate of 48.2, economic dependency ratio 133.7 and unemployment rate of
33.5 (CSA, 2004).

With regard to ethnicity, Oromo, Amhara, Somali, Gurage and others comprises 31, 40, 13.8,
6.5 and 8 percent respectively. In terms of religion composition Orthodox Christian (50%),
and Muslims (46.5%), Chirstian of Protestant and Catholic denomination constitutes 2.1% and
0.5% of the total (CSA, 1994).

Education: Education is the basis for development and it enables a country to identify and
develop untapped natural resources. There are 24 kindergartens, 36 primary and 9 secondary
schools available in the town. Additionally, government owned university, technical college
and 5 private colleges do exist. The Gross enrollment ratio, GER for primary school for
2004/05 academic year was 97.2 percent and 61.7 for the secondary school.

Health: Health service is one of the basic social services, which is required for a healthy and
productive life of any society. The potential health service coverage of the town in 2005 is
estimated to be 100 percent. According to information from Health Bureau currently there are
three hospitals, two health centers, 15 clinics of different grade, 20 pharmacies and one health
post.

Infrastructure: The town has modern postal and telecommunication services including internet,
fax, mobile telephone, and twenty-four hour hydroelectric power supply and international
airport.

Administratively the town of Dire Dawa had been divided in to four ‘Kefitegnas’ and 25
kebeles, however the recent restructuring process has merged the number of kebeles in to eight
with more power and responsibility. The restructuring process of the administration was not
restricted to kebeles only but extended to government offices too. With the newly practiced

27
structure 52 offices/agencies were organized under the City Administration, which is
accountable to Ministry of Federal Affairs.

Municipality, Small and Micro Scale Enterprise and Housing Agency are among the newly
emerged offices trying to create temporary employment opportunity in their construction and
business oriented activities that have impact on reducing household food insecurity. The
livelihood of the residents predominantly depends up on manufacturing, commerce,
government employment, transport service and other sectors.

Since the restructuring of the administration council in 2003, few activities of income and
employment generating have been implemented in the town by city administration. Out of
3218 registered loan needy people, 493 of them have got a total of 1,053,100 Birr through
small and micro enterprise agency. Provision of skill training in the field of metal and wood
work, garment, construction and business were given to 7591 participants. However, it is
worthwhile to mention that selection of training participants and borrowers was not with
specific criterion instead it was based on first comes first served principle. Road side shops
were also started to legalize the informal sector in the city.

28
 4. RESEARCH METHODOLGY

4.1. Sampling and Data Collection

The primary data source for this research was the Dire Dawa urban households’ socio
economic data collected by survey. Wealth ranking of Dire Dawa urban kebeles was done by
ACCORD (1994) for the purpose of identifying intervention areas of urban development
programme. Hundred informants selected from each kebeles have ranked kebeles based on
accessibility, location, infrastructure availability, housing condition, sanitation and other
relevant criterion. The informants were from different social groups, age category and both
sexes. The result obtained from informants had been discussed with government officials and
found being reflected the reality.

Wealth ranking of kebeles was found to reflect social status of residents and convinced to
consider the wealth rank to employ stratified random sampling for this research. The two
conditionalities namely, stratification criterion and number of households in each strata, were
available and employed three stage stratified random sampling technique with probability
proportional to size sampling to draw a total of five kebeles and 200 households. At the
primary stage Dire Dawa kebeles were stratified in to two each having ten and fifteen kebeles
and at second stage probability proportional to size was employed to draw two kebeles from
one strata and three from the remaining ones. Finally, 200 households were randomly selected
from each representative kebeles by using probability proportional to size.

The sample frame used were registered household list collected through census by Population
and Vital Statistics Office of the Administration Council with technical support of Central
Statistical Authority and the list is updated whenever new household come to the town as a
resident and seeks to obtain any service.

29
Table 3. Distribution of sample kebeles and household

Stratum Number Total Sampled Households Sampled

of kebeles households kebeles per kebele households
1 10 16692 01 563 19
12 2134 70

2 15 21227 03 1615 38
10 393 9
20 2746 64
___________________________________________________________________________
Total 38189 5 7451 200

A structured interview questionnaire had been prepared and pilot test of questionnaires was
conducted by trained enumerators after which some minor adjustments were made. The
primary data collected by structured questionnaire can be categorized in to four parts that
focus mainly on those factors believed to have an effect on the state of food security. The first
part covers basic household information covering household size, sex, age, marital status,
educational background, main current activity and religion. The second part consists of food
consumption by households either from purchase and non-purchase during the last seven days
prior to the survey date with total value and non - food expenditure spent during last month.
Third section covers income earner of household member and monthly income level by
source, change in household income for the last five years with main reasons, and length of
residence in Dire Dawa town. Then household major asset holding, involvement in urban
agriculture, credit service and saving issues were covered. The final section touches up on
availability of major infrastructure such as water, electricity, housing and related issues.

In practice, some difficulties were encountered in collecting household data especially from
the assumed better-off one and unavailability of responsible person during day time. For the

30
latter problem, evening time interview were arranged but for the former the researcher
personally had to contact and convince those households by explaining the purpose of the
study (Revealing University’s letter).

To ease and keep uniformity among enumerators, standardizing unit on local measurement of
different food items were carried out and provided to each enumerator.
.
Before analysis of household data was started, quantity of each item consumed during the last
seven days were converted into their equivalent kilocalorie using ENHRI food composition
table for use in Ethiopia (ENHRI, 1998) and household members to their adult equivalent.

4.2. Method of Data Analysis

Food security at household level is best measured by the direct survey of dietary intake (in
comparison with appropriate adequacy norms). The level of, and changes in, socio economic
and demographic variables can be properly analyzed, and can serve as proxies to indicate the
status of and changes in food security (Von Braun et al, 1992).

Food security at the household level is measured by direct survey of income, expenditure, and
consumption and comparing it with the minimum subsistence requirement. In this regard,
income and expenses are used to compute the status of food security. The minimum level of
income, which is required per adult equivalent, was calculated on the basis of amount of food
required by an adult person. The government of Ethiopia has set the minimum acceptable
weighted average food requirement per person per day at 2100 kilo calorie (FDRE, 1996; cited
in Kifle, 1999), which is estimated to be 225 kg of food (grain equivalent) per person per year.
Consequently, a threshold level is set by computing the value of this amount of cereal by the
existing local market price of grain. Thus, those households beyond this thresholds level will
be deemed to be food secured otherwise not food secured.

Having identified the food insecure and food secured groups of households, the next step is to
identify the socio economic characteristics that are correlated with the food insecurity. In light

31
of this, it was hypothesized that there are some household characteristics like household size,
income, household head educational level, etc that have got relative importance in determining
the state of food security at household level.

In order to test the hypothesis a probabilistic model is specified with food security as a
function of series of household characteristics as explanatory variables. The dependent
variable in this case is dummy variable, which takes a value of zero or one depending on
whether or not a household is food insecure. Thus, the main purpose of a qualitative choice
model is to determine the probability that an individual with a given set of attribute will fall in
one choice rather than the alternative (in this study becomes a food insecure or secure).

Regression models in which the dependent is dichotomous can be estimated by linear

probability model (LPM), logit or probit models. Although linear probability model is the
simplest method, it is not logically attractive model in that it assumes that the conditional
probability increases linearly with the value of explanatory variables. Unlike linear probability
model, logit model guarantee that the estimated probabilities increases but never steps outside
the 0 – 1 interval and the relationship between probability ( Pi) and explanatory variable (Xi) is
nonlinear (Gujarati, 1995).

Usually a choice has to be made between logit and probit models, but as Amemiya (1981) has
pointed out, the statistical similarities between the two models make such a choice difficult.
However, Maddala (1983) and Kementa (1986) indicated that many authors tend to agree in
that the logistic and cumulative normal functions are very close in the mid range, but the
logistic function has slightly heavier tails than the cumulative normal distributions.

Gujarati (1995) and Pindyek and Rubinfeld (1981) also illustrated that the logistic and probit
formulations are quite comparable, the main difference being that the former has slightly
flatter tails, that is, the normal curve approaches the axis more quickly than the logistic curve.
Therefore, the choice between the two is one of convenience and ready availability of
computer programmes. Hosmer and Lemeshew (1989) pointed out that a logistic distribution
has got advantage over the others in the analysis of dichotomous outcome variable in that it is

32
extremely flexible and easily used model from mathematical point of view and results in
meaningful interpretation.

Thus, a logistic model was specified to identify the determinants of food insecurity and to
assess their relative importance in determining the probability of being in a food insecure
situation at household level. The analysis of the logistic regression model will show that
changing an independent variable alters the probability that a given individual becomes food
secure, and will help to predict the probability of achieving food security.

Following Gujarati (1995), the functional form of logit model is specified as follows:

1
Pi = E (Y = 1 X i ) = − (β o + β i X i ) (1)
1+ e

For ease of exposition, we write (1) as:-

1
Pi = (2)
1 + e − Zi

The probability that a given household is food insecure is expressed by (2) while, the
probability for not food insecure is:-

1
1 − Pi = (3)
1 + e Zi

Therefore we can write:-

Pi 1 + e Zi (4)
=
1 − Pi 1 + e − Z i

33
Now (Pi/1-Pi) is simply the odds ratio in favor of food insecurity. The ratio of the probability
that a household will be food insecure to the probability of that it will not be food insecure.

Finally, taking the natural log of equation (4) we obtain:-

 P  (5)
L i = ln  i  = Z i = β o + β X + β X + ... + β n X n
1 − P  1 1 2 2

Where P1 = is a probability of being food insecure ranges from 0 to 1

Zi = is a function of n explanatory variables (x) which is also expressed as:-

Z i = β o + β X + β X +...+ β X (6)
1 1 2 2 n n

ß0 is an intercept
ß1, ß2 ------ ßn are slopes of the equation in the model
Li = is log of the odds ratio, which is not only linear in Xi but also linear in the parameters.
Xi = is vector of relevant household characteristics

If the disturbance term (Ui) is introduced, the logit model becomes

(7)
Zi = β o + β 1 X 1 + β 2 X 2 +...+ β n X n + U i

The logit model can not be estimated by the usual ordinary least square method because to
apply OLS we must know the value of the dependent variable ln (Pi / 1 - Pi), which obviously
not known and more over the methods of OLS doesn’t make any assumptions about the
probabilistic nature of the disturbance term. If there is data on individual observations the
method of maximum likelihood can be used to estimate the coefficients of the equation
(Gujarati, 1999 ).

It needs to be clarified that prior to the estimation of the logistic regression model, the
explanatory variables were checked for the existence of multicolinearity. In this study among

34
the other methods Variance Inflation Factor (VIF) was used to measure the degree of linear
relationships among the continuous explanatory variables. Where each continuous explanatory
variable is regressed on all the other continuous explanatory variables and coefficient of
determination for each axillary or subsidiary regression will be computed.

Following Gujarati (1995), VIF is defined as:

 
VIF (X ) =
j
 1
1− R 2


 j 
(8)

Where:

Xj = the jth quantitative explanatory variable regressed on the other quantitative explanatory v
variables.
R2j = the coefficient of determination when the variable Xj regressed on the remaining
explanatory variables.

As a rule of thumb, if the VIF of a variable exceeds 10 that variable is said to be highly
collinear and it can be concluded that multicolinearity is a problem (Gujarati, 1995).

It is also evident that there might be interaction among qualitative variables, which could lead
to the problem of multicollinearity. To detect this problem, contingency coefficients were
computed for each pair of qualitative variables.

The contingency coefficients are computed as follows:

χ2
C= (9)
n + χ2

35
Where, C= coefficient of contingency, χ2 = a Chi-square random variable and n = total sample

size.

To estimate head count ratio, food insecurity gap and to assess the severity of household food
insecurity the Foster, Greer and Thorbeck (FGT) index is employed. This model is recently
used by IFPRI for the analysis of household food insecurity (Hoddinot, 2001).

Several researchers used the FGT model to determine the incidence and severity of poverty
and food insecurity (Edilegnaw, 1997; Ayalneh, 2002; MoFED, 2002; Abebaw, 2003).

The FGT model can be expressed as follows

P(α ) = 1 n ∑ 
(
q m−y
i
)
α
(10)
m 
i =1 

Where: n is the number of sample households

yi is the measure of per adult equivalent food calorie intake of the ith household

m represents the cutoff between food security and insecurity (expressed in caloric
requirement)
q is the number of food insecure households
α is the weight attached to the severity of food insecurity

In equation (10), m - yi = 0 if yi > m

As far as the weight to α is concerned, Hoddinot (2001) further explained that giving no
weight to the severity of food insecurity is equivalent to assuming that α = 0. So then, the
formula collapses to P (0) = q/n, this is called the head count ratio.

36
Giving equal weight to the severity of food insecurity among all food insecure households is
equivalent to assuming that α = 1. Summing the numerator gives the food insecurity gap;
dividing this by m expresses this figure as a ratio. This index p (1) will provide the possibility
to estimate resources required to eliminate food insecurity through proper targeting. That is,
the product (n*m*p1) gives the total calorie commitment required to bring the food insecure
households to the given daily calorie requirement level.

Further giving weight to the severity of food insecurity among the most food insecure
households is equivalent to assuming that α > 1. The most common approach in poverty
literature is to set α = 2, yielding

q 2
P(2 ) = 1 n ∑ (m − y i ) 
 m (11)
i =1  

Hence P (0) is percentage of food insecure households

P (1) is food insecurity gap
P (2) is the severity of food insecurity.

Moreover, based on survey data result socio economic characteristics of sampled households
were described with respect to food security status by employing some descriptive statistics.

4.3. Definition of Variables and Hypothesis

The literature on the determinants of household food insecurity makes it clear that the choice
of dependent and independent variables have been identified by different researchers,
international and national development organizations. This section describes the variables used
in the econometric analysis.

37
Dietary intake is used as a proxy to measure household food security status. Household
consume a variety of food, mainly from purchase that are converted in to their calories using
ENHRI food composition table for use in Ethiopia (1998).

Household food insecurity (HFINS): is a dichotomous dependent variable in the model

taking value of 1 if a household is food insecure and 0 otherwise. Food security status of a
household is identified by comparing total kilocalorie consumed in a household per adult
equivalent per day with daily minimum requirement of 2100kcal and those getting 2100kcal
and above is food secure and food insecure otherwise.

Independent variables: Household socio-economic characteristics such as household size,

sex of household head, marital status of head, educational status of household head,
dependency ratio, access to credit, ownership of saving account, daily income per adult
equivalent, and proportion of food expenditure are selected variables for the model analysis.

Household size (HSZE): An increase in household size implies more mouth to be fed from
the limited resources and especially in males dominant household the situation becomes more
than this due to high possibility of accustoming to bad habits. As a result in this study, the
household size and status of food insecurity is expected to be related positively.

Sex of household head (HHSX): HHSX is a dummy variable taking a value of 1 if male and
0 otherwise. Household head is a person who economically supports or manages the
household or for some reason of age or respect is considered as head by other members of the
household. It could be male or a female. There is no generally accepted relationship between
sex of household head and level of food security. In the study area where females are actively
engaged in various activity as compared to males, it is hypothesized that households with
female head and food insecurity are related negatively.

Marital status of household head (MRTSTHH): Marriage is biological and social

engagement to support each other both socially and economically. Marriage is established with
a view of helping each other and married people pool their resources and also reduce cost that

38
would have been spent separately. Moreover, married households put aside some of resources
for unforeseen circumstances to smoothen their life. In this study marriage and food insecurity
are hypothesized to be related negatively.

Education level of household head (EDUSTHH): Education level is important for gauging
income earning potential of a household which has significant influence on consumption
behavior of the household. Education is a dummy variable taking a value of 1 if household
head is literate and 0 otherwise. Educational level of household head and food insecurity are
expected to be related negatively.

Dependency ratio (DEPNDRTO): Household members aged below 15 and above 64 are
considered as dependent and dividing it by household members whose age is between 15 - 64
resulted in dependency ratio. These groups are economically inactive and burden to the other
member of the household. It is hypothesized that dependency ratio and food insecurity are
positively related.

Access to credit (HGTCRDT): is a dummy variable taking a value of 1 if household received

credit and 0 otherwise. Credit serves as a means to be involved in income generating activities
and to reap derived benefit based on the amount and purpose of credit. It also normalizes
consumption at hard time. Thus, access and getting credit is negatively related with food
insecurity.

Owning saving account (HSAVACC): is a dummy variable taking a value of 1 if a

household has bank account or maintain credit and saving association and 0 otherwise. Unlike
rural households, urban residents usually maintain bank account and liquidate when faced with
difficult circumstances. Owning saving account or maintaining credit and saving association is
hypothesized to be negatively related to food insecurity.

Daily income per AE (DYINCPAE): One of the major determinants of household food
insecurity is income of a household. Total amount of daily income per adult equivalent in Birr
from different source is computed and the higher the level the lesser the likelihood of

39
household become food insecure. Income is hypothesized to be negatively related with food
insecurity status of the household.

Proportion of household food expenditure (PRPNFDEX): The variable has paramount

importance in reducing the risk of food insecurity. The proportion of income spent on food
expenditure matters the status of household food security. It is hypothesized that proportion of
food expenditure and food insecurity are related negatively.

40
 5. RESULTS AND DISCUSSION

Study results are presented in two categories as a descriptive and econometric model analysis
of the survey data. Descriptive statistics such as mean, standard deviation, percentage and
frequency distribution were employed and binary logistic, econometric model, was used to
identify determinants of food insecurity at household level. Dimensions of household food
insecurity, in terms of extent and severity, were computed by using an FGT index.

5.1. Demographic and Socio-Economic Characteristics

Socio - economic characteristics of sample households by age, sex, household size, education
level and employment status are summarized in relation to the food security status at
household level. Possible explanations on factors supposed to have contribution on household
food insecurity are also presented from analyses of model output.

5.1.1. Age and sex composition

The summary of basic household characteristics for the 200 sample households indicated a
total size of household members of 988 people where females accounted for about 543 (55%).
Percentage of male and female in each category followed similar pattern where age group of
15 – 25 are found to be the largest. Age group of the sample household showed children aged
0 – 14 consisted 23 percent, age group 15 – 64, 74 percent and old age above 64 years of age
amounted to 3 percent. Over all dependency ratio, defined as the ratio of people aged 0 – 14
and above 64 divided by those people aged 15 – 64, was 0.35 (Table 4).

41
Table 4. Characteristics of household by age and sex

Age Group Male (N= 445) Female (N= 543) Total (N = 988)
________________________________________________________________
Percent Percent Percent

0 –7 6.5 9.21 8
8 – 14 18.2 12.52 15.1
15 – 25 34.8 35.73 35.3
26 – 45 27 28.36 27.7
46 – 64 0.6 11.05 10.8
Above 64 2.9 3.13 3
____________________________________________________________________________
Total 100 543 100
Source: Survey result

5.1.2. Characteristics of household by headship

Female as household head comprise 30 percent of sample households while, the majority that
is 70 percent were male headed household. Mean number of household size, age of household
head in years and number of adult equivalent of sample households were found to be 4.94,
45.94, and 4.09 respectively. Mean family size and adult equivalent were found to be higher in
male headed households whereas mean household head age and dependency ratio were higher
in female headed sample households. Fifty three percent of male-headed households had
household size number below seven while 25 percent of female headed household had the
same size of household. Seventeen percent of male headed households had more than six
members whereas only 5 percent of female headed households had family size of more than
six. Male headed households in each group had greater percentage of family size (Table 5).

42
Table 5. Descriptive characteristics of households by headship

Characteristics All Male Female

household headed household headed household
____________________________________________________________________________
Mean household size 4.94 5.10 4.55

Mean age of head 45.94 45.72 46.43

Dependency ratio 0.35 0.34 0.37

Adult equivalent 4.09 4.25 3.72

HH size group
1 – 2 persons 22 (11) 12 (6) 10 (5)

3 – 4 persons 68 (34) 46 (23) 22 (11)

5 - 6 persons 66 (33) 48 (24) 18 (9)

7 – 8 persons 34 (17) 27 (13.5) 7 (3.5)

More than 8 10 (5) 7 (3) 3 (1.5)

____________________________________________________________________________
Total 200 (100) 140 (70) 60 (30)
Figures in bracket are percentage
Source: Survey result

43
5.1.3. Household food security status and family size

It is hypothesized that family size has positive relationship with food insecurity status of a
household. The survey result revealed that 33.3 percent of food secure households have family
size of 1 – 3 persons whereas only 15.1 percent of food insecure households have the same
family size. About 26.8 percent of food insecure and 18.4 percent of food secure households
have family size of more than six persons. Households with larger family size were more
likely to be at risk of becoming food insecure. The survey result indicated that there is a
significant difference in mean family size at less than 5 percent probability level between food
secure and food insecure sample households. The mean household size for food insecure and
food secure households was found to be 5.43 and 4.55 respectively. The minimum and
maximum family size of sample households is 1 and 12 persons (Table 6).

Table 6. Household food security status by family size

Family size Food secure (N = 114) Food insecure (N = 86) Total (N = 200)
group Percent Percent Percent

1-3 33.30 15.1 25.5

4-6 48.2 58.1 52.5
7-9 17.5 22.1 19.5
≥ 10 0.9 4.7 2.5
Total 100 100 100
Mean 4.55 5.43 4.94
SD 1.98 2.24 2.12
Minimum 1
Maximum 12
t-value 2.933**
**significant at less than 5 percent probability level
Source: Survey result

Family size in terms of adult equivalent (AE) and food insecurity are related positively. The
number of adult equivalent within the household does not necessarily imply job opportunity or
access to income and the same was reflected on the survey result. Households having less than

44
3.51 adult equivalents constituted 54.4 percent of food secure and 29 percent of food insecure
households. Similarly 55.8 percent of food insecure and 34.2 percent of food secure
households have AE within a range of 3.51 – 6 (Table 7).

A significant mean difference of adult equivalent was revealed from survey result between the
two groups at probability level of less than one percent. Though, adult equivalent and family
size explain similar household characteristics, difference of significance level was observed
due to family composition interms of age and sex. The mean household adult equivalent for
food insecure and food secure households were 4.51 and 3.77 respectively (Table 7). This
implies that unless households with higher AE are supported by better income levels, the
vulnerability to be food insecurity becomes more serious.

Table 7. Household food security and adult equivalent

AE group Food secure (N = 114) Food insecure (N = 86) Total (N = 200)

Percent Percent Percent
≤3.50 54.4 29 43.5
3.51 - 6.00 34.2 55.8 43.5
6.01 - 8.50 11.4 10.5 11
More than 8.50 0 4.7 2
Total 100 100 100
Mean 3.77 4.51 4.09
SD 1.68 1.89 1.81
Minimum 0.75
Maximum 10.3
t-value 2.832*** 10.3
***significant at less than one percent probability level
Source: Survey result

5.1.4. Food security status by kebeles

Table 8 shows the food security status in Dire Dawa town by kebeles. Kebeles were classified
on the basis of wealth ranking where kebele 12 and 01 are from the most poor and the rest
belongs to the higher wealth rank strata. The location of these kebeles differs where kebele 01

45
and 12 are on periphery of the town and the rest three are located from the middle to center of
the town. Out of 86 total food insecure households kebele 12 and 20 accounted for about 45.3
and 29.1 percent of food insecure households respectively. These two kebeles comprised 61.4
percent of food secure households. There is significant difference at less than five percent
between food secure and food insecure households among kebeles. Residents of kebele 12 are
mainly dominated by ex-soldiers, pensioners and displaced people. Location difference may
contribute to access to different socio-economic services that will contribute to food security
condition of the households.

Table 8. Household food security status by kebeles

2
Kebeles Food secure (N = 114) Food insecure (N = 86) Total (N = 200) χ
Percent Percent Percent
Kebele 03 24.60 11.60 19 10.21**
Kebele 10 3.50 5.80 4.5
Kebele 20 34.20 29.10 32
Kebele 01 10.50 8.10 9.5
Kebele 12 27.20 45.30 35
Total 100 100 100

**significant at less than 5 percent probability level

Source: Survey result

5.1.5. Household food security status and headship

Sex of household head was hypothesized to be one of the variables that make a difference on
the level of food security. Female headed households accounted for about 30 percent of the
sample households. The survey result indicated that 30.2 percent of food insecure households
were female headed whereas, the corresponding figure for male headed households was 69.8
percent. Male headed households comprise 70.2 percent of food secure and the remaining 29.8
percent food secure are female headed households. The survey result showed no significant
difference (p > 0.10) on food security status of household in terms of household head sex.

46
Female headed households had 4.55 mean household sizes and it is 5.10 for male headed
households (Table 9).

Table 9. Household food security by sex of household head

2
Household Food secure (N = 114) Food insecure (N = 86) Total (N= 200) χ
head Percent Percent Percent

Male 70.2 69.8 70 0.004

Female 29.8 30.2 30
Total 100 100 100
Source: Survey result

5.1.6. Household food security and marital status

Marital status of sample household heads indicated that married, divorced, widowed and
single household head accounted for about 67, 8, 20.5 and 4.5 percent respectively. 67.5, 9.6,
18.4 and 4.4 percent of married, divorced, widowed and single households were found to be
food secure whereas, food insecure households consisted of married (66.3%), divorced (5.8),
widowed (23.3), and single (4.6%). The result of the survey showed no significant difference
(p > 0.10) among the marital status with respect to household food security status (Table 10).

Table 10. Household food security by marital status of household head

2
Mariatl Food secure (N = 114) Food insecure (N = 86) Total (N = 200) χ
status Percent Percent Percent
Married 67.5 66.3 67 1.48
Divorced 9.6 5.8 8
Widowed 18.4 23.3 20.5
Single 4.4 4.6 4.5
Total 100 100 100
Source: Survey result

47
5.1.7. Household food security and education of household head

Most household heads in the survey were found to be in secondary cycle education (34.5%)
followed by primary cycle and illiterate with equivalent percentage. There was not a specific
pattern that indicated the higher the level of education of the household head, the more food
secure a household will be. It was revealed that households headed by illiterate persons were
more vulnerable to food insecurity followed by primary and secondary school. In societies
such as Ethiopia where household heads are the major breadwinners of the households,
household head’s educational status could determine food security status of the entire
household. Among other things, household heads play a pivotal role in shaping family
members towards educational attainment thus reducing the probability of being food insecure.

It was hypothesized that household food insecurity and education of household head has
negative relationship. The survey result showed insignificant relationship at 10 percent
probability level when household educational level was categorized in to illiterate, write and
read, primary, secondary etc and became significant while categorized as literate and illiterate
at less than 5 percent. Categorization of household head as literate and illiterate exhibited that
77 percent of household heads were literate. Among the literate households 61.7 percent were
found to be food secure and out of 46 illiterate households 58.7 percent were food insecure
(Table 11).

The possible implication is that in addition to other factors, while some level of education is
important to household food security, its marginal contribution beyond primary education is
small unless continued further or supported by vocational training that enable to be self
employed or competent to be employed. Motives to work in low paid activities especially by
those found in secondary cycle might contribute to the justification of the result.

48
Table 11. Household food security by educational status of household head

2
Level of education Food secure (N= 114) Food insecure (N= 86) Total (N = 200) χ
Percent Percent Percent
Illiterate 16.7 31.4 23 12.506
Write and Read 7.9 6.98 7.5
Primary 1st Cycle 5.3 6.98 6
Primary 2nd Cycle 14 19.77 16.5
Secondary 1st Cycle 8.8 4.65 7
Secondary 2nd Cycle 31.6 22 27.5
Certeficate 1.7 1.16 1.5
Diploma 10.5 6.98 9
B.Sc/B.A & above 3.5 0 2
100 100 100
Literate 61.70 38.3 77 6.005**
Illiterate 41.30 58.7 23

Total 57 43 100
**significant at less than 5 percent probability level
Source: Survey result

5.1.8. Household food security and urban agriculture

Table 12 depicts that only 9 percent of sample households were engaged in urban agriculture.
The result of the survey explained that 93.9 percent of food secure households did not engage
themselves in urban agriculture and only 6.1 percent of food secure were found to be involved
in urban agriculture. On the other hand, those households involved in urban agriculture and
found to be food insecure accounted for about 12.8 percent. Statistically no significance
difference was observed between the two groups on the basis of their participation in urban
agriculture.

49
Table 12. Household food security and urban agriculture

Households involved Food secure (N=114) Food insecure (N=86) Total (N= 200) χ2
in urban agriculture Percent Percent Percent
Yes 6.1 12.8 9 2.647
No 93.9 87.2 91
Total 100 100.0 100
Source: Survey result

5.1.9. Household food security and occupation

Occupation categories of the survey revealed that sample household heads that were in formal,
informal, government, NGOs, other sector and unemployed accounted for about 21.5, 15, 35,
3.5, 2.5 and 22.5 percent respectively. Out of 86 food insecure households those household
heads involved in government, informal activity and unemployed contribute 30.2, 30.2 and
23.3 percent of food insecure households respectively whereas those involved in formal
activity consists of a quarter of food secure households.

Table 13. Household food security status by occupation

2
Food secure (N=114) Food insecure (N= 86) Total (N= 200) χ
Occupation Percent Percent Percent

Formal 27.2 14.0 21.5 27.56***

Informal 8.8 23.3 15
Government 38.6 30.2 35
NGOs 6.1 0.0 3.5
Unemployed 16.7 30.2 22.5
Others 2.6 2.3 2.5
Total 100 100 100
*** significant at less than one percent probability level
Source: Survey result

Those household heads classified as unemployed but are food secure obtain their income from
other household members, gift and remittance and house rent. Not surprisingly, households
whose household heads were employed in NGOs were found to be food secure. The result of

50
the survey revealed a significant difference in terms of occupation of household head among
the food secure and food insecure groups at less than one percent probability level (Table 13).

5.2. Household Income Source and Food Security

In the survey, an attempt was made to collect household income level by source for the last 30
days prior to the survey period to see whether there is a relationship between income level and
food security status. The sampled household heads and members were asked on the amount of
income earned during the last month from all sources.

How do households use their income to meet their material need and what are the sources of
their income? The share of expenditure spent on food commodities and non – food items,
number of income sources, daily income per adult equivalent, contribution of various food
groups to caloric requirement and other related issues in relation to food security status of
households will be discussed under this section.

5.2.1. Household income sources and level

For the purpose of this study income source of sampled households includes monthly salary,
urban agriculture, house rent, business, pensions, daily wage, gift and remittance and other
sources. The monthly income of sampled households revealed the share of income comes
from monthly salary (38.7), business (30.84%), gift and remittance (22.36%), pensions
(2.10%), house rent (2.04%), urban agriculture (0.17%), and other sources (1.24%). The mean
income level of sampled households was found to be Birr 1654.98 with standard deviation of
2680.19. The minimum and maximum monthly income was Birr 20 and 26250 respectively
(Table 14).

51
Table 14. Household income source and level

Source of income Total monthly income (Birr) Percent

Monthly salary 128,116 38.7

Urban agriculture 564 0.17
House rent 6735 2.04
Business 102090 30.84
Pensions 6935 2.1
Daily wage 8410 2.54
Gift and remittance 74,026 22.36
Others 4120 1.24
Total 330,996 100
Mean 1654.98
SD 2,680.19
Source: Survey result

Twenty three percent of sampled households had gained mean monthly income Birr 232.48
which accounts 3.3% of total income. On the other hand 28.5 and 3.5% of sampled households
obtained an average monthly income of Birr 586.78 & 11750 respectively (Table 15).

Table 15. Mean monthly income of household by income group

Income group Percent (N = 200) Mean income Income share

0 - 399 23.5 232.48 3.3
400 - 799 28.5 586.78 10.1
800 - 1199 10 964.3 5.8
1200 - 1599 11 1387.13 9.2
1600 - 1999 5 1765.8 5.33
2000 - 2999 9 2364.5 12.86
3000 - 3999 2.5 3362 5.07
4000 - 4999 1.5 4170 3.77
5000 - 5999 3 5363.33 9.72
6000 - 6999 2.5 6570 9.92
7000+ 3.5 11750 24.84
100 1654.98 100
Source: Survey result

52
The share of income and percentage of cases in different income group revealed large
variation among sampled households.

5.2.2. Household food security status by number of income source

The sampled households reported that 64% of them had engaged in one to two income
generating activities. Diversifying income source are important to reduce risk in urban
economic environment especially for low income groups. The average number of income
generating activities or sources per household for the whole sample was found to be 2.3. The
corresponding figure for food secure and food insecure was found to be 2.16 and 2.5
respectively. The reasons for higher mean of income sources for food insecure households
might be associated with type of activity households had been engaged and insufficiency of
income to cover households food and non food expenditure.

Table 16. Household food security by number of income sources

Number of income Food secure (N = 114) Food insecure (N = 86) Total (N = 200)
sources Percent Percent Percent
1-2 71 54.7 64
3-4 24.6 37.2 30
5-6 4.4 8.1 6
Total 100 100 100
Mean 2.16 2.50 2.30
SD 1.07 1.22 1.15
t - value 2.096*
*significant al less than 10 percent probability level
Source: Survey result

It was hypothesized that number of income sources and food insecurity were related
negatively. Households with income source of 1 – 2, 3 – 4, and 5 – 6, accounted for about
54.7, 37.2 and 8.1 percent of the food insecure groups where as 71 percent of food secure
households earn their income from one to two sources. Within food insecure group, the higher
the number of income sources, the lower the percentage of food insecure households. The

53
number of income source exhibited a significant mean difference at less than 10 percent
probability level between the two groups (Table 16).

5.2.3. Household food security and daily income per adult equivalent

Household income has a paramount importance in achieving household food security

especially in urban situation where people largely depend on earning capacity rather than on
natural resources like rural areas.

Table 17. Household food security status by daily income per adult equivalent

Daily income per Food secure (N = 114) Food insecure (N = 86) Total (N = 200)
adult equivalent Percent Percent Percent
≤ 4.00 15.79 38.37 25.5
4.01 - 8 18.42 29.07 23
8.01 - 12 21.93 19.77 21
12.01 - 16 11.40 5.81 9
16.01 - 20 7.89 4.65 6.5
≥ 20 24.56 2.33 15
Total 100 100 100
Mean 16.73 6.93 12.51
SD 18.80 6.30 15.54
t - value -5.188***
*** significant at less than one percent probability level
Source: Survey result

Daily income per adult equivalent was hypothesized to have negative relationship with
household food insecurity. Households with daily income per AE of less than or equal to Birr
4, 4.01 – 8, 8.01 – 12, 12.01 – 16, 16.01 – 20 and above 20 comprised 25.5, 23, 21, 9, 6.5 and
15 percent respectively. 15.8 percent of food secure and 38.4 percent of food insecure
households were found to earn a daily income per adult equivalent of Birr 4 or less. Hence, as
daily income per AE increases, the percentage of food insecure households exhibits a
declining tendency. The mean daily income per adult equivalent of food secure and food

54
insecure household were found to be Birr 16.73 and 6.93 respectively. The survey result
depicted a significant mean difference in daily income per adult equivalent at probability level
of less than one percent between food secure and food insecure household group (Table 17).

5.3. Household Food Security and Expenditure

Households usually allocate their income to meet food and non – food needs of their family.
The sample households were asked on the quantity and value of food consumed for the last
seven days and then monthly food expenditure was computed.

Table 18. Monthly household food and non - food expenditure (Birr)

Expenditure type Total monthly Percent

expenditure
Food expenditure
Ceraels 18576.56 19.08
Vegetables 12712.75 13.06
Pulse 2336.4 2.4
Fruit 3440.35 3.53
Prepared food 17662.71 18.14
Livestock products 24542.35 25.21
Otther food 18069.08 18.56
Total 97340.2 100
Mean 486.7
Non - food expenditure
Housing 10641.1 6
Water, energy and regular 38183.95 21.53
Clothing 27028.5 15.24
Health & education 16139.6 9.08
Transportation 15447.4 8.71
Communication 13496.5 7.61
Annual and ocassional 46484 26.21
Others 9931.24 5.6
Total non- food expenditure 177352.29 100
Mean 886.76
Source: Survey result

55
The average monthly household expenditure on food and non food expenditure for sampled
households were found to be Birr 486.68 and 886.76 respectively. The minimum and
maximum monthly food expenditure was Birr 22.5 and 2712.42 respectively whereas for non
– food expenditure the minimum was Birr 9.85 and the maximum was Birr 4486 (Table 18).

The mean food budget share of sample household indicted that cereals, vegetables, pulse, fruit,
prepared food, livestock and other accounts for about 19, 13, 2.4, 3.53, 18.14, 25 and 18.56
percent of food expenditure respectively. Expenditure on prepared food within food insecure
household constitute the largest share that may signify the cost minimization ways for this
group that would have been spent on food preparation process. Non - food expenditure of
sample household consist outlays on housing(6%), water, energy and regular expenditure
(21.53%), clothing (15.2%), medical care and education (9.1%), communication and
transportation (16.3%), annual and occasional expenditure (26.2%) and others (5.6%) as
indicated on Table 18.

The different income group versus food and non - food expenditure were compared to monthly
income. The result of the data exhibited that the reported income was underestimated
especially for low income groups while overall expenditure was overestimated beyond
monthly income. Nevertheless, there is an indication that as level of income increased,
proportion of income spent on food decline that is consistent with Engle’s Law. Proportion of
expenditure on food commodity is higher as compared to non - food expenditure for
households with monthly income of less than Birr 800 and the reverse is true for income group
of more than Birr 800. Moreover, those sample households who had earned monthly income
of below Birr 600 have spent larger proportion of their income on food i.e. more than 60
percent of their income (Table 19). This group comprises more than one third of sample
households.

56
Table 19. Monthly household income and expenditure proportion

Income Total Total monthly Mean Per household monthly Proportion of household
group monthly expenditure on monthly expenditure on expenditure on
income Food Non - food income Food Non - food Total Food Non- food Total
0 - 99 305 983.36 640.70 50.83 163.89 106.78 270.68 322.41 210.07 532.48
100 - 199 1584 1276.29 461.25 158.40 127.63 46.12 173.75 80.57 29.12 109.69
200 - 299 3271 2668.50 1228.09 233.64 190.61 87.72 278.33 81.58 37.54 119.13
300 - 399 5767 3970.03 1467.39 339.24 233.53 86.32 319.85 68.84 25.44 94.29
400 - 499 7565 4931.36 4769.82 445.00 290.08 280.58 570.66 65.19 63.05 128.24
500 - 599 5839 3558.43 3246.66 530.82 323.49 295.15 618.64 60.94 55.60 116.55
600 - 699 10389 5345.61 3894.85 649.31 334.10 243.43 577.53 51.45 37.49 88.94
700 - 799 9654 4261.33 3952.95 742.62 327.79 304.07 631.87 44.14 40.95 85.09
800 - 899 5833 2435.91 3163.50 833.29 347.99 451.93 799.92 41.76 54.23 96.00
900 - 999 4573 1565.36 2186.64 914.60 313.07 437.33 750.40 34.23 47.82 82.05
1000 -2999 99616 35017.16 62156.94 1717.52 603.74 1071.67 1675.42 35.15 62.40 97.55
3000+ 176600 31326.86 90183.50 6792.31 1204.88 3468.60 4673.48 17.74 51.07 68.81
Total 330996 97340.19 177352.29 1654.98 486.70 886.76 1373.46 29.41 53.58 82.99
Source: Survey result

57
5.3.1. Household food security and daily food expenditure per adult equivalent

Urban food expenditure in Dire Dawa sample households were characterized by heavy
reliance on purchased food commodities such as cereals, pulses, vegetables, and livestock
products. Food commodity from own production was almost non - existent.

The mean daily food expenditure per AE of the whole sample was Birr 4.04 and for food
secure and food insecure household the figure was 5.12 and 2.60 respectively. Households
with daily food expenditure per AE of less than Birr 4.51 comprised 88.4 percent of food
insecure and 47.4 percent of food secure. On the other hand, among those food insecure who
had spent more than eight Birr constituted only 1.16 percent while the corresponding figure
was 12.28 percent for food secure household. As the amount of daily food expenditure per AE
increases significant level of difference was observed between the two groups. The result of
the survey suggested a significant mean difference in daily food expenditure per AE at less
than one percent significant level (p < 0.01) between the two groups (Table 20).

Table 20. Household food security status by daily food expendituire

Daily food expenditure Food secure (N = 114) Food insecure (N= 86) Total (N = 200)
per AE Percent Percent Percent
≤ 4.50 47.37 88.37 65
4.51 - 8.00 40.35 10.47 27.5
8.01 - 11.50 8.77 1.16 5.5
11.51 - 15.00 2.63 0.00 1.5
≥ 15.01 0.88 0.00 0.5
Total 100 100 100
Mean 5.12 2.60 4.04
SD 2.61 1.53 2.54
t - value -8.538***
*** significant at less than one percent probability level
Source: Survey result

58
5.3.2. Household food consumption

This study measured household food security in terms of the adequacy of daily kcal
consumption per adult equivalent. Sampled households reported that over 22 food items were
used for consumption during the last 7 days prior to the survey period and for the purpose of
this paper they are categorized in to seven food types just to indicate the contribution of each
group to total kcal consumption. The most commonly eaten foods were teff, sorghum, wheat,
livestock products, vegetables, sugar and others.

In general the sampled households derive their calorie intake from cereals (44.11%),
vegetables (7.07%), prepared food (15.87%), pulses (5.6%), fruit (1.1%), livestock product
(4.86%) and others which includes sugar, salt, oil and others (21.38%).

5.4. Determinants of Food Insecurity

An econometric model, logistic regression, was employed to identify the determinants of

household food insecurity. The variables included in the model were tested for the existence of
multicollinearity, if any. Contingency coefficient and variance inflation factor were used for
multicollinearity test of dummy and continuous variables respectively.

Table 21. Contingency coefficient value for dummy variables

Variables HHSX MRTSTHH EDUSTHH HGTCRDT HSAVACC

HHSX 1
MRTSTHH 0.463 1
EDUSTHH 0.367 0.330 1
HGETCRDT 0.151 0.179 0.167 1
HSAVACC 0.197 0.167 0.273 0.100 1
Source: Own computation

59
Contingency coefficient value ranges between 0 and 1, and as a rule of thumb variable with
contingency coefficient below 0.75 shows weak association and value above it indicates strong
association of variables. The contingency coefficient for the dummy variables included in the
model was less than 0.75 that didn’t suggest multicollinearity to be a serious concern as
depicted on Table 21.

As a rule of thumb continuous variable having variance inflation factor of less than 10 are
believed to have no multicollinearity and those with VIF of above 10 are subjected to the
problem and should be excluded from the model. The computational results of the variance
inflation factor on Table 22 confirmed the non-existence of association between the variables
and were included in the model.

Table 22. Variance inflation factor of continuous variables

____________________________________________________________________
Variable R2 VIF
____________________________________________________________________
HSZE 0.003 1.00
DAYINCPAE 0.116 1.13
DEPNDRTO 0.055 1.05
PRPNFDEXPH 0.076 1.08
___________________________________________________________________
Source: Own computation

In total, nine independent variables were used for estimation. To identify determinants of food
insecurity among hypothesized explanatory variables that are supposed to have influence on
Dire Dawa urban households, binary logit model was estimated using a statistical package
known as SPSS version 11. Types, codes and definition of the variables and estimates of the
logit model are presented on Table 23 and Table 24 respectively.

60
Table 23. Types, codes and definition of variables in the model
_________________________________________________________________________
Types Codes Definition
_________________________________________________________________________
Dummy HHSX 1, if household head is male;
0 otherwise
Dummy MRTSTHH 1, if household head is married;
0 otherwise
Dummy EDUSTHH 1, if household head is literate;
0 otherwise
Dummy HGTCRDT 1, if household got credit;
0 otherwise
Dummy HSAVACC 1, if household has saving
account; 0 otherwise
Continuous HSZE Household size in number
Continuous DENPDRTO Dependency ratio
Continuous DYINCPAE Daily income per adult equivalent
Continuous PRPNFDEX Proportion of food expenditure
__________________________________________________________________________

61
Table 24. The maximum likelihood estimates of the logit model
________________________________________________________________________

Variables Coefficient Wald - statistics Odds ratio

________________________________________________________________________

HSZE 0.413 15.528*** 1.512

HHSX 1.797 4.572** 6.033
MRTSTHH -1.472 3.327* 0.229
EDUSTHH -1.161 5.992** 0.313
DYINCPAE -0.147 18.625*** 0.863
PRPNFDEX -0.008 3.276* 0.992
HGTCRDT -0.862 3.248* 0.422
HSAVACC -0.296 0.527 0.744
DEPNDRTO -0.156 0.262 0.855
Constant 0.563
Pearson Chi-square 66.673***
-2 Log likelihood 206.653
Sensitivity 69.8
Specificity 78.9
Percent correctly predicted (Count R2) 75
Sample size 200
*** significant at less than 1% probability level
** significant at less than 5% probability level
* significant at less than 10% probability level
Source: Model output

The likelihood ratio has a chi – square distribution and it is used for assessing the significance
of logistic regression. Model chi – square provides the usual significance test for a logistic
model i.e. it tests the null hypothesis that none of the independents are linearly related to the
log odds of the dependent. It is an overall model test which doesn’t assure every independent

62
is significant. The result is significant at less than one percent probability level revealing that
the null hypothesis that none of the independents are linearly related to the log odds of the
dependent is rejected. Additionally, goodness of fit in logistic regression analysis is measured
by count R2 which works on the principle that if the predicted probability of the event is
greater than 0.50 the event will occur otherwise the event will not occur. The model result
show the correctly predicted percent of sample household is 75 percent which is greater than
0.50. The sensitivity, correctly predicted food insecure is 69.8 percent and that of specificity,
correctly predicted food secure is 78.9 percent. This indicates that the model has estimated the
food insecure and food secure correctly.

5.4.1. Explanation of significant independent variables

Nine independent variables that are hypothesized to have influence on household food
insecurity in the study area were included in the model, of which seven were found to be
statistically significant even though the level of statistical significance for the independent
variables included in the model was different for individual variable and the sign of the
significant parameters were as expected. The model output revealed that household size
(HSZE) and daily income per adult equivalent (DYINCPAE) were significant at less than one
percent probability level. Educational status (EDUSTHH) and sex of household head (HHSX)
were found to be significant at less than 5 percent probability level and the rest three variables
namely, marital status of household head (MRTSTHH), household access to credit
(HGTCRDT) and proportion of food expenditure (PRPNFDEX) were significant at less than
ten percent probability level. The remaining two variables, namely overall dependency ratio
(DEPNDRTO) and owning of saving account (HSAVACC) were not statistically significant.

In light of the above summarized model results possible explanation for each significant
independent variable are given consecutively as follows:

Household size (HSZE): Given the strong positive relationship between household size and
food insecurity already noted in the descriptive part, it is not surprising that the estimated
parameters are positive and highly significant. This positive relationship shows that the odds

63
ratio in favor of the probability of being food insecure increase with increase in household
size. Other things remaining equal, the odds ratio in favor of food insecurity increases by a
factor of 1.512 as household size increases by one. The possible reason is that with existing
high rate of unemployment and less employment opportunity coupled with low rate of
payment, an additional household member shares the limited resources that lead the household
to become food insecure.

Sex of household head (HHSX): Sex of household head is significant at less than 5 percent
probability level and positively related with household food insecurity. The result is inline
with apriori expectations. Other things being equal, the odds ratio in favor of food insecurity
increases by a factor of 6.033 as the household head becomes male. Possible reason is that
female household head is more responsible and give due attention to their family and having a
woman as head of household impacts higher caloric availability reflecting differences in
spending priority between male and female headed households.

Marital status of household head (MRTSTHH): The result of the model depicted that marital
status of household head and food insecurity are related negatively in the study area. The
negative relation indicated that the odds ratio in favor of food insecurity, ceterius paribus,
decreases by a factor of 0.229 as the household head becomes married.

The possible explanation is related to the economic scale of consumption item purchased and
pooling available resources in one way or another and possibly, married households reduce
expenditure that would have been spent separately. The results of marital status and sex of
household seems to be contradictory but headship is not only gifted to male as observed from
sampled households. There were female household heads in the presence of male (husband)
either due to economic reasons or absence of male household head in the area for any reasons.

In general, being married by itself is not an assurance to escape the risk of food insecurity.
Rather it is mainly because of the fact that household size, level of income and other factors of
household affect food security status in relation to marital status. The result of the survey
confirms prior hypothesis of the study.

64
Educational status of household head (EDUSTHH): Although, educational status of other
income earner household members have great importance, that of head plays a significant role
in shaping household members by being exemplary and willing to invest on education.
Holding other variables constant, negative relation of educational status of household head and
the dependent variable brought the odds ratio in favor of food insecurity to reduce by a factor
of 0.313 as head of the household becomes literate. It is explained interms of contribution of
education on working efficiency, competency, diversify income, adopting technologies and
becoming visionary in creating conducive environment to educate dependants with long term
target to ensure better living condition than illiterate ones. Thus, being literate reduces the
chance of becoming food insecure in the sample households.

Daily income per adult equivalent (DYINCPAE): The survey result showed a negative
relation between daily income per adult equivalent and food insecurity and the coefficient is
highly significant at less than one percent probability level. The odds ratio in favor of food
insecurity, holding other variables constant, decreases by a factor of 0.863 as daily income per
adult equivalent increases by one Birr. The result corresponds with the prior expectation and
the possible explanation is that income determines purchasing power of the household with the
prevailing price so that those households having higher daily income per adult equivalent are
less likely to become food insecure than low income households.

Proportion of food expenditure (PRPNFDEX): Proportion of food expenditure spent by the

household is significant at less than 10 percent probability level and related negatively with
food insecurity. Under ceterius paribus condition, the odds ratio in favor of food insecurity
decreases by a factor of 0.992 as proportion of food expenditure increases by one. As
proportion of expenditure on food increases, access to food by household also increases to the
amount needed for household consumption. In situation where some covariant shocks for
instance rise in price of food commodity happens increasing proportion on food expenditure
helps to overcome the change and keep households in accessing needed food and it also leads
to the consumption of better quality food.

65
Household access to credit (HGTCRDT): The results of the survey revealed that the variable
under consideration is negatively related and significant at less than 10 percent probability
level with food insecurity. Holding other things constant, the odds ratio in favor of food
insecurity decreases by a factor of 0.422 as a household has access to credit. The possible
explanation is that credit gives the household an opportunity to be involved in income
generating activities so that derived revenue increases financial capacity and purchasing power
of the household to escape from risk of food insecurity. Access to credit also smoothen
consumption when household faces with hard time.

5.5. Extent and Incidence of Food Insecurity

The quantitative measures of poverty index developed by FGT and recently employed by
Hoddinot (2001) and others in food security study are head count ratio, food insecurity gap
and severity of food insecurity. The three measures vary with the weight attached to severity
of food insecurity. Based on food energy intake at household level, head count ratio or
incidence of food insecurity indicates the percentage of households who fall below the
predetermined kcal amount. Though head count ratio is simple to compute and interpret, it is
insensitive to difference in depth of food insecurity.

The second index, food insecurity gap, measures the aggregate food insecurity deficit of the
food insecure population relative to the recommended caloric requirement i.e. it reflects total
kcal deficit of all household below the subsistence energy requirement level. It can be also
interpreted as a potential indicator of eliminating food insecurity by transferring required
resources to food insecure. The draw back of this measure is that it doesn’t capture the
difference in severity of food insecurity among the poor.

The last index that measures the mean of squared proportional shortfalls from the cut off
points is known as severity of food insecurity. The problem with this measure is that it is not
easy to interpret.

66
5.5.1. Extent and severity of food insecurity

The three FGT indices used are head count index, food insecurity gap and severity of food
insecurity. The results of the survey revealed that the head count ratio or incidence of food
insecurity are 0.43 which implies 43 percent of the sampled households can not meet the daily
recommended caloric requirement.

To know how far the food insecure households are below the recommended daily caloric
requirement, food insecurity gap was calculated. Food insecurity gap provides the possibility
to estimate resources required to eliminate food insecurity through proper targeting. The
calculated value for food insecurity gap was found to be 0.13. This indicates if the city
administration mobilizes and distributes resources that can meet 13 percent of caloric need of
every food insecure households and distribute to each household to bring up to the
recommended daily caloric requirement level, then theoretically food insecurity can be
eliminated. It means that assuming sampled households are representative there are about
41344 households with 169,097 adult equivalents. Considering the daily recommended 2100
kcal per adult equivalent, a resource needed to push all households to daily subsistence
requirement is estimated to be 46,163,481 kcal per day. Taking a Kg of cereal produce 3700
kcal, total amount of cereals needed per day becomes 124.76 quintals. This shows a
requirement of 45,337.4 quintals of cereal or equivalent amount of money to purchase
45,337.40 quintals of grain to bring all households to obtain daily subsistence caloric energy in
a year.

Finally, to approach the most food insecure sample households, severity of food insecurity was
calculated by assigning a higher weight, α = 2. Thus, the survey result indicated that the
severity of food insecurity becomes 0.059.

67
5.5.2. Incidence of food insecurity and household characteristics

The incidences of the food insecurity with some household characteristics are depicted on
Table 25. Food insecurity is more than three times less prevalent with households of less than
or equal to three members as compared to those households with more than nine members. On
the other hand, households with family size ranging from 7 to 9 have almost twice more
incidence of food insecurity as compared to those having less or equal to three family
members.

The prevalence of food insecurity decreases as household head education status improved i.e.
literate household head has 38.3 percent prevalence of food insecurity and is higher for
illiterates. Households with higher daily income per adult equivalent have much lower
incidence of food insecurity than lower daily income per adult equivalent households. As
indicated in Table 25 the incidence of food insecurity is four times lower for households who
earned Birr 16.01 and above of daily income per adult equivalent than those with less than Birr
4.01.

The negative relations of food insecurity and access to credit revealed higher incidence for
households who didn’t get credit and had no saving account. The prevalence of food insecurity
of households who had no saving account and didn’t have access to credit was 55.2 and 44.1
percent respectively.

68
Table 25. Incidence of food insecurity
_______________________________________________________________________
Characteristics Household Number of food Total Food insecurity
grouping insecure household incidence
_______________________________________________________________________
Family size 1–3 13 51 25
4–6 50 105 48
7–9 19 39 49
≥ 10 4 5 80
Overall 86 200 43
________________________________________________________________________
Education Illiterate 27 46 58.7
Literate 59 154 38.3
Overall 86 200 43
________________________________________________________________________
Owning saving No 53 96 55.2
account Yes 33 104 31.7
Overall 86 200 43
________________________________________________________________________
Daily income ≤4 33 51 64.7
per AE 4.01 - 8.00 25 46 54.3
8.01 - 12.00 17 42 40.5
12.01 - 16.00 5 18 27.8
≥ 16.01 6 43 13.9
Overall 86 200 43
________________________________________________________________________
Borrowed credit No 75 170 44.1
Yes 11 30 36.7
Overall 86 200 43
________________________________________________________________________
Source: Survey result

69
 6. SUMMARY AND POLICY RECOMMENDATIONS

6.1. Summary

The study was conducted with the specific objective of examining food insecurity situation,
estimating the food insecurity gap and severity and identifying the determinants of food
insecurity at household level in Dire Dawa town. The research objectives were realized
through conducting household survey in five administrative kebeles of the study area.
Household demographics, educational status, income, expenditure and other data deemed to be
relevant were collected, organized, analysed and interpreted to come with possible results.

The analysis employed both descriptive statistics and econometric methods. Descriptive
statistics were employed to describe household characteristics with food security status.
Binary logistic model was specified and estimated to identify determinants of food insecurity.
FGT indices were used for the computation of incidence and severity of food insecurity among
sample households.

The sample households were classified into food secure and food insecure groups based on
kcal actually consumed by the households during the previous seven days of survey date either
through purchase or other means. Total amount of food commodity consumed by each
household during the seven days were converted in to equivalent daily kcal per AE and then
compared with recommended daily kcal per adult equivalent. Total daily food energy per adult
equivalent of less than 2100 kcal was considered as food insecure and food secure otherwise.

The descriptive statistics showed the existence of a significant mean difference in daily
income per AE and daily food expenditure per AE at less than one percent probability level
between food secure and insecure households. The T - test for household size and adult
equivalent showed a mean difference between the two groups at less than 5 and 1 percent
probability level respectively whereas at less than 10 percent significance level for number of

70
income sources. Education status and occupation of household head were found to be
significant at 5 and 1 percent significance levels respectively.

Binary logit econometric model was employed to estimate determinants of the probability of
being food insecure as a function of various household characteristics among sampled
households of Dire Dawa town. Seven out of nine variables namely household size, sex of
household head, marital status of household head, education of household head, daily income
per adult equivalent, proportion of food expenditure and access to credit were found to be
statistically significant with the hypothesized sign as determinants of household food
insecurity in the study area. Household size and daily income per AE were significant at less
than one percent probability level and are related with the dependent variable positively for the
first and inversely for the latter.

Household head education and sex of household head were significant at less than 5 percent
and the remaining three namely, proportion of food expenditure, access to credit and marital
status of household head were found to be inversely related at less than 10 percent significance
level.

The head count ratio revealed that 43 percent of sampled households are found to be food
insecure. The gap and severity of food insecurity were estimated to be 13 and 5.9 percent
respectively. Considering the daily recommended 2100 kcal per adult equivalent, a resource
needed to bring all households to daily subsistence requirement amounted to 46,163,481 kcal.
This shows daily requirements estimate of 124.76 quintals of cereal per day which is
equivalent to 45,337.40 quintals per year.

This study has attempted to come up with the result of the analysis with defined scope
however a lot remained to be unanswered. To provide basic information on the patterns and
determinants of urban food insecurity, the social, political and environmental dimensions, role
of urban agriculture in urban food insecurity, descriptive data on purchasing patterns of food
insecure, specific characteristics that make urban poor more vulnerable to food insecurity and
their coping mechanisms demands future researchers attention.

71
6.2. Policy Recommendations

Possible policy recommendations that emanate from the results of the research study are
presented as follows:

1. Household size was found to be directly related with household food insecurity. The slow
downing of the Dire Dawa business condition coupled with poor investment performance
has contributed to the deterioration of income generation capacity of food insecure
households. With these scenario, having more household size aggravate the problem of
meeting food leave alone education, health and other non – food demands of household
that will bring future return. So, action based awareness creation on the impacts of
population growth at the family, community and national level should be strongly
advocated that lead to reduction in fertility and lengthen birth spacing resulted in smaller
household size. Moreover, development actors involved on population issue should
encourage households having acceptable number of children through provision of especial
offer such as covering schooling cost, giving training and other related incentives.

2 As income and food insecurity are negatively related on the model results, searching and
providing productive technical skill that make trainees competitive on the current market
and generate income should be sought and promoted. Additionally, budget allocated for
food security programme are in use only for rural areas even if 74 percent of regional
people live in urban Dire Dawa. So this budget should also be allocated and utilized for
employment generation scheme in urban Dire Dawa.

3. Access to credit can create an opportunity to be involved in economic activity that

generates revenue to households. Recently established small and micro business agency in
the region has started activity of organizing and training of every business community who
are interested. Provision of start up capital in the form of loan is effected through Dire
micro finance institution. Development partners operating in the study area should
implement provision of credit to eligible households using targeting criterion that reflects
actual characteristics of food insecure households. The other pressing issue related to

72
 provision of credit is the requirement of collateral and group lending procedure, which
discourages so many households. People are afraid of holding accountability for others so
individual lending should be considered as another option and collateral requirement
should be avoided if there is a need to lift food insecure households from their current
situation. Borrowers should be encouraged to save or contribute as matching fund to reach
the limited resources over large number of needy people.

4. The effect of education on household food security confirms the significant role of the
variable in consideration for betterment of living condition. The more household head
educated, the higher will be the probability of educating family member and familiar with
modern technology, which the twenty first century so badly demands. So, strengthening
both formal and informal education and vocational or skill training should be promoted to
reduce food insecurity in Dire Dawa town.

73
 7. BIBILIOGRAPHY

Abbi Mamo and Mckay, A. 2003. Chronic Poverty in Urban Ethiopia: Panel Data Evidence.
Paper Prepared for International Conference on Staying Poor: Chronic Poverty and
Development Policy. University of Manchester, UK.

Abebaw Shimeles. 2003. Dimensions and Determinants of Food Insecurity Among Rural
Households in Dire Dawa. An MSc Thesis Presented to the Schools of Graduate Studies of
Alemaya University. Alemaya.

Agency for Cooperation and Research in Development. 1994. Dire Dawa Urban Poverty and
Coping Strategy. Urban Development Program (Memo).

Amemiya, T. 1981. Quantitative Response Model: A Survey. Journal of Economic Literature.

19: 1483 – 1536.

Ayalneh Bogale. 2002. Land Degradation, Impoverishment and Livelihood Strategies of Rural
Households in Ethiopia: Farmers’ Perceptions and Policy Implication. Vol 8. Shaker Verlag,
Germany.

Bonnard, P. 2000. Assessing Urban Food Security: Adjusting the FEWS Rural Vulnerability
Assessment Framework to Urban Environment. Washington, D.C.

Central Statistical Authority. 1994. Population and Housing Census for Ethiopia: Result for
Dire Dawa Provisional Administration. Addis Ababa.

______.1999. Population and Housing Census for Ethiopia. Result for Dire Dawa
Administration. Volume II. Analytical Report. Addis Ababa.

______. 2004. Report on Urban Bi - Annual Employment Unemployment Survey. First Year
Round 2. Statistical Bulletin 319. Addis Ababa.

74
Debebe Habtewold. 1995. Food Security: A Brief Review of Concepts and Indicators. Food
Security, Nutrition and Poverty Alleviation in Ethiopia: Problems and Prospects. Proceedings
of the Inaugural and First Annual Conference of the Agricultural Economics Society of
Ethiopia, Addis Ababa, pp. 1 - 18.

Edilegnaw Wale. 1997. Household Income Inequality and Poverty During the Structural
Adjustment: The Case of Ghana. An MSc Thesis Presented to the University of Wageningen.
The Netherlands.

EHNRI. 1998. Food Composition Table for Use in Ethiopia. Part III and IV. Addis Ababa.

Engle, P., Ruel, M. T., Garrett, J. L., Morris, S. S., Maxwell, D., Oshaug, A., Menon, P.,
Slack, A., and Haddad, L. 1998. Urban Challenges to Food and Nutrition Security: A Review
of Food Security, Health and Caregiving in Cities. Food Consumption and Nutrition Division
Discussion Paper No 51. IFPRI, Washington, D.C.

Ethiopian Economic Association. 2003. Reports on the Ethiopian Economy: Industrialization

and Industrial Policy in Ethiopia. Vol.III. Addis Ababa.

Federal Democratic Republic of Ethiopia. 2001. Food Security Strategy. Update, Addis
Ababa.

Federal Democratic Republic of Ethiopia. 2002. Sustainable Development and Poverty

Reduction Program. Addis Ababa.

Frankenberger, T. 1992. Indicators and Data Collection Methods for Assessing Household
Food Security. Household Food Security: Concepts, Indicators, and Measurements. A
Technical Review. Rome, IFAD/UNICEF, pp. 73 - 134.

Garrett, J. L., and Ruel, M. T. 1999. Are Determinants of Rural and Urban Food Security and
Nutritional Status Different? Some Insight from Mozambique. World Development 27(11):
1959 – 1875.

Gujirati, D. N. 1995. Basic Econometrics. 3rd ed. McGraw - Hill Co. New York.

___________.1999. Essentials of Econometrics. 2nd ed. McGraw – Hill Co. New York.

75
Hoddinott, J. (ed). 2001. Methods for Rural Development Projects. Food Security in Practice.
IFPRI, Washington, D.C.

Hosmer, D., and S. Lemeshew. 1989. Applied Logistic Regression. A Wiley – Inter Science
Publication, New York.

Haddad, L., Ruel, M.T., and Garrett, J. L. 1999. Are Urban Poverty and Undernutrition
Growing? Some Newly Assembled Evidence. World Development 27(11): 1891 – 1904.

ICRA. Learning Resources: Livelihood – Key Concepts. Accessed on: 09/11/2005.

http://www.icredu.org/objects/anglolearn/livelihoods- key_concepts..

IFPRI. 2002. Living in the City. Challenges and Options for the Urban Poor. IFPRI Issue Brief
No.9. Washington, D.C.

Kifle Lemma and Yosef Gebrehiwot. 1999. The Food Security Situation status in Ethiopia:
Concept, Status and Trends. In:Taye Asefa (ed.). Food Security in Sustainable Land Use:
Proceedings of the First National Workshop of NOVIN Partners Forum on Sustainable Land
Use. NOVIN. Addis Ababa.

Kementa, P. 1986. Elements of Econometrics.2nd ed. Macmillan Pub.Co. New York. U.S.A

Maddala, G. S. 1983. Limited Dependent and Qualitative Variables in Econometrics.

Cambridge University Press. New York.

Maxwell, D. Urban Food Security in Sub-Saharan Africa. Accessed on: July 20 – 31, 2005.
http://web.idrc.ca/es/ev-30582-201-1-DO_TOPIC.html

Maxwell, D. 1998. The Political Economy of Urban Food Security in Sub-Saharan Africa.
World Development 27(11):1939–1953.

Maxwell, D., Levin, C., Armar-Klemesu, M., Ruel, M., Morris, S., and Ahiadeke, C. 2000.
Urban Livelihood and Food and Nutrition Security in Greater Accra, Ghana. Research Report
No.112. IFPRI, Washington, D.C.

76
Maxwell, S., and Smith, M. 1992. Household Food Security: A Conceptual Review.
Household Food Security: Concepts, Indicators, Measurements. A Technical Review.
IFAD/UNICEF, pp. 1 - 72.

Ministry of Planning and Economic Development. 1994. Report on Household Food Security
Study of Four Towns, Food and Nutrition Unit, Addis Ababa.

Ministry of Economic Development and Cooperation. 1997. Highlights of Development

Strategies for Selected Sectors, Macro Economic Planning and Policy Analysis Department,
Addis Ababa.

Ministry of Finance and Economic Devevlopment. 2002. Development and Poverty Profile of
Ethiopia, Welfare Monitoring Unit, Addis Ababa.

Mucavele, F. 2001. A Vulnerability and Food Security Study of Urban Maputo. Mozambique.

Mutangadura, G., and Makaudze, E. 1999. Urban Vulnerability to Chronic Poverty and
Income Shocks and Effectiveness of Current Social Protection Mechanisms: The Case of
Zimbabwe. Harare

Pindyek, R., Rubinfeld, S., and Daniel, L. 1981. Econometric Models and Economic forecasts.
2nd ed. McGraw Hill Book Company, New York.

Smith, M., Maxwell, S., and Pointing, J. 1992. Household Food Security: Concepts and
Definition – An Annotated Bibliography. Household Food Security: Concepts, Indicator
and Measurements. A Technical Review. Rome, IFAD/UNICEF, pp. 135 -192.

Ssewanyana, S.N. 2003. Food Security and Child Nutrition Status Among Urban Poor
Households in Uganda: Implication for Poverty Alleviation. African Economic Research
Consortium. Research Paper 130. Nairobi.

Strock, H., Bezabih Emana, Berhanu Adnew, Borowiecki, A. and Shimeles Woldehawariat.
1991. Farming System and Farm Management Practice of the Smallholders in Hararghe
Highlands: A Base Line Survey. Farming System and Resource Economics in the Tropics, Vol
11. Kiel:Vauk.

Tegegne Gebre Egziabher, Mulat Demeke, and Roth, M. 1999. Food Availability & Resource:
A Micro Perspective: The Case of South Wollo and East Gojjam in Ethiopia. Addis Ababa.

77
United Nations Centre for Human Settlements. 2000. An Overview of Urban Poverty in Addis
Ababa. Poverty Study. Addis Ababa.

United Nations. 1998. World Urbanization Prospects. The 1996 Revision. New York.

Von Braun, J., Bouis, H., Kumar, S., and Lorch, R. 1992. Improving Food Security of the
Poor: Concept, Policy and Programmes. IFPRI, Washington, D.C.

Wiebe, K., and Maxwell, D. 1998. Land Tenure and Food Security: A Review of Concepts,
Evidence and Methods. Research Paper No 129. University of Wisconsin, Madison.

World Bank. 1986. Poverty and Hunger: Issues and Options for Food Security in Developing
Countries. A World Bank Policy Study. Washington, D.C.

78
8. APPENDICES

79
Appendix 1. Summary of survey questionnaire

1. General information: Kebele, Household number, Interview Date, Interviewer’s

2. Basic household information: Household list (number), Relation to household head, Sex,
Age, Marital status, Final grade achieved, Main current activity, Religion
3. Food consumption and expenditure during last 7 days: Item code, List of food items, Unit,
Purchased food available to household with unit price and quantity, Non – purchased food,
Total consumption from all source.
4. Non – food expenditure during last month: Item code, Unit, List of type of expenditure,
Total value
5. Household income during last month: Income earner of household, Income source and
amount in Birr
6. Has there been any change in household income during the last five years? Yes = 1 No= 2
If yes what changes? 1= Increased 2= Decreased
If decreased what were the reasons?
7. For how long have you lived in Dire Dawa town?
8. Household asset holdings: List of assets and quantity
Have you sold any of these assets during the last one year?
If yes specify the reasons
9. Do the household involved in urban agriculture?
If yes indicate type of activity and purpose
10. Do you feel that it is important or necessary to provide credit for your household?
11. Have you or any one in the household borrowed any type of credit in 1996 E.C and if yes
for what purpose?
12. If borrowed credit, what was your source of credit?
13. Do you have saving account or a member of saving credit association?
14. Availability of infrastructure: Source of water, Power supply, Residential house

80
Appendix 2. Calorie value of food items consumed by sample households
_____________________________________________________________________
Food item Unit Kcal
_____________________________________________________________________
Teff Kg 3589
Wheat Kg 3623
Sorghum Kg 3805
Maize Kg 3751
Barley Kg 3723
Oat Kg 3599
Peas Kg 3553
Lentils Kg 3522
Fenugreek Kg 3824
Irish potato Kg 1037
Sweet potato Kg 1360
Onion Kg 713
Meat Kg 1148
Milk Lt 737
Egg Each 61
Butter Kg 7364
Edible oil Lt 8964
Coffee Kg 1103
Sugar Kg 3850
Spaghetti/Macaroni Kg 3550*
________________________________________________________________________
Source: EHNRI, 1998; * Dire Dawa Food Complex

81
Appendix 3. Conversion factor used to calculate adult equivalent
__________________________________________________________________________
Age category (Years) Female Male
___________________________________________________________________________

Less than 10 years 0.60 0.60

10 – 13 0.80 0.90
14 – 16 0.75 1.00
17 – 50 0.75 1.00
Greater than 50 0.75 1.00
Source: Institute Pan African Pour le Development (1981); cited in Strock et al. 1991

82
83