Вы находитесь на странице: 1из 121

ETHIOPIAN CIVIL SERVICE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

FOOD SECURITY AND ITS DETERMINANTES IN RURAL ETHIOPIA: THE CASE OF KAMBA DISTRICT IN GAMO GOFA ZONE

BY MELKAMU MADA DILNESHU

June, 2011 Addis Ababa

Food security and its determinants in the rural Ethiopia: the case of Kamba district in Gamo Gofa zone

A Thesis submitted to the Department of Development Economics Institute of public management and Development Studies Ethiopian Civil Service University In Partial Fulfillment of the Requirements for MSc Degree in Development Economics

By Melkamu Mada

June, 2011 Addis Ababa

Ethiopia Civil Service University


School of Graduate Studies
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: Melkamu Mada Entitled: Food security and its determinants in the rural Ethiopia: the case of Kamba district in Gamo Gofa zone and recommended that it be accepted as fulfilling the thesis requirement for the degree of: Master of Science in Development Economics.

Approved by Board of Examiners:Chairman, Department Graduate Committee Signature ___________________________________ Date _____________ Major Advisor Signature ___________________________________ Date _____________ Internal Examiner Signature ___________________________________ Date _____________ External Examiner Signature ___________________________________ Date _____________

DEDICATION
I dedicate this thesis manuscript to both my uncle

MELESSE

DILNESHU, for he brought up us after the sudden death of our mother, and to our mother HETEKE DUSSO whom we lost at our infancy

ABBREVIATIONS AND ACRONYMS


ADLI AE AIDS DAP DAs DPPA ECA EGS EHNRI FAO FEWS FGT FSCO FTCs GDP HFS HH HIV ICRA IFAD IFPRI ILRI Kcal Kg Km LDCs m.a.s.l ML Agricultural Development Lead Industrialization Adult Equivalent Acquired Immunodeficiency Syndrome Di-ammonium Phosphate Development Agents Disaster Prevention and Preparedness Agency Ethiopian Central statistics Authority Employment Generation Scheme Ethiopia Health and Nutrition Research Institute Food and Agricultural Organization Food and Early Warning System Foster, Greere and Thorbecke Food Security Coordination Organization Farmers Training Centers Gross Domestic Product Household Food Security Household Human Immunodeficiency Virus International Crop Research Agency International Fund for Agricultural Development International Food Policy Research Institute International Livestock Research Institute Kilocalorie Kilogram Kilo meter less Developed Countries Meter above sea level Maximum Likelihood
i

MoARD MoFED NGOs PAs PASDEP PSNP RBS RDFS RR10 SD SNNPR UN UNDP VIF WB WFP WoA

Ministry of Agriculture and Rural Development Ministry of Finance and Economic Development Non Governmental Organizations Peasant Associations A Plan for Accelerated and Sustain Development to End Poverty Productive Safety Net Programme Regional base line survey Rural Development and Food Security Rural Rode Standard Deviation South Nations Nationalities Peoples Region United Nations United Nations Development Programme Variance Inflation Factor World Bank World Food Programme Woreda Agricultural Office

ii

BIOGRAPHY
Melkamu Mada was born in Dorze, Chencha woreda, Gamo Gofa zone of South Nations Nationalities Peoples Region on November, 1976. He attended his junior and secondary education (9-12) at Chencha High School. He joined the then Jimma College of Agriculture in 1999 and graduated in diploma in Agriculture (Horticulture) on July 2000. After graduation, he was employed by the Ministry of Agriculture in the Kamba woreda and served for about two years. Then, he joined the then Alemaya University of Agriculture (AUA) in 2003 and graduated with B.Sc. Degree in Agriculture (Agricultural Economics) on July 2007. The author served as disaster prevention and preparedness expert and woreda agriculture Office Head from October 2000 to until he joined the school of graduate studies at Ethiopian Civil Service University on September 2009/2010 to specialize in Development Economics.

iii

ACKNOWLEDGEMENT I would like to express my deep gratitude to my research advisor Tadesse Kuma (Ph.D.) for his professional suggestions, thought guidance and over-all assistance from the beginning of the research proposal up to the end of thesis write up. Thus, I am very much indebted to him for all his support and willingness to advise me on my day-to-day efforts, and this has enabled me to finalize the study.

I wish to greatly acknowledge the financial support provided by Ethiopian Civil Service University for sponsoring and covering the costs associated with data collection for the thesis work.

My special thanks also go to the Kamba woreda administration and agriculture office without whose permission to join the M.Sc. program and without whose support this research document would have not been able to be materialized. My appreciation and thanks are also extended to Mr. Monaye Mosole, Mr. Asrat Agoze, Mr. Debebe Mulgeta, Mr. Alene Abeje and Dr. Tsgalem H/Mikael for their positive and immediate response and provision of computer facility. My genuine appreciation goes to all Kamba woreda agriculture office employees for their unreserved and unforgettable assistance and encouragements during my study and data collection campaigns.

I would like to extend my special thanks to Kamba woreda development agents Abraham Gamo, Asire Ayele, Andargachew Jemaneh, Mesafint H/Michael, Reta Wolde, Said Khalifa, Takele Mukulo, Tirunesh Lema, Wogaso Gassa, Zena Ali, Zena Dea who devoted their precious time for the data collection of this thesis. I am also very much grateful to 200 rural farmers for their

participation and for devotion of their precious time in answering the research interview questions.

No words can suffice to express my feelings of gratitude to my brother Ato Mitiku Mada(LLM) and his wife weizero Senait Tadesse and Akililu Mada for their financial and moral support which they offeried to me throughout the duration of study years.

Finally, my deepest gratitude goes to my wife Marta Beyene who had shown tremendous patience, made sacrifices, and shouldered the social, economic and family affairs including nursing of my sons during the time of my graduate program study.

iv

TABLE OF CONTENTS Page ABBREVIATIONS AND ACRONYMS.....I BIOGRAPHY ........................................ III ACKNOWLEDGEMENT ........................................ IV TABLE OF CONTENT...V LIST OF TABLESVIII ABSTRACT.. ........................................ IX 1 CHAPTER ONE ....1 INTRODUCTION.....1 1.1 Background ................................................................................................................................... 1 1.2 Statement of the Problem ............................................................................................................... 4 1.3 Objectives of the Study .................................................................................................................. 6 1.4 Significance of the Study ............................................................................................................... 6 1.5 Limitations of the Study ................................................................................................................. 7 1.6 Organizations of the Study ............................................................................................................. 8 2. CHAPTER TWO ..9 REVIEW OF LITERATURE9 2.1 Definitions and Overview of Food Security.................................................................................. 9 2.1.1 Food security defined .............................................................................................................. 9 2.1.1.1 Sufficiency ........................................................................................................................... 9 2.1.1.2 Access................................................................................................................................ 10 2.1.1.3 Security.............................................................................................................................. 10 2.1.1.4 Time .................................................................................................................................. 10 2.1.1.5 Vulnerable ......................................................................................................................... 10 2.1.2 Household food security: An economic perspective ................................................................... 11 2.1.3 Determinants of Household food Insecurity............................................................................... 12 2.2. Measures taken to overcome the food security problem ............................................................... 17 2.2.1 Agricultural Extension Services............................................................................................. 18 2.2.2 Agricultural Research ............................................................................................................ 18 2.2.3 Food Security Programme ..................................................................................................... 19 2.2.4 Productive Safety Net Programme ......................................................................................... 20 2.2.5 Voluntary Resettlement Programme ...................................................................................... 20 2.2.6 Water harvesting ................................................................................................................... 21 2.2.7 Expansion of Small-scale irrigation ....................................................................................... 21 v

2. 3 Main challenges that affect the success of food security .............................................................. 22 2. 3.1 Climate change .................................................................................................................... 22 2.3.2 Technology adoption problem ............................................................................................... 22 2. 3.3 Rapid and unhindered population growth.............................................................................. 23 2.3. 4 Poor Input and output market system .................................................................................... 23 2.4 Empirical Literature ..................................................................................................................... 24 2.5 Methodologies of measuring food security: .................................................................................. 27 2.5.1 Household Caloric Acquisition .............................................................................................. 28 2.5.2 Method for generating data.................................................................................................... 28 2.5. 3 Advantages of this method ................................................................................................... 28 2.6 Expected Contributions of the study ............................................................................................. 29 2.7 Comments on the Reviewed Literatures.....29 CHAPTER THREE .....30 RESEARCH METHODOLGY....30 3.1 Description of the study area 30 3.1.2 Livelihood Strategies in the study area .................................................................................. 30 3.1.3 Population ............................................................................................................................. 31 3.1.4 Religion ................................................................................................................................ 32 3.1.5 Agriculture ............................................................................................................................ 32 3.1.6 Crop Production ................................................................................................................... 33 3.1.7 Livestock and Poultry Production .......................................................................................... 33 3.1.8 Agricultural Extension .......................................................................................................... 34 3.1.9 Input Supply ......................................................................................................................... 34 3.1.10 Rural road ........................................................................................................................... 34 3.1.11 Market......35 3.12 Unexploited opportunities that exist in the woreda.....35 3.2 Methods and Sources of data collection....35 3.2.1 Methods of data collection35 3.2.2 Method of data analysis............................................................................................................37 3.3 Theoretical model ............................................................................................................................ 38 3.4 Empirical model .............................................................................................................................. 42 3.5 Variables and working hypothesis.43 3.6 The dependent variable of the model.................................................................................................44 3.7 Independent variables....44 3.8 Interpretation of the coefficients of the logistic regression model49 vi

3.9 Marginal effects result in logistic model........................................................................................... 50 3.10 Testing Multicollinearity...50 3.11 Measuring intensity of Food security ....51 4. CHAPTER FOUR ...52 RESULTS AND DISCUSSION..52 4.1 Measuring food security status of households .............................................................................. 52 4.2 Demographic and socio-economic characteristics ......................................................................... 53 4.3 Incidence of food security by households factors.66 4.4 Extent of Food security.................................................................68 4.5 Summary of mean difference and household scores 68 4.6 Analysis of determinants of food security.. 70 4.7 Discussion on the Significant Explanatory Variables .................................................................... 73 4.8 Major Agricultural Problems........................................................................................................ 78 4.9 Extension services ...................................................................................................................... 79 4.10 Coping Strategies ....................................................................................................................... 80 CHAPTER FIVE ...................................................................................................................................... 84 SUMMARY AND RECOMMENDATION83 5.1 Summary of findings ................................................................................................................. 83 5.2 Recommendations........................................................................................................................ 87 Reference.....91 Appendices . 96 Appendix 1:- Conversion Factors to Estimate Tropical Livestock Unit equivalents .................96 Appendix 2: Conversion Factors Used to Compute Adult-Equivalent (AE) .96 Appendix 3: calorie value of food items consumed by sample households...97 Appendix 4: Summary of the Survey Questionnaire .98

vii

LIST OF TABLES

Page

Table 1 - population growth rate of Kamba woreda.. ......... 31 Table 2-Religion distribution of the study area....... 31 Table 3- Types, codes and definition of variables in the logit model.49 Table 4- Food security status of sample households .......... 53 Table 5-Distribution of sample households by family size in number.. ......... 54 Table 6- Household food security by sex of household head. ............ 55 Table 7- Household food security by educational status of household head. ............. 56 Table 8- Distribution of sample households by number of months food item purcased.57 Table 9- Distribution of sample household heads by age. .......... 58 Table 10 - Distribution of sample households by farm size per household(in hectare)..59 Table 11- Distribution of sample households by Livestock Holding by TLU.. ............. 60 Table 12- Distribution of sample households by oxen ........... 61 Table 13- Distribution of households by amount of food aid received in Kg ................ 62 Table 14- Distribution of sample household by status of use of technology...63 Table 15- Distribution of sample households by off-farm income in birr...64 Table 16- Distribution of sample households by irrigation farm size in hectare65 Table 17- Distribution of sample households by infrastructure distance in Km ............................. 66 Table 18:- Incidence of food security by sample households.67 Table 19- Summary Statistics of continuous variables included in the descriptive statistics.69 Table 20: Summary statistics of discrete variables included in the descriptive statistics .................. 70 Table- 21 VIF value of continuous variables .................................... 71 Table 22- Contingency coefficient for discreet variables....71 Table 23- The maximum likelihood estimates of the logit model.. ............ 72 Table 24-Major agricultural problems encountered in the study area ....................... 79 Table 25:- Coping strategies common in Kamba woreda........... 81

viii

ABSTRACT A better understanding of the factors that affect the status of food security at household level is required for the organization of technical research for formulation of the development of policies and for shaping the direction of action for food self-sufficiency. Consequently, this study is expected to generate ideas that would be useful to reveal the seriousness of the problem and identify the determinants of household food security and coping strategies and recommendations. In order to achieve these objectives biophysical, demographic and socio-economic data were collected from 200 randomly selected households in Kamba woreda of Gamo Gofa Zone in South Nations Nationalities Peoples Region. A two stage sampling procedure was used to select 13 PAs, and 200 sample respondents from a total of 38 PAs in the woreda. A survey was conducted to collect the primary data from sample respondents. Supplementary, secondary data were collected from various sources. The data was analyzed using descriptive statistics like mean, standard deviation, percentage and frequency distribution. Univariate analysis such as T-test and Chi-square (2) tests were also used to describe characteristics of food secure and food insecure groups. The survey results show that only 39.5% of sample farmers were food secure. A logistic regression model was fitted to analyze the potential variables that affect households' food security status in the study area. Among 13 explanatory variables included in the logistic model, 10 of them were significant at less than 5% level of significance. These were, family size, farm size, number of livestock owned, total annual off-farm income, educational level of household, technological adoption of household head, household head participation in public meeting, household head extension contact trend, number of months food purchased and quality of land .The estimated model correctly predicted 96.5% of the sample cases, 96.1% food secure and 96.7% food insecure. Thus, identifying, analyzing and understanding those factors that are responsible for household food security status and its determinants is important to combat the problem of food security at the household level. The study findings suggest that in selecting priority intervention areas, the food security strategy should consider statistically significant variables as the most important areas.

ix

CHAPTER ONE INTRODUCTION 1.1 Background


Ethiopia, as one of the low income countries of the world, is facing repeated macro-and micro level food insecurity coupled with environmental degradation and depletion. In the last three decades, it has not been possible to produce adequate food to meet the needs of the fast growing population, attributed mainly to fragmented land holdings, successive droughts, untimely and unpredictable rainfall, antiquated farm technology, lack of farm input, low producer prices and other ecological factors. Earlier studies have estimated Ethiopias food insecure people to be around 40-50 percent of total population and at least 50% of farm house production does not satisfy basic needs and most of them face a hunger season every year (Yohannes, 2002; Devereux, 2001). The proportion of people who are unable to attain their minimum nutritional requirement is reported to be thirty three (33%) for rural and thirty (30%) for urban population (IFPRI, 2009). In Ethiopia, food insecurity is seen as the most important feature of development challenges. Every year, more than 4 million people, particularly in the rural areas have problems of getting enough food for themselves (Tassew, 2004). Currently, 5.23 million people will continue to require emergency food assistance up to December 2010 according to the Joint Government and Humanitarian Partners February 2010. Report of FAO/WFP in 2004, 2007 and 2011 shows the existence of bumper harvests at the national level. But achieving food security at national or regional level does not necessarily guarantee food security at district or households level. There exists disparity among districts and households in production function and motives for productivity. Even if a households is food secure it does not ensure that each member of the households is food secure due to discrimination in food distribution within households'. The geographical, environmental and medical factors that affect food security are important for their respective fields but social factors are significant for policy making and use by development practitioners.

Food insecurity in SNNPR remains a multifaceted and complex problem in which lack of access, as well as availability and quality of food still play an essential role. Several factors should be taken into consideration including extreme poverty, poor access to infrastructure, lack of productive assets, weaknesses in the marketing system and transport bottlenecks and others. According to a recent estimate, about 59 percent of the households in the region are labeled as food insecure (RDFS 2009).Regional average per capita calorie availability is about 1800 which is less than the international minimum standard of 2100 calories and much less than the standard for an adequate diet of 2400calories (SNNPR Regional base line survey,2007). Despite the overall improvement in alleviate food security, most poor and very poor households in these areas (20 to 35 percent of the total population) face chronic food shortages due to very high population density, shortage of land, and declining soil fertility. According to the regional baselines, even in a normal year, the poor and very poor in most of these regions rely on food aid (emergency or PSNP programs) to meet 525 percent of their basic food requirements (FEWS NET 2010). In the region, population growth is as high as 3.6% in some years, far above the national average of 2.9%, implying population may have exceeded the carrying capacity of the already fragile environment production capacity (SNNPR Regional base line survey, 2009). This shows that in the region food supply and demand are not moving in equal pace. Kamba woreda, where this research was done is one of fifteen woredas of Gamo Gofa zone. It has 38 peasant associations (PAs) with total population of 155,748. Out of this, 5,612(3%) live in urban and 150,132 (97%) live in rural (CSA, 2007). It has three agro-ecologies: highlands (dega), midlands (woinadega) and lowlands (kola). It is about 635kms away from Addis Ababa and 105km from Arbaminch, zonal town.

Kamba woreda is one of the major food deficit and famine-prone woreda in Gamo Gofa zone: SNNPR. Food insecurity, poverty and vulnerability to livelihood crises have increased in the woreda since the drought years of the middle 1980s and early 1990s (WFP 2003). Despite the overall improvement in the level of food security in the region and in the country, most poor and very poor households in Kamba face chronic food shortages due to very high population density, shortage of land, poor access to infrastructure and declining soil fertility. According to
2

the livelihoods baselines survey (1994/95), even in a normal year, the poor and very poor in Kamba woreda rely on food aid (emergency or PSNP programs) to meet 525 percent of their households food requirements. A review of food aid recipients in the different kebeles of the woreda shows that the number of households that depend on seasonal food assistance has increased from year to year. According to Kamba woreda agricultural office early warning report (2003-2009) the number of food aid recipient increased from year to year, 14,200 beneficiaries in 2003 and 22,062 beneficiaries in 2009. Food security problem in Kamba derived directly from dependence on undiversified livelihoods based on low-input, low-output, full dependence on rain fed agriculture, decreasing land holding size and increasing population, week agricultural extension services, recurrent drought and natural resource degradation (land, water, forest, and rangeland), persistent livestock disease and lack of market access, lack of alternative off-farm employment and others in the study area have made the food security situation worse. Kamba woreda is not among those areas that are showing agricultural growth and technological change in the region. Rather, Kamba agrarian conditions represent continuity more than change, as demonstrated by its inability to meet the food demand of the woredas rapidly growing population.

The dominant factors that contribute for the chronic food insecurity problem in the Kamba woreda are poor crop and pasture production, infrastructure problem mainly road, abnormal food price and livestock disease. The poor performance of the agricultural sector in the area directly creates supply problems and indirectly creates demand problems by contradicting the producers access to sufficient income. The agricultural productivity in the area strongly depends on rain fed, which mostly behaves an irregular start and end (poor distribution). In that area, below normal rain is common and leads to poor harvest, poor pasture and poor food availability. Kamba woreda is known with its road problem. Before fifteen years, the woreda faced serious drought and due to road problem life saving rations were dropped from air. Even nowadays no food crops are transported in to the woreda and out of the woreda. Households in the area face unusual prices during normal time and farmers cannot get fair price for their products during good harvest seasons.

Food self-sufficiency has remained the declared goal for federal, regional and woreda government of Ethiopia. Similarly, the problem of food security has continued to persist in the rural areas like Kamba. The Kamba woreda administration and different NGOs implement food security programs like productive safety net, voluntary resettlement, provision of improved technologies for the farmers, water harvesting programs have been implemented in the area. But the progress is not significant, so assessment of food security challenges at woreda and households level to identify the major determinants is fundamental. Having this background, this study tries to investigate the food security and its determinates in rural households in Kamba woreda, Gamo Gofa zone: SNNPR. 1.2 Statement of the Problem Over the past ten decades, Ethiopia has been challenged by food insecurity problem. In Ethiopia, the trend in growth of domestic food production matched population growth only in the 1960s (Markos, 1997). In spite of the fact that Ethiopia has abundant natural resources, most of its socioeconomic indicators are extremely low and discouraging. Numerous studies have confirmed that there is a problem of food security in Ethiopia with wide range of area to be covered and large number of people to be attended for different identified causes of food security problem. Among these causal factors, per capita land holding with increasing population growth, livestock availability, education, per capita income of the households from agricultural and non

agriculture activities, soil fertility, conflict, under-funded agriculture are the major and commonly mentioned factors (Gebre-Selassie, 2005, Negatu, 2004,Ramakirshina et al, 2002, Madeley 2000).

Ethiopian government and international donors have been implementing different categories of responses to attain food selfsufficiency and to reduce food aid dependency. These categories are based on supply based responses (increasing the level and stability of production, increasing food reserve, and influencing international food markets), demand based responses (improving income, productive assets available to vulnerable groups, and other market and non-market transfer), and disaster prevention and preparedness capabilities having adequate early warning systems (IFPRI, 2003). Despite such efforts, food insecurity remains the main challenging

problem in our country and the need for food aid has been increasing. This shows that there is mismatch between food demand and supply.

In the last two decades, Kamba woreda have been identified as chronically food insecure area and that cannot adequately feed its rapidly growing population .In the woreda, farmers do not produce enough food even in good harvest seasons to meet their own consumption requirements. Therefore, they have continued to depend on relief assistance to fill the food deficit. As a

result of this annual food deficit people migrate from highlands every year in search of jobs outside the district and within the district. High dependence on relief assistance and the existence of migration shown there is high annual variability in food production and availability, which confirms the existence of food security problem among the households of the district. Making their living on marginal, moisture stressed, heavily degraded and less productive land, households in the Kamba woreda have been facing persistent food shortages. On top of the ever decreasing size of land holding due to increasing population, continuously decreasing per-capita production, high dependence on rain fed agriculture, recurrent drought and natural resource (land, water, forest, and rangeland)degradation, lack of infra-structures, lack of access to market, persistent animal diseases, week extension system(absence of credit service, lack of improved technologies, absence of crop diversification), low family income, lack of alternative off-farm employment, practicing bad cultural habits and others in the study area have made the food security situation worse. Realizing this issue, many governmental and non-governmental organizations are intervening at least to decrease the adverse effects of the food security problem. But, there is yet little success. Aware of these facts, this study was designed to identify the factors that contribute to households food insecurity problem and to the severity of the problem in the rural areas of Kamba woreda, and through that make recommendations to improve the effectiveness of interventions. Having this background in mind the study has put forward the following research questions.

Research questions: What is food security? What efforts have been made by the international and national governments to eradicate the problem? What are the success stories and the challenges that have been faced? What are the determinant factors that affect food security problem in the rural households in the Kamba woreda? What coping strategies do food insecure households uses to make their livelihood during time of sever food shortage? What is the contribution of governmental and non- governmental programs to address the problem? To address the above research questions, the study has applied logistic regression method of analysis by using data generated from 200 households.

1.3 Objectives of the Study The main objective of this study is to analyze the food security status of the rural households and its determinants in the Kamba woreda of Gamo Gofa zone: SNNPR. The specific objectives of this study include: 1. To investigate how households level socio-demographic structure affect food security. 2. To identify the major factors that contribute for some households attain food security while others not. 3. To record coping mechanisms of households during adverse covariant shocks. 4. To investigate how households food security status is related to households food intake and dietary diversity. 5. To identify policy measure those improve the households status of food security problem in the area. 1.4 Significance of the Study Population growth coupled with environmental degradation, drought, a largely stagnant agricultural technology, poor institutional arrangements, inefficient output and input markets, inadequate infrastructure and external factors have all contributed to the aggravation of food security problem in the study area. The woreda administration of the woreda study area has been
6

trying to increase food security at both households and woreda levels through these small-scale farmers. Despite all these efforts, majority of households in the area have been facing series food security challenges. Identifying determinates of households food security problems have been studied by many researchers in the country. This study has focused on assessing determinates of households food security problem in Kamba woreda in which no similar research has been done to the best of own knowledge. The various factors that affect households food security status were discussed so that recommendations can be made for the adoption of better strategies and actions to assist small-scale producer farmers to address households food security problem in the woreda. This study helps to identify gaps in smallscale farmers and come up with reasons why small- scale farmers are food insecure.

The result of the study provides policy related information that helps to prioritize among the many possibilities depending on the relative extent of influences of its determinants. More specifically, it helps concerned bodies in their effort to formulate policies and develop intervention mechanisms that are tailored to the specific need of the study area. Furthermore, this study attempts to make further contribution to the previous studies and can be used as a source material for further studies.

1.5 Limitations and scope of the study The study specifically has focused on identifying major determinants of food security problem at households level by comparing direct calorie consumption per adult equivalent with the minimum requirement by classifying sample households as food secure and insecure and then looking at the extent of food security problem in Kamba woreda. The study covers only 13 of the 38 PAs of the study area. Moreover, the study deals with a limited number of households and focused on the dimensions and determinants of food security problems. Besides to this, the data were collected at one time period and during the time of food shortage faced by the households in the study area. The scope of this study was limited by time, budget and other resource limitation. Even if the study was restricted in terms of its coverage its outputs can be used as a spring board for more detailed and area specific studies.

1.6 Organization of the study The rest of this thesis is organized in six chapters. Chapter two deals with review of literature that includes theoretical frameworks of food security and empirical studies made in the country and elsewhere in the world. Chapter three presents a brief description of the study area while chapter four deals with methodology of the research. Results obtained are discussed in detail in chapter five. Chapter six presents summary and conclusions of the study.

CHAPTER TWO REVIEW OF LITERATURE


2. 1 Definitions and Overview of Food Security This section reviews the literature on food security. This will be done in four sub-sections. Section one provides brief definitions of what food security is and determinants of households food security are discussed in this section. In section two much of the discussion will revolve around what efforts have been made at global, regional and at country level to close down the food security problem and what are success and failures of the past efforts. In section three a brief discussion of empirical research that has been carried out on food security in the country and outside the country, and in the last section how food security has been measured and expected contribution of this study for the literature world is discussed.

2.1.1 Food security defined Food security is defined in different ways by international organizations and researchers. According to Smith et al. (quoted in Maxwell, 1996), there are close to 200 definitions of food Security but the one most commonly used and this study concentrates is the one given by the Food and Agricultural Organization (FAO) which states that, Food security exists when all people, at all times, have physical and economic access to sufficient, safe and nutritious food to meet their dietary needs and food preferences for an active and healthy life (FAO 1996, Rome Declaration 1996, page 8). Food security may be generally understood as, "Secure Access to Sufficient Food at All Times. This definition integrates access to food, availability of food, and the biological utilization of food as well as the stability of all these. These details are provided below: 2.1.1.1 Sufficiency: Sufficiency is commonly understood and measured by the number of calories an individual requires to live an active and healthy life (S. Maxwell & M. Smith, 1992).Although daily caloric requirements for men, women and children have been estimated by nutritionists, the broad applicability of these caloric intake standards has been questioned(K.
9

Bagachi et al,2002).An individual's perception of his/her own hunger and the amount of food required to live an active and healthy life have been found to vary greatly(J. Coates, et al, 2002).Despite these debates, food sufficiency is generally understood as the amount of food required by an individual to meet dietary energy requirements as defined by the Food and Agricultural Organization(R. Gardiner, 2002).

2.1.1.2 Access: Access to the means of acquiring food, whether through production, purchase
or exchange, is central to contemporary definitions of food security. The key determinants of food access are physical, economic, political and socio-cultural. Human and physical capital, households assets, property resources, markets, and a variety of social contracts at the households, community and state levels, directly affect an individuals level of access (S. Maxwell & M. Smith, 1992).Access to the means of food acquisition is highly dependent on gender biases and social forces.

2.1.1.3 Security: Security refers to the extent to which the access to the means of acquiring
food is vulnerable. Vulnerability is a function of risk and insurance. Asset ownership, human and social capital, government safety nets, and the activities of non-governmental organizations (NGO's) all contribute to an individual or household level of insurance, while variability in crop production, changes in food prices and food supply, employment, and wages all influence the risk profile (S. Maxwell & M. Smith, 1992).

2.1.1.4 Time: Food insecurity may be chronic, cyclic or transitory based on factors such as seasonal patterns and income regularity. Households or individuals that are continually unable, or are at risk of being unable, to meet caloric needs are categorized as chronically food insecure. Households experiencing a temporary decline in the ability to meet caloric needs is categorized as having transitory food insecurity. Households experiencing re-occurring, regularly timed periods of food insecurity (e.g. hunger seasons) are understood to be facing cyclic food insecurity (S. Maxwell & M. Smith, 1992).

2.1.1.5 Vulnerable: Vulnerability is strongly related to the concept of food insecurity, highlighting the element of risk that households face in their production, income, and
10

consumption activities. Vulnerability can be defined as the likelihood that a specific population group will experience an acute decline in their food access. In addition to the risks that households face, vulnerability further implies that these groups are unable to sufficiently cope with those threats to effectively protect their basic food access. Female-headed households, the aged, the disabled, and other disadvantaged groups with low levels of households labor and insufficient means of support from family members and the community are also typically included as being among the most food insecure and vulnerable as well. Other households are vulnerable because they live in areas susceptible to natural or man-made disasters. households under the extreme threat of conflict, drought, and other risks, particularly those households lacking a diversified income and asset base to cope with those risks, are also considered among the most food insecure and vulnerable groups.

2.1.2 Households food security: An economic perspective Economic approaches to food security have traditionally focused on assessing aggregate levels of food supply, agricultural production, and the balance of agricultural trade (World Bank1986). In the 1970s, food security was defined at the macro level as the ability to avoid short-term deficits in aggregate food supply (Staatz JM Econ 1990) and it was directly linked to grain stocks at the global and national level (Holmboe-Ottesen G1990). During the 1980s, this macro-level understanding was replaced by a focus on individual-level and households-level concepts of food security, emphasizing access, vulnerability and entitlements (S. Maxwell & M. Smith.1992). This shift was largely influenced by the pioneering work of economist Amartya Sen.
From an economic perspective, malnutrition was increasingly recognized as the individual-level manifestation of a complex combination of households, community, regional, national, and international factors (Marek T, 1992). Seminal work on the phenomenon of famine by Sen (an essay on entitlement and deprivation,1981) brought attention to the issue of access to food by households and by individuals, which could be constrained by economic, social, and cultural factors and was most often a chronic, not transitory, condition at the households level. Food insecurity could occur at the households level, and was occurring, in the absence of regional and national food insecurity.

The neoclassical economic theory of households production added further to the concept of food security by emphasizing the decision-making processes within the households that determine
11

how scarce resources are allocated. Since households have limited access to resources and struggle to fulfill a variety of basic needs, procurement of food competes with acquisition of health services and other goods and services. Therefore, food needs are necessarily the most dominant basic needs for a given households subsistence or survival (Maxwell S, Frankenberger TR 1992). Macro-level food security focused on food availability (supply), most of the recent households food security frameworks are concerned primarily with households access to food, although all recognize that access is just one component of households food security as discussed above. At the households level, food security is determined by a household current food supplies, past stable food supply, and potential future supply. 2.1.3 Determinants of households food insecurity To obtain their food, households typically either: (a) grow it and consume from their own stocks; (b) purchase it in the marketplace; (c) receive it as a transfer from relatives, members of the community, the government, or foreign donors; or (d) gather it in the wild. The factors that limit the ability of households to grow, store, purchase, gather or receive transfers of food will vary by location, across socio-economic groups, and over time. Once the basic sources of food have been identified, it is necessary to investigate the often complex interaction of agro-physical and socio-economic processes that limit a households ability to obtain sufficient quantities of food from each source. Food security is generally affected by two major determinants: Availability of food and accessibility to it (Andersen, 1997). The Same source also showed that human resource development, non-food factors, including education, health care, and clean water, population growth, urbanization and displacement of people greatly influence food insecurity and human nutrition. This source further stipulated that natural resource and agricultural inputs are critical determinants of food security. Food insecurity is due to a variety of reasons, and the FAO/UNDP (1987) cited in Getachew (1995) suggested, i) the relatively high density of human and livestock populations and the resulting squeeze of land resources; ii) the inability of agricultural practices to sustain the required productivity levels of land; iii) insufficient level of adoption of modern
12

farm technology; iv) extensive and often irreversible levels of land degradation; v) the value placed on livestock, specially cattle, in the social economic system and the accomplishing desire to maintain large livestock holdings.

According to Hoddinott (2001) households food security (HFS) issues cannot be seen in isolation from broader factors. He viewed these factors as physical, policy and social environment. And he argued that the physical factor play a large role in determining the type of activities that can be undertaken by rural households. Government policies on the other hand toward the agricultural sector will have a strong effect on the design and implementation of households food security interventions. Likewise the presence of social conflict, expressed in terms of mistrust of other social groups or even outright violence, is also an important factor in the design and implementation of interventions. Hoddinott (2001) expressed that resources or endowments that food security of households can be divided into two broad categories: labor and capital. Labor refers to the availability of labor for production. It incorporates both physical dimension-how many people are available to work as well asknowledge and human capital dimensions. On the other hand, capital refers to those resources such as land, tools for agricultural and non agricultural production, livestock, and financial resources; that when combined with labor produce income. In turn the house- holds allocate these endowments across different activities such as food production, cash crop production and non-agricultural incomegenerating activities in response to the returns each activity generates. In addition, households may receive transfer income from different sources, which determines households income.

Hoddinott (2001) further described that households face a set of prices that determine the level of consumption that can be supported by the given level of income. Accordingly, consumption is divided between those goods that affect households in individual food security and all other goods. Goods that affect food security include food consumption at the households level (referred to as food access in much of food security literature), goods directly related to health care; and goods that affect the health environment. These three goods affect illness & individual food intake, which in turn generates nutritional status or food utilization. Finally Hoddinott

13

(2001) noted that food security is not static over time. There are second rounds, or feedback effects.

Lathan (1997) has clearly indicated that income received from cash crops or wage earnings and prices paid for purchased items influence a rural populations food security. Further, the author stated that inadequate land holdings; landlessness and sharecropping are all powerful causes of family insecurity. Lathan has also identified that a shock often precipitates households food insecurity. The shock can adversely influence food production (suddenly threatening farm food availability). There are many different kinds of shocks, like serious illness, which may result in reduced agricultural production in a farm family; loss of rural job; farm production crises, such as failure of the rains, or a plague of locusts or some other agricultural catastrophe. Any crisis that has an adverse impact on the livelihood of the family may also result in households food insecurity.

In the Horn of Africa, for example, a leading determinant of food insecurity is low levels of per capita food production. The primary constraints to improved food production in the region are a combination of low and erratic rainfall, high population densities, deforestation and, as a result, an accelerated deterioration in soil quality and crop yields. Poor market infrastructure and an unfavorable policy environment which leads to high and variable prices for inputs and low producer prices further undermine productivity in many countries in the region (Frank Riely, Nancy Mock1999).

A case study of resource and food security of Wobera District of East Hararghe Zone (Getachew, 1991) showed that sufficient conditions exist for chronic and transitory food insecurity among the households'. These conditions are: first, land, one of the most important resources for food production, is scarce among the study households. Second, other households resources such as livestock have fallen dramatically. Third, due to climatic hardship, even cereal major producing areas remain deficit, leaving both cereal and cash crop dependent households in a disadvantaged food supply position. Fourth, the administrative apparatus of Ethiopia (both

14

past and present) neglected the rural sector with no or realistic development strategies to reduce risks of food insecurity.

The same source further showed that agro-ecological induced variation of holding size and plot distribution and ox-ownership, as an important factor in determining households resource endowment and the ability to perform agricultural activities, came out to be factors which determine the food security situation among the sample households. Moreover, other factors that were given due attention in the study were labor, land-to-man ratio, ability of the area to offer cash crop and off-farm income, grazing land, households indebtedness, cash block (off farm employment income, cash crop income, livestock income and borrowing), market price, households expenditure (obligation to the state, rural institution, the households itself and other households).

In a case study of Social and Demographic Characteristics Habro woreda, using logistic regression model, Getachew (1993) showed that there is a statistically significant relationship between resources held by a households and its level of food security. It was confirmed that those households which hold land less than three Timad, do not own any oxen, have a small households adult equivalent size and earn non-farm income of less than Birr 500 (or nothing at all) are those most at risk of food insecurity among the sample population. Consequently, the researcher showed that the levels of income and farm size are the most important resources determining food security when other factors such as favorable climatic conditions and low pest outbreak are satisfied. In other words, a larger land size and high income increase the chances of maintaining food security. Poor target groups often lack access to institutions and services which could help them in improving their subsistence production and income (SLE, 1999; cited in Ayalneh 2002).

Moreover, it is a combination of availability, access and the chance of receiving external assistance that determines the households food security. As explained in FAO (1991) the problem of households food security is not simply one of agricultural output, but encompasses all factors affecting a households access to an adequate year round supply of food. Thus, the
15

problem of households food security is not simply one of next seasons crops, but can also include factors as diverse as deforestation, seasonal variations in food supply, availability of fodder and other forest foods, shifts from subsistence to the cash economy, and even the timing of cash needs as school fees.

Ayalneh (2002) in his study of Land Degradation, Impoverishment and Livelihood Strategies of Rural households in Ethiopia, showed that factors that have contributed to transitory and chronic food insecurity in rural Ethiopia are various and varied, ranging from political and socioeconomic to environmental. Among the political factors he listed inappropriate agricultural and marketing policies, and political conflict both at national and local level. Among the socioeconomic factors are demographic characteristics of rural households inadequate resource endowments, inadequately developed infrastructure such as school, hospital and roads, etc. The same source further stated that food security concerns in rural households depend to a large extent on the size and age structure of households members. The number of the households member capable of contributing to food production and/or who can be employed in non-farm income earning activities will greatly determine households own production and its capacity to acquire food through enhancing exchange entitlement.

Further, while Ayalneh (2002) explaining the feature of food insecure groups, he also implicitly explained the factors that determine food insecurity. In that the largest group of food insecure households is those who live on the edge of subsistence, often located in remote areas far from markets. They lack the important asset of good quality land and access to productive assets. Lack of draught power severely handicaps farmers, lack of access to credit, agricultural input and technology. Lack of male labor in female-headed households is another important constraint. They usually work in an insecure and low productivity occupation. Another determinant of food insecurity is gender discrimination. Subordination of women in society, their over-burdening and the greater difficulties faced by female-headed households contribute to food insecurity (Lathan, 1997).

16

On the other hand Gezahegn (1995) explained that the major causes of transitory food insecurity are failure in agricultural production or instability in food supplies resulting from stochastic shocks such as recurrent draught, lack of incentives to small scale food producers, and poor extension services for the small peasant households. The weak system of marketing and transport operations to procure and collect agricultural products from widely dispersed rural producers and to distribute essential agricultural inputs on time contributes not only to the fall in production in some years, but also to the problems caused by failure to move the available food itself to needy areas.

Getachew (1993) in a case study of Adama Boset Using logistic regression model he showed that there is statistically significant relationship between food insecurity and its determinants except farming system. He also reported that there are statistically multiple relationships between resources owned by a household and level of food security. Accordingly, it was confirmed that amongst the sample population it is those households which hold land less than or equal to 3 Timads, do not own any oxen, have a small households adult equivalent size, are unable to use fertilizer, and earn a non-farm income of less than Birr 500 (or none at all) which are most at risk of food insecurity. Thus ox-ownership, level of income and land size is the most important resources determining food security when other factors such as favorable climatic conditions and low pest out break are satisfied. In other words, an increased size of land, ox-ownership, high income and use of fertilizer increase the chances of maintaining food security. In his study using logistic model of Kembata and Hadiya district Getachew (1993) tested the significance of the relationship between households resources and food security. For the test he included six variables farming systems, land size, production output, livestock, households size and fertilizer. Moreover, in this study all the variables are negatively related with food insecurity except households size.

2.2. Measures taken to overcome the food security problem The government of Ethiopia has implemented a policy response specific to Ethiopia's food security and agricultural productivity challenge, including the Agricultural Development Led Industrialization (ADLI) strategy. This policy strategy divides the country into three main agro17

ecological zones, which include regions with adequate rainfall, moisture stress areas, and pastoral areas. In regions with adequate rainfall, the focus is on exerting all possible efforts to efficiently utilize available rainwater to bring about the maximum possible rate of agricultural development, and promoting irrigation in areas where it is feasible. In moisture stress areas, major activities undertaken to enhance food security are focused on increasing off-farm income opportunities, and voluntary resettlement to more productive areas. In terms of pastoral areas, special efforts are made to enhance specialization in livestock production and marketing through the provision of water supply for the community and their livestock as well. Major interventions in this area also include improving livestock quality, expansion of animal health services, water points, feed production and improvement of breeds and development of market infrastructure. Specific policy measures practiced by the government to improve food security and show positive results include: Agricultural extension and research, food security programme Productive safety net programme, Voluntary resettlement Programme, water harvesting programme, expansion of small-scale irrigation. 2.2.1 Agricultural Extension Services: A key feature of this innovative policy measure is the employment of medium level trained extension workers to every rural kebele in Ethiopia to facilitate sustained knowledge and skills transfer to smallholder farmers through the establishment of Farmers Training Centers (FTCs) to transfer improved agricultural technologies and give adequate services at a closer reach. Throughout the planning period, all of the female headed households have access to extension services and an estimated 30% of women in maleheaded households will also get access to training and extension services in the type of

extension packages that will benefit those most. This has contributed to increased agricultural productivity particularly for cereals, pulses, and oil seeds and decreased the number of food insecure farmers (UN- development strategies). 2.2.2 Agricultural Research: Effective agricultural research is also at the core of improving food security. The major outputs of the public research system are varieties with improved agronomic and protection practices that can be used in crop and livestock diversification and specialization, for both traditional food crops and livestock, as well as high value crops such as vegetables, spices and other horticultural crops. Research is also conducted on food science,
18

socioeconomic and post harvest technology as well as farm implements. The research centers maintain improved varieties and multiply breeder and pre-basic seeds and seedlings of released varieties of crops, and distribute them to farmers. Those farmers who applied improved inputs increased their productivity per area and decreased vulnerability to food insecurity (UNdevelopment strategies 2009). 2.2.3 Food Security Programme: The National Food Security Programme rests on three pillars: increasing the availability of food through domestic (own) production; ensuring access to food for food deficit households and, strengthening emergency response capabilities. But Ethiopia is constantly dependent on international aid for nearly 10% of its annual food need. This figure at times reaches about 25% in periods such as the drought years. This means that 4 to 5 million people are living in continuous risk of food shortage and their existence is directly related to external help. The effort to reduce food insecurity problem is a central part of the PASDEP strategy. Measures are being put in place to reduce the variance in crop production and food availability overall through more irrigation and water control, diversification of crops, and better integration of markets, transport, and information links. Maintaining macroeconomic stability; expanding offfarm employment and income-earning opportunities, and better functioning credit markets, improving health services and nutrition, and, innovative measures, such as experiments with crop and weather-based insurance mechanisms -are key components of these measures. The Government tries to strengthen and expand rural micro financing institutions and cooperatives to provide banking services especially in food insecure areas. The Government of Ethiopia recognizes the importance of improved credit services for food insecure rural and urban households in order to address both supply and demand side problems. The Government will continue to increase the availability of rural financing and provide special support in reducing the credit administration cost extended to food insecure communities. Cooperatives are expected to play a big role in this regard. Furthermore, the Government has gradually shifted away from food assistance (assistance in kind) towards financial assistance for the purchase of food from the domestic market. This has
19

helped augment the stocks of food security reserve in good times. This in turn contributed to the creation of effective demand through stabilization of prices. Since the launching of the Programme, remarkable achievement has been made in narrowing overall food gap from domestic production. The recent report by the Disaster Prevention and Preparedness Agency (DPPA) shows that the number of people requiring emergency assistance stood at 2.3 million in 2010/11from5.2 million in 2009 /10, evidence that the food security situation in the country has improved significantly (FEWS NET 2010).

2.2.4 Productive Safety Net Programme: The Productive Safety Net Programme is intended to bridge the income gap of chronically food insecure households and engage such households in community asset building efforts especially during the lean season and times of drought. Priority for households asset-building interventions is given to beneficiaries of the Safety Net Programme. The programme started in 2005 and covers 287 woreda. It has two components - labor-intensive public works and direct support for labor-poor households. The able-bodied are engaged in public works for which they are paid a minimum amount, while the labor poor are provided the same amount free. A key feature of the Safety Net Programme is its households focus. It is linked to the households asset-building efforts of the Food Security Programme in that the priority for households asset building interventions is assigned to those covered by the Safety Net Programme, as they are the chronically food insecure. The Safety Net Programme through its predictable transfer of resources will help prevent asset depletion, which is an important factor for the attainment of food security. This is in addition to the community assets it helps build. 2.2.5 Voluntary Resettlement Programme: Over the years, a large portion of the country's population has lost the productive capacity mainly due to land degradation and high population pressure, while at the same time Ethiopia has a considerable amount of land suitable for agriculture currently under-utilized. To rationalize resource use, and thereby help food insecure households, the Government is supporting voluntary resettlement as part of its food security
20

programme. To date, over 149,000 households have been resettled. Resettlement is on a purely voluntary basis, and each settler households is guaranteed a package of assistance that includes provision of up to 2 hectares of fertile land, seed, oxen, hand tools, tools, and food rations for the first eight months. Settlers are also provided access to essential social infrastructure (clean water, health post, feeder road), and logistics support. 2.2.6 Water harvesting:- In the area of water harvesting, the technique is considered to be one of the major interventions to overcome the challenges of households food insecurity in the country. In this regard, efforts are in progress to construct water-harvesting structures across the country. Water harvesting schemes are generally viable, with an economic rate of return of 32.7%, and an annual earnings of about ETB 6,400 households /annum. Accordingly, 139,462 different types of water harvesting structures with different capacities have been constructed in the four major regions. These structures are at households level and are meant for life saving irrigation for field crops as well as for vegetable gardening where conditions are favorable. Extension and training programmes are also designed to pay particular attention to enhancing farmers' capacity to use water resources efficiently, and to boost productivity for both households consumption and for market. 2.2.7 Expansion of Small-scale irrigation: The most direct and significant contribution to an improvement in food security for rural households has been made by the support furnished for small-scale irrigation development. Some 31,000 households in densely populated drought-prone areas have been reached through the support provided for small-scale irrigation projects, and many of these households are gradually seeing an improvement in their incomes. Findings suggest that increases in crop yields over the traditional yields are in the range of from 25 to 40 percent, and in cases where irrigation facilities have been built around springs, the increases have been between 75 and 100 per cent. Thanks to these irrigation developments, the targeted irrigation farmers physical and financial assets have started to increase, although experience suggests that it may take six or more years for irrigation farmers to experience the full benefits (IFAD 2009).

21

2.2.3 Main challenges that affect the success of food security: The extent of the food insecurity has become sound the alarm. As much as 45 percent of the population is affected during drought years. The serious and accelerating problems of food insecurity in Ethiopia are now considered a national security threat. So breaking the cycle of food insecurity problem has remained one of the difficult challenges towards poverty reduction in Ethiopia. Factors that may aggravate the situation are:2.2.3.1 Climate change: is an immediate and unprecedented threat to food security of millions of households that depends on small-scale agriculture for their livelihood. Climate change in Ethiopia affects all four dimensions of food security: Food production and availability, stability of food supplies, Access to food and Food utilization. ILRI (2006) and Stige et al. (2006) identify Ethiopia as one of the most vulnerable countries to climate change. Climate affects food production directly through changes in agro-ecological conditions and indirectly by affecting growth and distribution of incomes, and thus demand for agricultural produce. Greater fluctuation in crop yields and local food supplies can adversely affect the stability of food supplies and food security. Climatic fluctuations will be most pronounced in semi-arid and subhumid regions and are likely to reduce crop yields and livestock numbers and productivity which increase vulnerability. Climate change decrease access to food due to food price increases and declining rates of income growth. Climate change may initiate a vicious circle where infectious diseases, including water-borne diseases, cause or compound hunger, which, in turn, makes the affected population more susceptible to those diseases. Results may include declines in labour productivity and an increase in poverty, morbidity and mortality (Schmidhuber and Tubiello 2007). 2.2.3.2 Technology adoption problem: The country extension system inability to transmit improved technology and farming practices to producers in participatory and accountable ways worsen the food insecurity. Over the years, many development agendas and practices were prescribed in a top-down approach to the rural people and farmers. Potentially well designed programs and practices usually failed simply because farmers were not given the chance to ask why, where, how and when they should adopt these practices. Extension workers

22

mandated to persuade or teach farmers at grassroots level are quite often practically unskilled or even irresponsible.

2.2.3.3 Rapid and unhindered population growth: Rapid and unhindered population growth is a significant factor in exacerbating food shortages in Ethiopia the second most populated country in Africa according to a 2009 report by the UN Emergencies Unit for Ethiopia. That year, 20 percent of the population was dependent on foreign-supplied food aid. Of the countrys current 80 million people, an estimated 12 million Ethiopians are facing serious threats from food insecurity and famine. More than half of the countrys children under five years of age are stunted in growth and 47 percent are underweight. With one of the highest birth rates in the world, Ethiopias population is projected to increase by 20 million in the next 10 years and double by 2045. Forty-five percent of the population now is under age 15. Each Ethiopian woman gives birth to an average of six children, and 36 percent of married couples that desire to use contraception to space or limit childbirths do not have access to contraceptives. Population growth is putting unprecedented and increasing pressure on vital natural resources, including cropland and fresh water. Dozens of countries already have reached alarmingly low levels of available cropland. Currently, 17 million people live in countries with less than 0.5 hectare of cropland per person the minimum cropland capable of supplying a vegetarian diet for family (starvation plot). 2.2.3.4 Poor input and output market system: Lack of efficient and effective input and output market systems is a major obstacle to the Ethiopian agriculture and food security. In Ethiopia, surplus food may be produced in some parts of the country, but because of weak transport and market links it cannot be easily transported to food deficit areas. The countrys high transport costs and poor infrastructure aggravate the food insecurity situation for poor farmers. Studies on the grain market (Mulat et al. 1998) and input use confirm the existence and consequences of these problems. Improvement in market policy and investment in infrastructure is necessary to reduce food transaction costs and enhance the incentive to produce more and improve food security situation in the country.

23

2.4 Empirical Literature Food insecurity could be caused by several factors. Following the usual practice, we classify them into two broad categories- immediate and underlying community conditions. Under immediate conditions, we have low rates of agricultural production; low access to food resulting from low income; poor roads and infrastructure facilities. On the other hand, underlying community conditions include existing factors which could interrupt availability, accessibility, utilization, and stability of food. For example, if a community is characterized by poor infrastructural conditions, productive capabilities of farmers could be hampered as they will have limited access to new technologies, to credit, and also to storage and transportation facilities for inputs and outputs. Subsistence farming is also characterized by low yield and growing levels of soil fatigue as people remain in one area for extended periods and become less peripatetic than in the past.

The food security status of the community could also be affected negatively by bad local and international market conditions which could result from ill-designed domestic and international trade policies. These could reduce access to food by the community from local as well as outside sources. In addition, the food security status of the community could be frustrated by HIV and AIDS and other pandemics that harm economically productive sections of the society.

Studies show that food prices have increased sharply over the past five years. This has been attributed to a number of factors: rising energy prices and subsidized bio-fuel production; income and population growth; globalization and urbanization; land and water constraints underinvestment in rural infrastructure and agricultural innovation; and lack of access to inputs and water disruptions (FAO, 2008). This has made poorer households highly vulnerable to food insecurity by decreasing their purchasing power. Food constitutes the lions share of poor households expenditure or budget.

Study made in Zimbabwe on determinates of households food security in the semi-arid areas of lupine and hwange districts by (Misery Mpuzu 2008) using logistic regression conclude that having access to irrigation by communal farmers has a positive and significant effect on
24

households food security. Farmers who are on irrigation schemes are more likely to be food secure than dry land farmers. Thus, the hypothesis Irrigation farming in communal areas enhances households food security is accepted. Access to irrigation does not only increase crop production but also increases incomes that can be used on non-farm goods and services. The same study further found that lack of access to credit and cash crop production displace food crops and households consumption of own production is reduced. Thus the households vulnerability to food insecurity tends to increase. However another study in Malawi by Diagne .A. (1998) found that formal credit has marginally beneficial effects on households annual income. However, these effects are very small and do not cause any significant difference between the per capita incomes, food security, and nutritional status of credit program members and non-current members. Choosing timing and volume of maize harvest as an indicator (Shiferaw T.Feleke 2005) in southern Ethiopia to assess determinates of food security by using regression model: he found that based on results from tests of the full/reduced models and the magnitude of changes in conditional probabilities of food security, he conclude that the supply-side variables are more powerful than the demand-side variables. Results also show that an increase in per capita aggregate production is not necessarily translated into households food security at the households level. The study highlights the deterioration of rural households food security due to the income effect of declining prices associated with increased per capita aggregate production. The same study also future confirmed that based on the positive and significant relationship between technology adoption and households food security, adopters of improved technology are more likely to be food secure than no adopters. Since the effect of technology adoption can be explained through increased income, it can be argued that technology adoption can also have an important effect on the distribution of benefits through increased demand for nonfarm goods. Although the study shows that technology adoption improves households food security through increasing food availability or households cash incomes in the short run, we cannot infer that this positive relationship will hold in the long run or at the macro level. This is because with increased adoption of improved technologies, the proportion of marketed surplus will generally increase, as non adopters become adopters. However, given a very low overall price elasticity of demand for cereals in developing countries, the increase in marketed surplus
25

causes prices to fall further, and eventually makes the producers worse off. This was inferred based on the negative relationship between per capita aggregate production and households food security. Studies done to assess determinates of households food security(Bartfeld & Dunifon, 2006; Kidane, Alemu, Khundhlande, 2005; Shiferaw, et al. 2005; Abebaw & Ayalneh, 2007) found that demographic factors also play a influential role in households food security status.

households whose heads are educated are often more likely to be food secure; on the contrary, households with relatively more mouths to feed (i.e. with a higher dependency ratio) are more likely to face food insecurity (Bartfeld & Dunifon, 2006; Shiferaw et al. 2005; Abebaw & Ayalneh, 2000); women-headed households face gender specific obstacles, adversely affecting their ability to produce food (FAO, 2008; Bartfeld & Dunifon, 2006). And even if they are employed, they dont earn as much as their male counterparts. They also spend much of their time doing unpaid work in the households and chances are slimmer for households in affluent neighborhoods (higher income earners) to be food insecure (Bartfeld & Dunifon, 2006; Kidane, et al., 2005; Shiferaw, et al. 2005).

Amhara regional state of Ethiopia, the case of North Wollo, the data analysis was based on food balance sheet and aggregate food security index reveal that the north wello zone is highly food insecure area and the majority of the sampled households depends on famine relief assistance. In addition they tried to find the cause of food insecurity using logit model and found that cereal production, education, fertilizer consumption, livestock, land size, reduce the probability that households food insecure while, family size increase the probability of insecurity (Ramakrishna et al 2002).

Off-farm employment opportunities in rural Ethiopia are limited in both availability and incomegenerating potential. Only 44% of rural households surveyed by the Ministry of Labor in 1996 reported any non-agricultural sources of income and these contributed only for 10% to households income (Befekadu and Berhanu 2000). Another survey in Hararghe Region confirmed that off-farm activities generated only petty incomes: women collect and sell firewood and forage, men and women seek irregular, low-paid work as farm laborers, and some men
26

migrate seasonally (ICRA et al. 1996). In a survey conducted in the Amhara region, 25% of households had one or more members migrate during the dry season in search of work, mostly to nearby rural areas. One in three migrants had difficulty securing employment, while half brought back no food or income for their families (FSCO 1999).

A number of studies have been taken by (Bigsten A, Kebede B, Shimelis A and M Taddesse2007, Dercon S 2006, MoFED 2006) to examine the determinants of food security and poverty in rural areas of Ethiopia using logistic regression and found that. Socio-economic variables such as asset holding (mainly cultivated land, farm income and livestock holding) and access to services like credit are found to be important correlates which affect households food security favorably. While controlling for all other variables, households with better access to irrigation are found to have significantly higher wellbeing and so more likely to be food secure. However, among demographic variables considered in this study only households size was found to have a negative and statistically significant effect on households food security. Contrary to usual expectation, the coefficient of education level of the households head was not statistically significant. This may imply that education of households head has not yet enhanced households capabilities to adopt better production technologies, accept technical advice from extension workers and diversifying their source of income than the illiterate ones which would have reduced the risk of food insecurity among households. The results also suggest that both food secure and food insecure households have the same access to food aid resources.

2.5 Methodologies of measuring food security The next section outlines four ways of measuring households food security outcomes: individual intakes, households caloric acquisition, dietary diversity, and indices of households coping strategies. This ordering of methods is deliberate, moving from methods that are very time and skill-intensive but are regarded as being more accurate, to those that can be implemented quickly, are relatively undemanding in terms of the skills required by the implementers, but are more impressionistic. In this study households caloric acquisition will be applied.

27

2.5.1 Households Caloric Acquisition: This is the number of calories, or nutrients, available for consumption by households members over a defined period of time. The principal person responsible for preparing meals is asked how much food she prepared over a period of time. After accounting for processing, this is turned into a measure of the calories available for consumption by the households (John Hoddinott 1999).

2.5.2 Methods for generating data: A set of questions regarding food prepared for meals over a specified period of time, usually either 7 or 14 days, is asked to the person in the households most knowledgeable about this activity(John Hoddinott 1999). There is no consensus regarding the optimal recall period between 7 and 14 days. In our experience, 7 days seemed to be the most appropriate. A shorter recall period would have risked missing foods served infrequently, say on Fridays (in Muslim areas) or Sundays (in Christian areas). A longer recall period has proved problematic as difficulties of remembering what was prepared appear to increase. However, other organizations such as the World Bank in its Living Standard Measurement Surveys have used the 14-day recall period.

In constructing these questions, the following considerations should be borne in mind: it is extremely important that the list of foods specified in the questionnaire is detailed and exhaustive. Experience has shown that using short lists typically leads to an understatement of consumption on the order of 25 to 75 percent (Deaton and Grosh 1998); the phrasing of the questions needs to be unambiguous in the sense of distinguishing between the amount of food purchased and the amount prepared for consumption and the amount of food served; and it is not uncommon for individuals to report consumption in units other than kilograms or liters. In such cases, it is necessary to obtain information on the size of a "kirchat" or the quantity contained in a "Silicha" or whatever units are used locally.

2.5.3 Advantages of this method: This measure produces a crude estimate of the number of calories available for consumption in the households. It is not obvious to respondents how they could manipulate their answers. Because the questions are retrospective, rather than prospective, the possibility that individuals will change their behavior as a consequence of being observed is
28

lessened. The level of skill required by enumerators is less than that needed to obtain information on individual intakes. On average, it took around 30 minutes per households to obtain these data, an amount of time considerably less than that required to obtain information on individual intakes (John Hoddinott 1999).

2.6 Expected Contributions of the study Ethiopia is an agrarian economy with more than 80 percent of its people dependent on agriculture. The Government of Ethiopia has been trying to achieve food security at both households and national level through these small-scale farmers. Identifying determinates of households food security has been studied by many researchers in the country. This study focuses on assessing determinates of households food security in Kamba woreda in which no similar research has been done. Methodological gap and variables not included in previous studies were taken as the gap that this research going to fill. Various factors contributing to households food insecurity will be discussed so that recommendations can be made for better strategies and actions to assist small-scale producer farmers to address households food insecurity challenge in the study area.

2.7 Comments on the Reviewed Literatures Much of the reviewed literature on household food security concentrated on describing qualitatively and quantitatively the extent of household food security and identifying the factors and examining their implications. Almost all reviewed studies applied logistic regression in modeling relationships between variables. However, the central task of regression analysis: the parameter estimation techniques and variable selection methods were not addressed. Most of the reviewed model did not check model adequacy: detection and treatment of outliers, influence diagnostics and multicollinearity. Almost all reviewed studies did not examine the sustainability issues in detail. Hence, in this study in addition to the prediction, a sustainability issues will be used taking into account the limitations described in the reviewed literature.

29

CHAPTER THREE RESEARCH METHODOLGY


3.1 Description of the Study Area The study area, Kamba woreda, is found in Gamo Gofa zone, which is located in the Southern part of SNNPR. Wolayta and South Omo administrative zones border the zone in the north and south, respectively. Kamba is one of the 15 administrative woreda of Gamo Gofa zone. It is bordered by Bonke woreda in the East, Zala woredas in the West, Daramalo woreda in the North and South Omo zone in the South.

According to annual report of the Office of Agriculture (2002), Kamba has an estimated total land area of 118,054 hectares of which only 50,787 hectares are arable. From arable land 20,345 hectares are used for annual crops 2,722 hectares are used for perennial crops and 27,720 hectares of arable land is the potential for the future and 22,364 are forest covered 27,664 are pasture land, and 5,300 are unclassified. The woreda has nine big rivers of which three of them have been used as source of modern irrigation.

Kamba, the capital of the woreda, is located 105 Kms from the zonal capital, Arbaminch and 635 kms away from Addis Ababa. The agro-ecological classification of this woreda consist of low lands(kola), middle altitude(weynadega) and high lands(Dega).The woreda has 27% mid-

altitude,28% highland and 45% lowland agro-ecological zone. The top of the escarpment is approximately 3340 m.a.s.l and the foothills and ridges raise to700 m.a.s.l

3.2 Livelihood Strategies in the study area Mixed farming, both rain fed and irrigation based, agriculture is the primary source of livelihood with mainly maize, wheat, sweet potato and enset grown as staple food crops, vegetables predominantly onion and Shiferaw, coffee and corerrima are some of perennial cash crops. However, even though all these crops are grown in the area, the livelihood of the farm households heavily depend on the success and failure of maize production. The other important livelihood activity, which plays an indispensable role in the mixed farming operation, is
30

livestock production. Of the different livestock species in the production system holders pay greater emphasis to the large and small ruminants, cow, ox, sheep and goat, production. Because, their capacity generate income in a shorter period is very high. In other words, a household who has large number of livestock, especially chattels, deserves greater respect and influential power in the locality. 3.3 Population Based on the 2007 population and housing census results the Gamo Gofa zone has a total population of 1,595,570 (CSA, 2007). From this, 1,436,601 (90%) of the population lives in the rural areas depending on agriculture, while the remaining 106,234 (10%) lives in the urban areas. The population distribution of Gamo Gofa zone is the combined effect of temperature, rainfall, and soil fertility. The central highlands of the zone that have enough rainfall and other suitable natural conditions for cultivations support large number of population. On the other hand, the surrounding lowland, which is characterized by insufficient rainfall, high temperature, and higher incidence of diseases such as malaria has very small number of population or sparsely populated (2007). Hence, population distribution in Gamo Gofa is uneven. According to the same source, the average population density of the zone is 85 persons per Km2. Children less than 15 years accounts for 46.9%, the economically active population aged 15 to 64 accounts for 50.9 % and those aged 65 and above constitute 2.18%. Thus about half of the population of the zone is under the category of the independent age group. However, the dependency ratio of the zone in terms of percentage is 97, which is quite high.

The total population of the study woreda is 155,748 of which 49.2% are women and 50.8% are men (CSA, 2007). According the same source, children whose age is 14 or less account for 46.1%, this is similar to the situation in the zone average. Economically active population ages between 15 and 64 accounts for 51.6% and those aged 65 and above are 2.3 %, which is also similar to the zone average. According to Gamo Gofa Planning and Economic and Development Office (2001), the average households size of the zone was 6.34, which is more than the national average of 5. And, the average households size of the study area is 6.65, which is more than the zonal average. The average population growth rate in the study area is also above the

31

country average growth rate (2.8% CSA, 2006). Table 1 shows the woreda population growth rate in the past ten years as compared to country growth rate. Table 1: - Population growth rate of Kamba woreda Growth rate Year 2001 2002 2003 2004 2005 Kamba 3.4 3.8 3.0 3.2 3.2 Country 2.65 2.62 2.60 2.56 2.56 Year 2006 2007 2008 2009 2010 Growth rate Kamba 3.0 5.2 3.2 3.4 2.9 Country 2.60 2.59 2.59 2.58 2.45

Source: - World Bank2008 report and Kamba Woreda council 3.4 Religion Religion in the woreda is classified and presented in Table 2. As it is indicated in the Table, majority of the population are Protestants (about 58%). Next to Protestant Orthodox Christian accounts about 24.5 percent and households with no religion 16%. There is no conflict among different religion followers. Therefore, food security problem due to religion conflict is not observable. Food source and feeding habit of all religion is basically similar Table 2:- Religion distribution of the study area Religion Orthodox Protestant Muslim No religion Source:-Woreda council 2010/11 3.5 Agriculture As elsewhere in the country, agriculture is the major occupation of people living in the study woreda. Except for few, the livelihood of the population (residents of both rural and urban areas) in the woreda depends directly or indirectly on agriculture. More specifically, agricultural production (livestock and/or crop production) is the main source of income and employment to the society, though the degree of importance varies from one Keble to another.
32

Percentage 24.5% 58% 1.5% 16%

The farming system of the study woreda is governed by the agro-ecology of the area, which consists of highlands, midlands, and lowlands. In the highlands, mixed farming is practiced which cultivate cereal as their annual major crop and enset as their perennial crop and animal rearing as their minor on-farm activities. In the midlands, mixed farming is also practiced but their main farm activities are cash crop and animal rearing. In the lowlands, mixed farming is also common farming practice but focusing on maize production and animal rearing as major activities. Pastorals depend strongly on animal rearing and to some extent on sorghum farming. 3.6 Crop Production The major crops produced in the Gamo Gofa zone in general and Kamba woreda in particular are cereals, roots and tubers, vegetables, and fruits and spices. Maize, wheat, barley, and teff are the dominant cereal crops and sweet potato, Irish potato and enset are also the dominant roots and tubers produced in the woreda. Coffee and coriander (corerrema) are also mostly produced spices. The distribution of these major crops varies from place to place depending on crop suitability factors (soil, rainfall, temperature, etc). For example barley, wheat and Irish potato are cool weather crops, and are predominantly grown in the highlands above 1500 m.a.s.l where the average annual rainfall ranges from 1000 mm to 2000 mm. Maize and sorghum are warm weather crops they grow under rainfall of 800 mm to 1500 mm in midlands and lowlands. Pulses such as haricot bean and field peas are widely grown in the study woreda and usually intercropped with maize, sorghum and perennial crops like enset and coffee.

3.7 Livestock and Poultry Production According to Kamba woreda Agricultural Development Department (2002), the Wordes total population of livestock and poultry is estimated to be 338,011. Among this, cattle population accounts for 107,338(32%), followed by poultry 94,686(28%), sheep 78,368 (23%) and goat 45,796(13%). The proportion of the horses, donkeys, and mules are 2.5, 0.5 and 1percent respectively. Here, cattle are the major source of farm power for plowing, threshing as well as for manure supply. The highest concentration of goats is found in the lowlands of the woreda where pastoralists live. Largest number of sheep and equines are mainly found in the highlands because of disease problem in the lowlands. According to zonal agricultural department annual report (2002) the study area is the main sources of meat livestocks, especially for oxen and got.
33

3.8 Agricultural Extension Agricultural extension is one of the most important inputs, which encourages farmers to use improved techniques and creates an enabling environment for agricultural input and output markets in the case of Gamo Gofa in general and Kamba woreda in particular. Agricultural extension focuses on use of improved hybrid seeds, fertilizer, animal husbandry, and soil and water conservation to increase farmers productivity. The agricultural extension serves in the woreda also focus mainly on crops, livestock and natural resource development activities.

There are118 Development Agents (DAs) serving 23,420 farming households in Kamba woreda (WoA, 2002). The development agents have diploma level education and recruited from the community and live within the community to provide extension services. The ratio of households to DA in the study area is 1:198. This ratio is more than double compared to the national ratio of 600. This is one of the issues to be considered.

3.9 Input Supply Key agricultural inputs used by farmers in Kamba woreda are improved seeds, chemical fertilizer, improved chicken and protection chemicals. Improved seeds include maize and wheat. Chemical fertilizers commonly used are DAP and urea. Plant protection chemicals are pesticides and herbicides. The ten years average of fertilizer use in the woreda is 825 quintals DAP per annum and 19 quintals of urea (WoA 2002); the quantity used per households is almost 3.5kg of DAP and very small amount of urea per households which is too small to bring about a change in production of the woreda in general (WoA 2002).

3.10 Rural road Road network is not well developed and the majority of the existing roads are not in a position to function in wet seasons. Due to this problem the movement throughout the woreda is restricted and the development effort is hampered. Until the year 2003 there were a total of 48km of roads in the woreda out of which only 20 km were RR-10 standard roads all weather rode. According to Kamba woreda council road network problem has been number one problem in the study area for the problem of food security. Food transportation is very costly and farmers and consumers in the study area pay abnormal price during normal production seasons.
34

3.11 Market There are three dominant and high monetary exchange market places located in open rural villages. These markets are traditional in nature and are characterized by inadequate marketing facilities and services. Particularly, feeder roads and roads linking rural areas with urban consumption centers are inadequate. Thus, the majority of the areas are inaccessible by vehicles making it imperative to use pack animals (such as donkeys).

3.12 Unexploited opportunities that exist in the woreda Agriculture and livestock development are almost certainly the only viable alternatives for improving the food security status of kamba poor majority people. Kamba district has three constructed irrigation scheme covering 470 hectares and there are more than four big rivers feasible for irrigation scheme with potential covering about1,200 hectares. The woreda has one of the largest livestock populations in the zone and Kamba woreda is the known supplier of meat livestocks. According to Kamba woreda tax office report in each year averagely 8,000 oxen and 15,000 goats have been transported from the woreda for export or local consumption. The woreda has all the three agro-ecologies and known coffee producer next to Mallo woreda in the Gamo Gofa zone. The woreda agricultural office report shows that about 13,456 hectares are suitable for coffee and Correrima production. In the highland farmers are producing highland fruit, especially apple. The woreda has 27,720 hectares arable land which is suitable for agriculture and still not owned by farmers.

3.2 Methods and Sources of data collection


3.2.1 Methods of data collection

A detailed households survey was conducted to collect both primary and secondary data on the dimensions of food security and livelihood strategies in Kamba woreda. The woreda has three agro-ecological zones Dega (14PAs), Woynadega (10PAs), and Kola (14PAs).The data collection
was done efficiently through a structured questionnaire to collect the required data from selected sample households during the 2010/1011 production year. The questionnaire was first pre-tested and modified before the execution of the survey (See the questionnaire in Appendix 4). For this study Development

agents (DAs) wear used to collect primary data after two days training on the questionnaires, selection was based on their work performance, their ability to local language and their interest
35

to collect the data in the selected PAs. To ease and keep uniformity among enumerators, standardizing unit on local measurement of different food items and codes for some questions were provided to each enumerator. The enumerators who are stationed in the survey areas
administered the structured questionnaires under the continuous supervision of the researcher. The survey was carried out on March 2011. The primary data were supplemented by secondary data whenever necessary from governmental organizations.

Sample Size Determination:-There are several approaches to determine the sample size. These include using a census for small populations, imitating a sample size of similar studies, using published tables, and applying formulas to calculate a sample size. This study applied a simplified formula provided by Yamane, (1967) to determine the required sample size at 95% confidence level, degree of variability = 0.5 and level of precision =9%.

n=

------------------------------------------------- (1)

Where n is the sample size, N is the population size (total households size), and e is the level of precision. The above formula required a minimum of 123 responses but this study was carried out on 200 respondents.

Sampling Technique:-For this particular study a two-stage random sampling procedure was implemented to select 200 rural households in the woreda. To increase the precision of estimates

households were stratified by agro-ecological zone of the peasant associations (PAs) they live in. At the first stage, 13 peasant associations (PAs) were selected randomly out of 38 PAs which constitute about 34% of the total PAs in the study area. Thirteen PAs wear selected by applying size to agro-ecology proportionality across the three zones. This was done by selecting 34% of the total PAs in each zone, by doing this from both Kola and Dega zones 5 PAs were selected randomly and from Woynadega 3 PAs were selected also randomly. In the second stage, considering the PAs householdssize, random selection of households equal to probability proportional to households size sampling technique was employed to draw sample households. From randomly selected PAs maximum 34 households from larger kebeles and 10 households from smaller kebeles were selected and in each kebele two reserve households selected

randomly, if the households is not present in the kebele at the time of interview.
36

At the second stage of sampling, sample frame (tax payment list for land holding) was applied to pick out targeted households using lottery system. Responses from 200 households were however found useful for this study. Secondary data was mainly collected from woreda government offices.

The questionnaire helped to draw out information related to the different aspects of the households and individual characteristics including: households food habit and consumption, households demographic and socio-economic characteristics, consumption pattern, households income and expenditure, labor input in agriculture and non-agricultural activities, crop and livestock production, farming systems and productive resources, land use, access to services, access to infrastructure, amount of farm and off-farm income as well as coping strategies employed by the households during time of food shortage. Before analysis of households 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 19681997) and households members to their adult equivalent(see appendix 3).

3.2.2 Method of data analysis Food security at households 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).

A set of questions regarding food prepared for meals over a specified period of time, for this study for the last seven days, excluding interview day meals was asked to the person in the households most knowledgeable about this activity and other socio-economic and demographic datas wear also used to analyze data. The response on food consumption was changed to international unit (if local unit was used) and converted to calorie availability and compared with calorie demand or minimum calorie requirement. The government of Ethiopia has set the minimum acceptable weighted average food requirement per adult equivalent (AE) per day at 2100 kcal. The determination of the adult equivalent takes into account the age and sex of each
37

households member (Gassmann F and C Behrendt 2006) (see appendix 3). Hence, for this study 2100 kcal per adult equivalent per day was employed as a cut-off value between food-secure and food-insecure households.

Once the groups wear categorized as food-secure and food-insecure, the next step was to identify the socio-economic factors that are correlated with food security. A variety of statistical models can be used to establish the relationship between these households characteristics and food security. Conventionally, linear regression analysis is widely used in most economic and social investigation because of availability of simple computer packages, as well as ease of interpreting the results. However, results derived from linear regression analysis may lead to fairly unreasonable estimates when the dependent variable is dichotomous (Ayalneh Bogal and A. Shimelis volum9 No9 2009).Therefore, the use of the logit or probit models is recommended as a universal remedy of the drawback of the linear regression model (Gujarati DN Basic Econometrics, Fourth edition. McGraw- Hill, New York 2003).Which model to choose between logit and probit is, however, difficult for they are similar in most applications, the only difference being that the logistic distribution has slightly fatter tails. This means that there is no binding reason to choose one over the other but for its comparative mathematical and interpretational simplicity many researchers tend to choose the logit model (Hosmer DW and S Lemeshew Applied Logistic Regression New York. 1989). Therefore, this study applies the logistic regression model due to dichotomous behavior of dependent variable.STATA10 package was also used to see the dominant factors that significant affect food security in the study area. 3.3 Theoretical model Gary Beckers idea about family economy, households production function and family time allocation is the inspiration of agricultural households models (Becker, 1965). Agricultural households is a worker, a producer and a consumer all together, with his ultimate aim of maximizing utility. The models can also be used to address consumption and related nutrition policy issues (Strauss, 1984). The existence of imperfect market is common in the LDCs; one common cause of market imperfection is the existence of transaction costs. For the same food product, households typically sells it for a lower price than the price at which he can purchase in
38

the market. Thus, it is the budget constraint for the agricultural households. With this budget constraint in mind, we have a clear picture of utility maximization for the agriculture households. In Ethiopia, a households food security model was used to look at the importance of the supply side against demand side variables in determining households food security in Southern Ethiopia (Feleke et al., 2005). From the results, it was established that the supply side variables are more powerful determinants of households food security than the demand side variables. The extent of households food security in this study will be modeled within the framework of consumer demand and production theories following the modeling of production and consumption behaviors of a rural and Agricultural households models by Singhand Shiferaw(1986) and T. Feleke (2005). Following Agricultural households models 1986 by Singhand Shiferaw and T. Feleke 2005, the households utility function is specified as:

U = U (Fi, Fm, L; Dh) -------------------------------------------- (2) s.t G (Qi , L, A0, K0) = O PmFm = Pi (Qi - Fi ) - w(L-Lf) + N=0 T=Lf +
Where U is a utility function that is assumed to be well behaved (twice differentiable, increasing in its arguments, and strictly quasi-concave); Fi is a vector of home-produced goods and consumed by the households; Fm is a vector of market-purchased goods consumed by the households; and L is leisure. The utility that the households derives from various combinations and levels depends on the preferences of its members, which are shaped by the characteristics of the households, Dh.

39

As indicated in the Agricultural households Models 1986 by Singhand, Shiferaw and T. Feleke 2005, the households, as both producer (firm) and consumer, is assumed to maximize its utility from the consumption of these goods subject to farm production, income, and time constraints specified as where G() is an implicit production function that is assumed to be well behaved (twice differentiable, increasing in outputs, decreasing in inputs, and strictly convex); Qi is a vector of quantities of goods produced on-farm; L is total labor input to the farm; A0 is the households fixed quantity of land; K0 is the fixed stock of capital; Pi is the price of good i; Pm is the price of a market-purchased good; (QiFi) is the marketed surplus of good i; w is the wage rate; Lf is the households labor supply for on-farm use; N is nonfarm income that adjusts to ensure that Eq.(3 ) equals zero; and T is total time available to the households to allocate between work and leisure. The income and time constraints can be combined by incorporating the constraints equations.

PmFm = Pi (Qi Fi) w (L T + ) + N


Rearranging Eq. (3) gives

(3)

PmFm + PiFi + w = PiQi + wT wL + N (4)


The left-hand side of Eq. (6) is the households expenditure on food and leisure, and the righthand side is the full income equation. The expenditure side includes purchases of its own farmproduced good i (PiFi), the households purchases of the market good (PmFm), and the households purchases of its own leisure time (w). The full income side consists of the value of total agricultural production PiQi, the value of the households entitlement of time value of labor on the farm, including hired labor wL, and nonfarm income N. From the first-order conditions of the maximization of the constrained utility function, the relationship between production and consumption can be established in such a way that production decisions are made first and subsequently used in allocating the full income between
40

wT, the

consumption of goods and leisure (Strauss, J., 1983).This is based on the assumption that all the relevant markets function. It is important to have this assumption because we are considering that consumption (food security) depends on the production variables, but not vice versa. If the markets for inputs, product, and labor do not function, farm production decisions cannot be made separately from the consumption decisions. When a commodity has an incomplete market, or if a households is at a corner (i.e., if it consumes all of its output), there will exist a virtual (or shadow) price, which will be endogenous to the households (Singh et al., 1986). Given the assumption of separability between the production and consumption decisions, Pi and Pm are exogenous, the utility function to be well-behaved. We can mathematically derive the production-side and consumption-side equations separately. Starting with the production side, the first-order conditions can be solved for input demand (L*) and output supply (Q*) in terms of all prices, the wage rate, fixed land, and capital as

L* = L* (Pi, w, A0, K0), --------------------------------------------------- (5)


and

Q* = Q* (Pi, w, A0, K 0), ------------------------------------------------- (6)


These solutions involve the decision rules for the quantities of labor input used and outputs produced (production side). Once the optimum level of labor is chosen, the value of full income when profits have been maximized can be obtained by substituting L* and Q* into the right-hand side of the income constraint (Eq. 6) as

Y* = PiQi* + wT wL* + N ------------------------------------------- (7)


and

Y* = wT + *(Pi, w, A0, K0) + N ---------------------------------- (8)


41

Where Y* is full income under the assumption of maximized profit *. The first-order conditions can be solved for consumption demand in terms of prices, the wage rate, and income as

Fk = Fk (Pi, Pm, w, Y*), ------------------------------------------------- (9)


Where k=i, m. These solutions involve the decision for the quantities of goods and leisure consumed (consumption demand side). Equations (5), (6) and (9) give us a complete picture of the economic behavior of the farm households. They are combined through the profit effect, because income is determined by the households s' production activities, implying that changes in variables influencing production also changes income, which in turn affects consumption behavior. Incorporating the households characteristics that shape its preferences (Dh), the demand for food indicated in Eq. (9) can be rewritten as

Fk = Fk [ Pi , Pm ,w , Y* (w , A0 ,K0 ,N ) ,Dh ]
Where k=i, m. 3.4- Empirical model

---------------------------(10)

Determining the demand for both home-produced and market-purchased goods, we can now calculate the amount of calories (Ci) available in the respective food items. Then, the extent of households food security is determined by the difference between caloric availabilities and needs. Defining C*i=Cii, where Ci is caloric availabilities and i is the consumption needs for the ith households, C*i 2100 kcal indicates that the households is food secure while C*i < 2100 kcal indicates that the households is food insecure, implying that the same independent variables identified in Eq.(10) can be used in a food security model. Assuming a linear function, we can write the food security equation as

42

Pi= F(Zi) =F( + iXi) =

-------------------------------- (11)

Where e is the base of the natural logarithm, i represent the ith explanatory variables Pi is the probability that an individual is being food insecure given Xi, and i are regression parameters to be estimated. The dependent variable (food security) is measured using a proxy that the households observed to be food secure (Zi= 1) is assumed to have C*i2100 kcal; while the households observed to be food insecure (Zi= 0) is assumed to have C*i <2100 kcal. Since the dependent variable Zi is a discrete variable, the food security model can thus be cast as a qualitative response model where Pi is the probability of food security, which can be written as

Zi = ln(

) = + 1X1 + 2X2 + -------mXm-------------------------(12)

ln(

) = 0 + jXij+ i

here Pi is the conditional probability of food security; j's are parameters to be estimated; (X1)Family size, (X2) farm size, (X3) livestock owned(TLU), (X4) total off-farm income (X5) education of households head,( X6) amount of food aid received,( X7) technological adoption,(X8) access to infrastructure,( X9) participation in public meeting,( X10) extension service adoption, (X11) land quality (land quality measured by farmers perception of the fertility of their farmland),( X12) saving,( X13) number of month food purchased, (X14) age of households head, (X15) sex of households head,( X16) access to irrigation are identified potential variables that determines households food security problems in Kamba woreda.

3.5 Variables and working hypothesis The number of variables that would be included in the model should deliver optimum possible information. In this study the selection of the variable for the final logistic model was made by
43

looking the association between each predictor variables with the response variable. A minimum of 10 events per independent variable has been recommended for sample size and dependent variable efficiency (Hosemer-Lemeshow, 1989). The separate effect of each predictor variable in explain the outcome variable was made by postulating the null hypothesis that H0: i=0 against the alternative H1: i 0 for at least on i=1, 2, 3----n. The significance test for each coefficient in the model was done using the Wald-chi-square ([/s.e()]2 which is distributed as a chi-square with degree of freedom) and likelihood ratio test. A likelihood ratio (LR) chi-square test was also employed to examine the importance of each predictor variables to the outcome variable. Therefore, the following sixteen independent variables wear selected to analyze the hypothesis whether they explain a households food security status or not. Review of literature, working as an expert in the agricultural office in the study area, degree of attention given by the government policy, existence of chronic food security problem, unpublished local NGOs and local government reports wear used as important sources to identify the potential determinants of households food security in the study area. Some variables like access to irrigation, oxen ownership, age of households head and sex of households head are more used in descriptive rather than in logit model. 3.6 The dependent variable of the model (HFST): The households food security status is a discrete variable representing the status of households food security. It was represented in the models by two possible alternative ways: 1 for food secure and 0 for food insecure households. The information, which identifies the food secure from the food insecure, is obtained by comparing the total food calorie available for consumption in the households per AE to the minimum level of subsistence requirement per AE (2100 kcal). threshold is said to be food secured, otherwise not. Households beyond this

3.7 Independent variables: households socio-economic characteristics such as households size, farm size, livestock ownership, total off-farm income, educational status of households head, food aid received, technology adoption, access to infrastructure, sex of households head, age of households head, participation in public meeting, application of extension service,

44

saving, land quality and number of months food purchased are selected potential variables for the logistic model and descriptive analysis.

Family size (FASZ):- Family size is measured by the number of family members in the households. As family size increases, obviously the number of mouths to feed from the available food increases. Hence, it is hypothesized that family size and food security are negatively related. The existence of large number of family members with limited resources could affect the food security status of the households. This is due to increasing demand for food with limited food supply. Evidence in the literature indicates that larger family size have negative impact on food security (Mulugeta Tefera, 2002; Abebaw Shimeles, 2003 and Ayalew Yimer, 2003). Farm size (FRMS): Farm size is the total farmland owned by the households measured in hectares. Losses of farm land to other uses because of population pressure and limits to the amount of suitable new land that can be brought in to production is one of the constraints of food production (Brown et al., 1990). Fertile farmland is often sacrificed to meet the growing demands of population growth (Ehrlich et al., 1993). As the cultivated land size increases, provided other associated production factors remain normal, the likelihood that the holder gets more output is high. This variable represents the total cultivated land size of a households in hectare. It was hypothesized that farmers who have larger cultivated land are more likely to be food secure than those with smaller area. Livestock holding (TLU): livestock holding refers to the total number of livestock holding of the farmer measured in tropical livestock units (TLU). Livestock production constitutes a very important component of agricultural economy, a contribution that goes beyond direct food production to include multipurpose uses such as skins, fertilizer and fuel, as well as capital accumulation. Furthermore, livestock are closely linked to the social and cultural life of farmers for whom animal ownership ensures varying degrees of sustainable farming and economic stability (Sansoucy, 1995). Therefore, it was expected that a higher possession of livestock increase the probability to be food secure.

45

Access to infrastructure (INFRA): Physical access to the infrastructure is measured by the amount of
Km required to reach the nearest local infrastructure. Nearness to infrastructure creates access to

additional income by providing access to increase productivity, easy access to inputs, transportation and off-farm employment opportunities. The shortest distance to the infrastructure was used in this study to determine how often the farmers used the infrastructure to sell and purchase their produce. It was, therefore, expected that households nearer to infrastructure have better chance to improve households food security status than who do not have proximity to infrastructure centers. Proximity to infrastructure centers was measured in kilometer. Mulugeta Tefera (2002) reported that market distance has no significant effect on food security. In this study market distance was the main focus relative to other infrastructures. Age of the households head: Age matters in any occupation. Rural households mostly devote their lifetime or base their livelihoods on agriculture. It was argued as the age of the households head increases the farmer acquires more knowledge and experiences with possible negative impact on food insecurity. In other ways, it was expected that younger farmers are more likely to be food insecure than the older farmers that the older ones due to better possession of resources accumulation. In light of this, it is hypothesized that ages of the households heads and food security are positively correlated.

Irrigation: In areas like in Kamba where agriculture is the prime source of livelihood of the society, soil moisture is very crucial. Even if the climatic condition in a given area is conducive, then it would be far better to be supplemented with irrigation so that increased output could be attained. However, in the study area drought, erratic rainfall patterns and other factors limit the output per hectare, and made it one of the food insecure districts in the region. Hence, it was hypothesized that the use of irrigation and food security are positively related

Sex of households head (SEX): women farmer may need a long adjustment period to diversify their income sources fully and become food secure (Christina et al., 2001). Labor factor plays a great roll in the study area. With regard to farming experience and access to technology, males are better than female farmers. So sex of the households head is an important determinant of

46

food security. Therefore, it was hypothesized that male-headed households are more likely to be food secure. Off-farm income (TOFFI): When crop production and income earned from sales of livestock and livestock products become inadequate to subsist the farming households of the study area they often depend on external or other source of income to purchase food and farm inputs. So income earned from off farm activities is an important variable, which determines households food security in the study area. Having multiple livelihood strategies increase chance to be food secure for rural people like in Kamba. Hence, it was expected that the availability of off-farm income is positively associated with households food security status.

Education (EDUC): Education equips individuals with the necessary knowledge of how to make a living. Literate individuals are eager to get information and use it. Hence, it is supposed that households who have had at least primary education or informal education are the ones to be more likely to benefit from agricultural technologies and thus become food secure.

Total food aid (FAID): food aid is given as a copping strategy to food insecurity in the study area. Food aid can develop a dependency behavior among households which in turn will reduce farmers motivation towards food self sufficiency. Therefore, food aid was expected to have a negative relation to food security. The amount of food aid given was measured in kg. According to Mulugeta Tefera (2002); Abebaw Shimeles (2003) and Ayalew Yimer (2003) food aid has no significant effect on food security. Technological adoption (TEC): Technological adoption determines the households food security, those households who used improved seed, chemical fertilizer, apply proper agricultural practices increase productivity and increase chance to become food secure.

Participation in public meeting (PAPUM):- households who participate in public meeting have enough information about current input and output price, in put availability, credit access, technology adoption and have higher chance to increase production than non-participants. Therefore, public meeting was expected to have a positive relation to food security.
47

Extension service (EXTN):-It refers to the number of days per year a farmer was visited by a development agent for technical guidance has influence on both productivity and food security. The higher the linkage between farmers and development agents, the more the information flows and the technological (knowledge) transfer from the latter to the former. Those farmers who have frequent contacts with development agents are likely to produce better than others and would be in a better position of food security status.

Saving (SAVG) - Saving in the study area is in the form of either grain storing or purchasing animals. Hence, saving in the form of animal purchase and grain storing has several advantages in that source of income from the sale of their product; income from the sale of animal fatting and above all it serves as a hedge at time of crop failure (drought). Generally, it is assumed that households who have saving experience would have better food status than less saved.

Land quality (LNDQ):- Land quality measures farmers perception of the fertility of their farmland. Households were asked to indicate whether they consider their land as fertile and not fertile. Under optimal management, better land quality boosts crop production (Sah, 2002). Stephen (2000) found that a decline in soil fertility negatively affects food security. It is expected that this study was expected land quality affects food security status of households positively. This is because the increase in the fertility of the land is expected to contribute positively towards increase in crop output and consequently increase in farm income and food security.

Number of months food purchased (NMFP):- As the number of months food purchased increase the probability to be food secured decreased. Subsistence farmers like in Kamba produce food for home consumption. Farmers buy stable foods when their production does not meet food demand. Numbers of months that farmers buy food items indicate the level of food security. It has inverse relationship with food security statues of households.

48

Table 3:- Types, codes and definition of variables to be used logit model Variable FASZ AIR FRMS TLU TOFFI EDUC FAID TEC Variable type Continuous Continuous Continuous Continuous Continuous Continuous Continuous Dummy Family size in number Access to irrigation in hectare Cultivated land size in hectare Total livestock holding in TLU Total off farm income earned in birr(2010/11) Total grade of education Food aid obtained in Kg (2010/11) 1, if the households head used inputs 0, if the households head does not used inputs. Access to infrastructure in Km 1, if the households head participate in public meeting, 0 if the households head does not participate in public meeting. 1, if the households head got extension service, 0 if the households head does not got extension service. Amount of birr saved (2010/11) Number of months food item purchased Variable definition

INFRA PAPUM

Continuous Dummy

EXTNS

Dummy

SAVG NUMFP

Continuous Continuous

3.8 Interpretation of the coefficients of the logistic regression model The estimated coefficients for the predictor variables in logistic regression represent the slop or rate of change of a function of the outcome variable per unit of change in the predictor variables (Hosemer-Lemeshow, 1989).Thus, interpretation involves two issues:- i) determining the functional relationship between the outcome variable and the predictor variable. ii) Appropriately defining the unit of change for the predictor variable (Hosemer-Lemeshow, 1989).

49

The estimated logistic coefficients i's reflecting linear and non- linear relationships and they are interpreted as the change in the- log-odds for every unit increase/decrease (depending on the variable change in Xi) holding other variables constant. Odds, odds-ratio and the marginal effects are all important basic terms in logistic regression parameter estimates ( coefficients) of explanatory variables used in logistic regression equation to estimate the log-odds that dependant equals 1(binomial logistic regression).Odds ratio is the ratio of the probability that something is true or happen divided by the probability that something will not happen. Odds ratio above one refers to the odds that the dependent variable equals one in binary logistic regression. The closer the odds ratio to unit, the more the predictor variables categories are independent of the outcome variable, with 1 representing full statistically independency. 3.9 Marginal effects result in logistic model Marginal effect, which is one of the results of logistic regression, shows us what the effect of a change a given predictor on the output response variable would be at the sampled mean values for the continuous variables and when the value of dummy variables changes from zero to one. The marginal effect, in binary regression model, is the slop of probability curve relating Xi to Pr (Yi=1/Xi), holding all other variables constant. It measures the change in probability of occurrence for a unit change in Xi's at their mean value. The positive sign of marginal effect indicates that the probability of households to be food secure will be increased at the mean value of continuous predictors while the negative sign indicates that the probability of householdsto be food secure will decline at their respective mean. The marginal effect of the discrete variable indicates that the likelihood of households probability to be food secure will increase as the value changes from zero to one 3.10 Testing Multicollinearity Multicollinearity in logistic regression is a result of strong correlations between independent variable. Maddala (1989) described that high inter-correlation among the predictor variables by themselves need not necessarily cause any problems in conclusion. Whether or not this is a problem will depend on the magnitude of the error variance and the variance of the predictor
50

variables. Multicollinearity may be induced due to poor sampling method, miss measurement and over fitting of a model as well as improper use of dummy variables. There are a lot of statistically accepted thumb rules that have been proposed for detecting multicollinearity among categorical predictor variables. For this study variance inflation factor (VIF) was used to detect the existence of multicollinearity. The larger the value of VIF, the more 'troublesome' or collinear the variable Xi . As a rule of thumb, if the VIF of a variable is greater than ten then the variable is said to be highly collinear (Gujarati 4th edition). 3.11 Measuring intensity of Food security (FGT) With the increased awareness and availability of data, various measures of poverty have been developed overtime, among which the Foster, Greer And Thorbecke (FGT) class of poverty index is the most commonly applied (Ayalneh 2002). This index was initially suggested by Foster, Greer and Thorebecke (1984) and has several desirable properties that have enhanced recently by IFPRI for the purpose of food-insecurity analysis (Hoddinott, 2002). Therefore, this study employs the FGT index in order to estimate the food security gap and its severity among the rural households. The FGT index can be specified as follows:-

F () =

----------------------------------------

(13)

Where n is the number of sample households s; yi is the measure of per adult equivalent food calorie intake of the ith households ; m represents the cutoff between food security and

insecurity (expressed here in terms of caloric requirements); q is the number of food-insecure households s; and is the weight attached to the severity of food insecurity. In the above equation if m-yi = 0 if yi > m. Hoddinott (2001) further explained that giving no weight to the severity of food-insecurity is equivalent to assuming that = 0. So then, the formula collapses to F (0) = , this is called the head count ratio.

51

CHAPTER FOUR RESULTS AND DISCUSSION


This chapter presents the study findings in two categories as a descriptive and econometric model analysis of the survey data. Descriptive statistics such as mean, standard deviation, percentage, maximum and minimum were employed and binary logistic, econometric model was used to identify determinants of food security at households level in the study area. Dimensions of households food security, in terms of extent and severity, were computed by using an FGT (Foster, Greere and Thorbecke) index. 4.1 Measuring food security status of households Based on the methodology described in the previous chapter the following alternative results were found: The information to categorize households s into food secure and insecure groups was obtained by comparing the total households food or calorie acquisition per AE per day to the minimum level of consumption required to ensure survival per AE per day. Thus those households who have energy per AE beyond the minimum subsistence requirement (2100kcal) are deemed to be food secure, otherwise food insecure.

Considering 2100kcal as a benchmark, only 79 sample households(39.5%) were found to be able to meet the minimum subsistence requirement and 121 households (60.5%) were found to be unable to meet their minimum subsistence requirement. Based on the Core Food Security Module analysis, households were found to be grouped into four categories; food secure (above 2100Kcal) (39.5 %), mildly food insecure (1800Kcal-2100Kcal) (34.5 %), moderately food insecure (1500Kcal-1800Kcal) (16.5 %) and severely food insecure (less than 1500Kcal) (9.5 %) (Table3). Majority of the respondents were mildly food insecure followed by moderately food insecure. This empirical assessment results can serve as an instrument to guide decision as to which status should be given more emphasis to reduce the problem.

52

Table 4:-Food security status of sample households Level of food insecurity Food secure Mildly food insecure Moderately food insecure Severely food insecure Total Source: - Survey data (2011) 4.2 Demographic and socio-economic characteristics Socio - economic characteristics of sample households by family size, sex and education level are summarized in relation to the food security status at households level. Possible explanations on factors believed to have contribution to households food security are also presented from analysis of descriptive output. Family size Family size was considered and hypothesized as one of the potential variables that would have due contribution for food security. The proportion of sample households becoming food secure decreased as the family size increases. About 46 percent of the 36 food secure and 23 percent of the 28 food insecure sample households were found to have family size less than or equal to 5. The number of food secure households with family size six and above is less than insecure households which constituted 54 percent and 77 percent of the food secure and food insecure households, respectively. Amount of calories Above 2100Kcal 1800Kcal-2100Kcal 1500Kcal-1800Kcal Less than 1500Kcal Amount of HH in that interval 79 69 33 19 200 Percentage 39.5% 34.5% 16.5% 9.5% 100%

The survey result also revealed that there was significant difference in the mean family size at less than 1 percent probability level (p<0.01) between food secure and food insecure sample households. In that, the mean was found to be 4.32 and 7.04 for food secure and food insecure households respectively. While the overall mean family size of the sample households was 5.97. This was above the national average of 5 persons (CSA, 1994). This result is in agreement
53

with the prior expectation. The largest family size of the sample householdswas 12 and the smallest was 2 (Table 4). Table 5:-Distribution of sample households by family size in number Family size Food insecure N=121 5 6 Total Mean SD Maximum Minimum t-value 12.72 Source: - Survey data (2011) 28 93 121 7.04 0.17 12 2 Percent 23% 77% 100% Food secure N=79 36 43 79 percent 46% 54% 100% 4.32 0.13 9 2 Pr(T > t) = 0.0000*** *** Significant at less than 1% probability level N=200 64 136 200 5.97 0.15 12 2 Total percent 32% 68% 100%

Households food security status and headship Sex of households 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 9 percent of the sample households. The survey result indicated that 8.3 percent of food insecure households were female headed whereas, the corresponding figure for male headed households was 91.7 percent. Male headed households comprise 89.9 percent of food secure and the remaining 10.1 percent food secure are female headed households. The survey result showed no significant difference (p > 0.10) on food security status of households in terms of households head sex. The survey result showed insignificant which shows sex and food security status are unrelated (Table 6).The possible reasons may be the female headed households coudn't face any labour shortage problem due to capability of thire son for farming activeties.

54

Table 6:- households food security by sex households head Male Female Total 111 10 121 Food insecure(N=121) NO percent 91.7% 8.3% 100% Food secure(N=79) NO 71 8 79 percent 89.9% 10.1% 100% Total (N=200) NO 182 18 200 percent 91% 9% 100% 0.2024 2

Source: Survey result (2011) Households food security and education of households head Most households heads in the survey were found to be Illiterate (55.5%) followed by junior level education (23.5%) and primary (17.5%). There was specific pattern that indicated the higher the level of education of the households head, the more food secure a households will be. It was revealed that households headed by illiterate persons were more vulnerable to food insecurity followed by primary. In societies such as Ethiopia where households heads are the major breadwinners of the households, households heads educational status could determine food security status of the entire households. Among other things, households heads play a pivotal role in shaping family members towards educational attainment thus reducing the probability of being food insecure.

It was hypothesized that households food security and education of households head has positive relationship. The survey result showed significant relationship at 5 percent probability level when households 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 households head as literate and illiterate exhibited that 44.5 percent of households heads were literate. Among the literate households 59 percent were found to be food secure and out of 111 illiterate households 64.5 percent were food insecure (Table 7).The possible implication is that in addition to other factors, while some level of education is important to households food security

55

Table 7:- households food security by educational status. Level of education Illiterate 1-4 5-8 Above 8 Total Food insecure N=121 78 25 15 3 121 percent 64.5% 20.7% 12.4% 2.4% 100% Food secure N=79 33 10 32 4 79 percent 41.8% 12.7% 40.5% 5% 100% N=200 111 35 47 7 200 Total percent 55.5% 17.5% 23.5% 3.5% 100% Pr = 0.002*** 9.96 2

Source:-Survey result (2011) Number of months food item purchased

*** Significant at less than 1% probability level

With respect to the specific characteristics of food secure and food insecure households number of months food item purchased was hypothesized to be negatively or inversely related with food security. So, households with small number of months food item purchased tend to be food secure than those with large number of months food item purchased. Accordingly, the statistical analysis showed that number of months food item purchased is significantly different for the two groups at one percent probability level. The mean number of months food item purchased for food insecure households is about 4 months while that of food secure average number of months food purchased is almost 1.5 months. Almost half of the sample households fill the food demand gap by purchasing food items for three to four months (Table 8). The possible reasons for purchase is due to existence of inadequate own production.

56

Table 8:- Distribution of sample households by number of months food item purchased Number of months food purchased Does not purchase 1--2 3--4 Above four Total Mean SD Maximum Minimum t-value Food insecure N=121 25 41 55 121 3.7 1.15 7 1 -0.0104 Source:-Survey result (2011) percent 20.6% 33.9% 45.5% 100% Food secure N=79 25 36 14 4 79 1.5 1.13 5 0 Pr(T > t) = 0.0000*** *** Significant at less than 1% probability level percent 31.5% 45.5% 18% 5% 100% N=200 25 61 55 59 200 2.87 1.57 7 0 Total percent 16.5% 23.5% 49.5% 10.5% 100%

Age of households head: - The mean age of sample households heads was found to be 46.13 with standard deviation of 11.53. The statistical analysis revealed that there was no significant difference in the mean age of the households head between food secured and food insecure households heads at 5 percent probability level of significance. This finding was turned out to be opposite to the prior expectation, which argued as the age of the households head increases since he can acquire more knowledge and experience he would be more prone to be food secure.

57

Table 9:- Distribution of sample households heads by age households head age < 35 35-46 47-59 60 Total Mean SD Maximum Minimum t-value 0.28 Food insecure N=121 13 60 26 22 121 46.32 11.25 80 24 percent 11% 50% 21% 18% 100% Food secure N=79 10 39 14 16 79 45.83 12.00 75 28 Pr(T > t) = 0.3871 percent 13% 49% 18% 20% 100% N=200 23 99 40 38 200 46.13 11.53 80 24 Total percent 11.5% 49.5% 20% 19% 100%

Source:-Survey result (2011) Land holding per households and per capita From any other productive resources land is by far the most important resource in agriculture. The fertility status, location and other attributes of land in association with its size made it a binding resource in agriculture. In the study area, as witnessed by the survey result there is significant difference in the mean cultivated land size between the food secure and food insecure households. The mean farm size of food secure and food insecure households was found to be 1.19 ha and 0.91 ha, respectively. The overall mean farm size was 1.01 ha. As indicated in the table below, about 4 percent and 29 percent of the total food secure and food insecure households groups had farm size less than 0.47ha. But, percentage of sample households in the other extreme holding brackets differs sensibly. For instance, 71 percent of the food insecure and 96% of the food secure had farm size greater than 0.47 hectare. Similarly, it was the largest

proportion of sample households (81%) had farm size greater than 0.47 hectare.

58

Table 10: - Distribution of sample households by farm size per households (in hectare) Farm size/HH Food insecure N=121 < 0.47 0.47 Total Mean SD Maximum Minimum t-value 35 86 121 0.91 0.60 3 0.125 -3.38 Source:-Survey result (2011) Livestock holding Livestock production plays an important role both in the crop producing and pastoral areas of the study area. Livestock provide milk, meat, manure, traction power and transport. Livestock that are owned by the sample households include cattle, sheep and goat, equine and poultry. The total livestock population owned by the sample respondents was 742.558 in TLU. The percent share of goat and sheep is larger than any of the other types of livestock among the sample households. This signifies the importance of small ruminant production in the study area, both as a store of wealth and as check or control of food shortage during time of stress. Almost all households in the sample own livestock with different combination. Survey result demonstrated that the average numbers of livestock holding between the two groups of sample farmers differ. In order to make comparison of the animal size between the farmer groups, the herd size was converted into livestock units (TLU) based on Storck et al. (1991) (see Appendix 1).Food secure group own averagely relatively larger number of livestocks 6.6 in TLU and the food insecure households owns averagely 2.9 TLU only. Almost more than half (54%) sample households own less than 3.94 TLU and 46% sample households own greater than 3.94 TLU. The categories of livestock size indicate the wealth status of the households and the variation in this
59

Food secure N=79 3 76 79 1.19 0.58 3.5 0.5 Pr(|T| > |t|) = 0.0009 *** percent 4% 96% 100%

Total N=200 38 162 200 1.01 0.61 3.5 0.125 percent 19% 81% 100%

percent 29% 71% 100%

*** Significant at less than 1% probability level

aspect may indicate variation in vulnerability of the households to food security problem. The mean difference between the two groups is statistically significant at less than one percent level of significance. Table 11:- Distribution of sample households by Livestock Holding by TLU Livestock holding in TLU Food insecure N=121 < 3.94 3.94 Total Mean SD Maximum Minimum t-value -15.3 75 46 121 2.9 1.5 11.6 0.13 Pr(|T| > |t|) = 0.0000*** *** Significant at less than 1% probability level percent 62% 38% 100% Food secure N=79 33 46 79 6.6 1.7 14.3 0.13 percent 42% 58% 100% Total N=200 108 92 200 4.4 2.4 14.3 0.13 percent 54% 46% 100%

Source:-Survey result (2011) Ox ownership

Livestock is an integral part of crop production activities in the study area. It provides substantial non-human labor and manure to the soil. With regard to the contribution of labor, oxen ownership is an important variable. In the study area, as witnessed by the survey result only twenty households owns more than 2 oxen (10%). Otherwise the households are ox-less or own only one or two. As it is shown in the table below, above 31 percent of sample householdsare ox-less and owns only one ox and only 28 percent of the sample owns two oxen. Moreover, the result revealed that there is significant difference between the two groups with regard to ox ownership. This high variability between the two groups is correlated with the cost of maintaining an ox and some households use farm tools for their farm operation in the study area.

60

Table 12:- Distribution of sample households by oxen Number of oxen Food insecure N=121 No oxen 1 2 2 Total Mean SD Maximum Minimum t-value -3.68 41 44 28 4 121 0.81 0.86 4 0 percent 36.4% 36.6% 23% 4% 100% Food secure N=79 21 18 24 16 79 1.32 1.01 4 0 Pr(T > t) = 0.0003*** *** Significant at less than 1% probability level percent 26.6% 22.8% 30.4% 20.2% 100% Total N=200 62 62 52 20 200 1.01 0.95 4 0 percent 31% 31% 28% 10% 100%

Source:-Survey result (2011)

Relief food aid Food to the households is acquired either from own production or through purchase. When households exhaust their own produce, they attempt to entitle themselves to the food they want through purchase. However, households mostly fail to do so due to the fact that income from other sources is not sustainable and hence they depend on relief food aid. The study area is known with its sequential food aid receive. Food aid plays a role to lessen the households from being vulnerable to sever food insecurity. In this study it was hypothesized that households who received more aid will be more likely to escape from being vulnerable to food insecurity than those who received less. However, the mean amount of food aid received by the two sample households groups revealed no significant difference (Table 13). The statistical insignificant difference with respect to the mean amount of food aid received between the food secure and food insecure households groups is possibly because of the targeting problem. That is, the food

61

aid is distributed without discriminating the two groups and some times for all with no participation in EGS program. Table 13:- Distribution of households by amount of food aid received in Kg
Food aid received in Kg Food insecure N=121 Received no food aid Received less than 75Kg Received greater than 75Kg Total Mean SD Maximum Minimum t-value 1.01 65 22 34 121 46.86 63.57 375 0 Pr(T > t) = 0.1572 percent 54% 18% 28% 100% Food secure N=79 45 18 16 79 38.10 54.17 270 0 percent 57% 23% 20% 100% Total N=200 110 40 50 200 43.4 60.05 375 0 percent 55% 20% 25% 100%

Source:-Survey result (2011) Use of technology The table below shows the distribution of sample households by status of use of agricultural technologies. In the survey it was observed that 45.5% of the overall sample households are users of fertilizer and improved seed. To compare the two sample groups 22% of the food insecure households were users of fertilizer and improved seed. While the corresponding food secured households who were users of fertilizer and improved seed were 81%. Though such difference was observed between the groups, it was statistically significant. The result also showed that there is significant relationship between food secure and food insecure at less than 1percent.

62

Table 14:- Distribution of sample households by status of use of technology Agricultural in put Food insecure Food secure Total N=121 Use in put Does not use in put Total 27 94 121 percent 22% 78% 100% N=79 64 15 79 percent 81% 19% 100% N=200 91 109 200 percent 45.5% 50.5% 100%

Pearson chi2(1) 2

2= 85.1464

Pr = 0.000***

Source:-Survey result (2011) Off-farm income

*** Significant at less than 1% probability level

households in the study area perform various off farm activities like livestock trading, grain trading, wage employment, timber production, hand craft and vegetable trading etc. The income from such activities greatly improves the households purchasing power in the study area especially during time of stress. Some of the people in the study sample had no access to offfarm work. From 200 households sampled, 104 households heads had off-farm employment while the rest of the households did not. 49 percent of those who had off-farm employment were food secure while 53% were food insecure households. Table 15 shows the distribution of households by income from off-farm activity. The survey result revealed that about 48 percent of the sample households earn less than Br. 500 from these sources in the study area. But when we further look the results within the sample groups above 68 percent of the food insecure and above 51 percent of the food secure households earn less than Br. 500 from off farm activity per year. Moreover, the survey result also revealed that there is significant difference in the mean annual income from off farm activity between the two sample groups at less than 1 percent (p<0.01) level. This result is in a complete agreement with the prior expectation.

63

Table 15:- Distribution of sample households by off-farm income in birr. Off-farm income Food insecure N=121 500 501---1000 1001---2000 Above 2000 Total Mean SD Maximum Minimum t-value Source:-Survey result (2011) -5.01 83 22 12 4 121 399 813.7 7000 0 percent 68% 18% 10% 3.5% 100% Food secure N=79 51 12 13 3 79 951 676.17 3600 0 Pr(|T| > |t|) = 0.0000*** *** Significant at less than 1% probability level percent 51% 30% 15% 4% 100% Total N=200 134 34 25 7 200 percent 48% 36% 12.5% 3.5% 100% 617.3 807.3 7000 0

Irrigation Access to irrigation is expected to have a positive relationship with households food security (Burton et al., 2005). Farmers with plots on the irrigation schemes are able to grow crops throughout the year and meet households food requirements than those on dry land farming. Based on the analysis only 12% of sample households wear access to irrigation and 88% of sample households do not have access to irrigation and practice dry land and rain fed

agriculture. However, households that have irrigation farm size below one hectare were 7% for food insecure and 15% for food secure. The table also shows that it is impracticable to be food insecure holding irrigation farm size above one hectare. While the survey result also revealed that, there is significant difference between the two groups at less than 1 percent (p<0.01) level. This result is in a complete agreement with the prior expectation.

64

Table 16:- Distribution of sample households by irrigation farm size in hectare Food insecure Food secure Irrigation participation N=121 No irrigation farm size 0.51hectare Above 1hectare Total Mean SD Maximum Minimum t-value -2.48 112 9 121 0.74 0.26 1 0 Pr(T < t) = 0.0069*** percent 93% 7% 100% N=79 64 12 3 79 percent 81% 15% 4% 100% 0.18 0.39 2.25 0

Total N=200 176 21 3 200 0.12 0.32 2.25 0 percent 88% 10.5% 1.5% 100%

Source:-Survey result (2011)

*** Significant at less than 1% probability level

Access to infrastructure Nearness to infrastructure creates access to additional income by providing access to increase productivity, easy access to inputs, transportation and off-farm employment opportunities. The shortest distance to the infrastructure was used in this study to determine how often the farmers used the infrastructure to sell and purchase their produce. It was, therefore, expected that households nearer to infrastructure have better chance to improve households food security status than who do not have proximity to infrastructure centers. Table 17 depicts the statistical results of the two groups in relation to the effect of infrastructure distance on food security. The result was statistically insignificant at less than 5 percent probability level. Additionally, the mean distance of food secure group is greater than the food insecure ones. This result is contrary to the hypothesis stated for this study. The occurrence of this result may be due to the fact that when farmers are nearer to the infrastructure they would be tempted to sell their products and buy non food items. But if they were away from infrastructure centers they would be food secure by consuming their own product.

65

Table 17:- Distribution of sample households by infrastructure distance in Km Market distance(Km) Food insecure N=121 Less than 5Km 510km 1115km Above 15km Total Mean SD Maximum Minimum t-value Source:-Survey result (2011) -0.67 84 31 5 1 121 percent 69% 26% 4% 1% 100% 3.28 3.5 18 0 Pr(T > t) = 0.7476 Food secure N=79 52 23 1 3 79 percent 66% 29% 1% 4% 100% 3.66 4.44 25 0 Total N=200 136 54 6 4 200 percent 68% 27% 3% 2% 100% 3.43 3.89 25 0

4.3 Incidence of food security by households factors Table 27 shows the way in which some households factors affect food security by comparing the incidence of food security among households groups sharing similar characteristics. The occurrence of food security among households with less than four members is found to be 1.5 times more than that of households with five to eight and 2.7 times more than that of households with nine to twelve. This decrease incidence of food security with increase in family size confirms the result of the logit output discussed in the last section. In that the two variables have negative relationship. As hypothesized the incidence of food security also decreases as the proportion of children and elders increase in the family. This is also shown in the Table 27where incidence of food security in households with dependency ratio greater than 2 is found to be 1.4 times lower than that of households with dependency ration less than 1. Moreover, the head count index is 1.7 times higher in literate households heads than that of illiterate ones. Hence, the risk of incidence of food security decreases with education, i.e., to say when the households
66

turned from illiterate to literate status the incidence for the household to become food secure increased. With regard to gender of households head, female-headed households have lower incidence of food security than male-headed ones, i.e., 10%female and 90% male, respectively. Likewise prevalence of food security increase as farm size of the households increases. Table 18:- Incidence of food security by sample households Characteristics Family size households grouping 14 58 912 Over all Education Illiterate Literate Over all Dependency ratio <1 12 >2 Over all Sex Male Female Over all Land size < 0.5 0.51 1.12 Above 2 Over all Food secure 14 56 9 79 33 46 12 21 46 12 79 71 8 79 3 40 31 5 79 Total 23 137 40 200 111 89 200 65 109 26 200 182 18 200 62 75 55 8 200 Food security incidence 60.9 40.9 22.5 39.5 29.7 51.7 39.5 32.3 42.2 46.2 39.5 39.01 44.4 39.5 20.9 40 56.4 62.5 39.5

Source: - Survey result (2011)

67

4.4 Extent of Food security Food insecurity gap and severity of food insecurity were the indexes employed to capture the incidence and severity of food insecurity. In the study area the incidence of food security was found to be 0.395. That means only 39.5 percent of the sample households can meet the energy requirement recommended for subsistence. In other words, head count ratio of 0.395 for 200 sample households means 97 sample households are deemed food secure. The head count index or incidence of food security is good indicator to assess food security but it does not take into account the severity of the food security problem. Therefore, to address how far the food insecure households are below the subsistence energy requirement level, food security gap was calculated from the survey data. Accordingly, the food security gap index came out to be 0.73. This means that if the council mobilizes resources that can cover or meet the 73 percent of the daily calorie requirement for every food insecure households and distribute these resources to bring each households up to the given daily calorie requirement level, then at least in theory food insecurity will be eliminated. In other words, assuming that the sample households are representative to the rural population of Kamba and according to Office of Agriculture, it was estimated to be 26,258 farming households in Kamba woreda which wear on average equivalent to 140,480 in AE. Hence, based on the recommended subsistence energy (2100 kcal per day per person), the total resource required to bring all households at least to get the daily subsistence is amounted to 295,008,000kcal per day. When this amount of calorie is converted to cereals, assuming that cereals can produce an average of 3700 kcal per kg, it becomes 797 quintal of cereals per day. This implies that an estimated 291,702quintal of cereals per year is required to bring all households at least to get the daily subsistence energy in a year.

4.5 Summary of mean difference and households scores Table 18 and 19 below show summary statistics and households scores of sample households groups on the hypothesized continuous and discrete variables included in the descriptive analysis. According to the survey result depicted in Table 18, food insecure and secure households groups revealed significant difference with respect to some socio-economic variables like family size (FASZ), total farm size (FARMS), livestock holding in (TLU), total off-farm income (TOFFI), number of oxen households own (NOXEN), number of months food
68

item purchased (NUMFP) and access to irrigation (ACCIRR) at probability level less than 1 percent. On the other hand, Table 19 shows categorical variables with the chi-square value which shows the existence of significant relationship between food secure and food insecure sample households include the sex of households head (SEX), strong relationships between food security status and technology adoption statues of sample households (TEC) and existence of relationship between food security status and educational status of the households head (EDUC). Table 19:- Summary Statistics of continuous variables included in the descriptive statistics Variable Food insecure (121) Food secure (79) t- value Mean FASZ AGE NUMFP FARMS TLU NOXEN FAID TOFFI INFRA ACCIRR 7.04 46.32 3.7 7.04 2.9 0.81 46.86 399 3.28 0.74 SD 0.17 11.25 1.15 0.17 1.5 0.86 63.57 813.7 3.5 0.26 Mean 4.32 45.83 1.5 4.32 6.6 1.32 38.10 951 3.66 0.18 SD 0.13 12.00 1.13 0.13 1.7 1.01 54.17 676.17 4.44 0.39 12.72 *** 0.28 -0.01*** -3.38*** -15.3*** -3.68*** 1.0 -5.01*** -0.67 -2.48***

Source:-Survey result (2011)

***, is significant at 1% probability level.

69

Table 20:- summary statistics of discrete variables included in the descriptive statistics Variable Score Food insecure (121) Number SEX 1 0 TEC 1 0 EDUC 1 0 111 10 27 94 43 78 Percent 91% 9% 22% 78% 36% 64% Food secure (79) Number 71 8 64 15 33 46 Percent 90% 10% 81% 19% 42% 58% 9.96*** 85.15*** 0.2024 Chi-square value

Source-survey result (2011)

***, is significant at 1% probability level.

4.6 Analysis of determinants of food security


An econometric model, logistic regression, was employed to identify the determinants of households food security. However, before fitting the logit model, it was important to check whether serious problem of multicollinarity, heteroscedasticity and association existence among and between the potential continuous and discrete explanatory variables of the model estimation, respectively. For this purpose, variance inflation factor (VIF) and contingency coefficient were used for the continuous and discrete variables, respectively. Value of VIF greater than or equal to 10 is an indicator for the existence of serious problem of multicollinearity. Table 20 presents the value of VIF for each of the continuous variables. As it is shown in the Table, the VIF of all the variables were found to be smaller than 10. Hence, the problem of multicollinarity was not serious among the variables. As a result, all the hypothesized 9 continuous explanatory variables were included in the model estimation. Heteroscedasticity problem (different variance) was also diagnosed by using commonly used 'Robust' order in STATA-10.

70

Table: - 21 VIF value of continuous variables Variable FASZ FRMS TLU TOFFI EDUC FAID INFRA SAVG NUMFP Source: Own computation Similarly contingency coefficient was calculated for the discrete variables. Contingency coefficient value ranges between 0 and 1 and as a rule of thumb variable with contingency coefficient less than 0.5 assumes weak association between the variable. Contingency coefficients for technology adoption, participation in public meeting, land quality and extension service adoption of households heads were found to be lower than 0.25. Since this figure shows the absence of serious association (multicollinearity) between the variables, all the four discrete variables were included in the model. Table 22:- contingency coefficient for discrete variables LNDQ EXTN PAPUM TECH R2 0.503 0.464 0.503 0.209 0.144 0.091 0.295 0.213 0.222 VIF 1.42 1.87 2.01 1.26 1.17 1.10 1.42 1.27 1.29

LNDQ EXTN PAPUM TECH

1.0000 -0.0759 -0.1437 0.0750 Source: - Own computation 1.000 0.0857 0.1514 1.000 0.2321 1.000

The variable food security (FODS) was used as a dichotomous dependent variable, with an expected mean value of 1, indicating the probability of being food secure, 0 otherwise. Generally, there were 13 explanatory variables included in the model analysis. In order to
71

identify the most important factors from the hypothesized potential variables to influence food security, binary logit model was estimated from the survey data. For the purpose, STATA Version 10 was employed. Codes, types and definitions of the variables; and the maximum likelihood binary logit estimates are presented in Table 22 and Table 23below, respectively.
Table 23- The maximum likelihood estimates of the logit model Variable Coefficient p-value Wald Statistics FASZ FRMS TLU TOFFI EDUC FAID TEC INFRA PAPUM EXTN SAVG NUMFP LNDQ Constant -5.442 9.91 1.62 0.00 0.5 0.00 6.47 -0.07 7.18 5.27 0.00 0.82 8.32 -11.50 0.000*** 0.000 *** 0.000 *** 0.002 *** 0.049 ** 0.657 0.000 *** 0.087 0.000*** 0.010 ** 0.106 0.031** 0.000 *** 0.012 53.33*** 17.63 250.74 94.94% 97.52% 96.50% 200 *** Significant at less than 1% probability level 26.13 13.42 19.01 10.05 3.87 0.20 13.54 2.94 14.54 6.57 2.61 4.64 13.31 Marginal Effect (dF/dx) -0.033 0.06 0.01 0.000 0.003 0.000 0.0475 -0.0005 0.100 0.011 6.12 0.005 0.85

Wald chi-square -2Log likelihood Likelihood ratio test Sensitivity Specificity Percent correctly predicted (count R2) Sample size

Source: Model output

** Significant at less than 5% probability level

72

The likelihood ratio test statistics (250.74) exceeds the chi-square critical value (53.33) with 13 degree of freedom. The result is significant (p=0.000) at less than 1 percent probability level indicating that the hypothesis that the coefficient except the intercept are equal to zero is rejected. Another measure of goodness of fit used in logistic regression analysis is the count R2, which indicates the number of sample observations correctly predicted by the model. The count R2 is based on the principle that if the estimated probability of the event is less than 0.5, the event will not occur and if it is greater than 0.5 the event will occur (Maddala, 1989).The HL(Hosmer-Lemeshow) test also revealed that the model has H-L value 1.00 which indicates convergence between expected and observed probabilities, value is not statistically significant at less than 5% probability level, therefore the model is quite a good fit, or indicating that the model prediction does not significantly differ from the observed. In other words, the ith

observation is grouped as a food secure if the computed probability is greater than or equal to 0.5, and as a food insecure otherwise. The model results show the logistic regression model correctly predicted 193 of 200, or 96.5 percent of the sample households. The sensitivity (correctly predicted food secure) and the specificity (correctly predicted food insecure) of the logit model are 94.94 percent and 97.52 percent, respectively. Thus, the model predicts both groups accurately.

4.5 Discussion on the significant explanatory variables Out of the thirteen variables hypothesized to influence households food security, ten were found to be statistically significant. The maximum likelihood estimates of the logistic regression model showed that family size (FASZ), cultivated land size (FRMS), total livestock holding in (TLU), total off farm income (TOFFI), educational status of the households head (EDUC),

technological adoption of households head (TEC), land quality (LNDQ), households head participation in public meeting (PAPUM), extension service (EXTNS) and number of months food purchased (NUMFP) were important determinants identified to influence households food security in the study area. That means, the coefficient of family size, Cultivated land size, total livestock holding, total off farm income, technological adoption of households head, participation in public meeting and land quality were statistically significant at 1 percent probability level of significance whereas educational status of the households head, extension
73

service and number of months food purchased were statistically significant at less than 5 percent probability level of significance. Moreover, the results verified that except number of months food purchased, almost all of the explanatory variables obtained in the model had the signs that confirm with the prior expectations. In light of the above summarized model results possible explanation for each significant independent variable are given consecutively as follows: Family size (FASZ):- This households factor is found to be highly significant to determine households food security in the study area. Households size revealed a negative relationship with food security and statistically significant at 1 percent probability level. The negative relationship indicates that the probability of in favor of being food secure decreases with an increase in the family size. This means that the probability of a households to be food security is zero if households family sizes increase. The marginal effect of a unit change in family size, computed at sample mean of family size, the probability of food secure is -0.033. This means that the probability of food security decreased by -0.033(about -3.3%) for a one member increase in family size. The likely explanation is that in an area where households depend on less productive agricultural land, increasing households size results in increased demand for food. This demand, however, cannot be matched with the existing food supply so ultimately end up with food insecurity.

Cultivated land size (FRMS):- Cultivated land size was hypothesized to influence food security positively. The results of the logit model indicated that sample households which had larger farm size had more possibility of being food secure. This is assured by the positive coefficient of this variable indicating it is significantly influencing rural households food security at 1 percent level of probability. The possible justification is that farm households s which had larger farm size had better chance to produce more, to diversify the crop they produce and also have got larger volume of crop residues. The marginal effect of a unit change in farm size, computed at sample mean of farm size holding in hectare, the probability of food secure is 0.06. This means that the probability of food security increases by 0.06(about 6%) for a one hectare increase in farm size. This result is supported by the findings of Abebaw Shimelese (2003).

74

Total livestock holding (TLU):-The relationship between the amount of livestock holding in tropical livestock unit and food security turned out to be positive and statistically significant. The relationship is statistically significant at 1 percent probability level. This is an indication that ownership of livestock acts as a hedge against food security in the study area. Livestock, besides its direct contribution to subsistence need and nutritional requirement, is a vital input into crop production by providing manure and serves to accumulate wealth that can be disposed during times of need, especially when food stock in the households deteriorates. The marginal effect of a unit change in livestock ownership in TLU, computed at average TLU owned by sample households, the probability of food secure is 0.01. This means that the probability of food security increase by 0.01(about 1%) for a one unit increase in livestock number in TLU. This result was also supported by Abebaw Shimelese (2003).

Total off-Farm Income (TOFFI): This variable represents the amount of income earned in cash or in kind, during the year. In the areas like Kamba, where the farmers face crop failure and sales of livestock and livestock product is inadequate, income earned from off-farm activities is an important means of acquiring food. Accordingly, in the study area, the data revealed that 52% households have off-farm income access and from food secured households 51% have access to off-farm income and from food insecure 54% have access to the additional income. The result suggests that households engaged in off-farm activities are endowed with additional income and more likely to be food secure. Consistent with the hypothesis, off-farm income is positively and significantly associated with farm households food security status (at probability level of 1 %). The marginal effect of a unit change in amount of off- farm income in birr, computed at average amount of off- farm income by sample households increases the probability of food security by 0.002%.

Education (EDUC):- Education is positively and significantly related to the probability of food security in the study area. Education is also explained in terms of contribution to 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. The possible reasons are literate farm households heads are
75

more willing to adopt better production technologies, accept technical advice from extension workers, and diversifying their source of income than the illiterate ones. As a result literacy increases chance of becoming food security among the sample households. The marginal effect of a unit change in level of education, computed at average level of education by sample households, increases the probability of food security by 0.002%.

Technological adoption of households head (TEC):- similar to expectation, use of chemical fertilizer and improved seed was found to be positively and significantly affect the rural households food security in the study area. Thus, the study reveals that most of the farmers who applied chemical fertilizer and improved seed in their farm were more food secure than those who did not apply chemical fertilizer and improved seed. The probable reason for this is that application of chemical fertilizer and improved seed improves the mineral content of the soil and restores the nutrients required by the crops, thus crop yields are increased. Input use by farmers increases the probability of food security in the study area. Dry land farmers like in Kamba, who are risk averse, can also be food secure from its use especially when the rains are adequate for crop production. Keeping other factor constant, the marginal effect of change from technology non user to technology user, computed at mean technology adoption status especially for chemical fertilizer and improved seed by sampled households increases the probability of food security by 5%.

Participate in public meeting (PAPUM):- The objectives of public meeting are to transmit the governments plan and strategy to boost production. Participation in public meeting changes farmers outlook towards production, by helping them to talk the main challenges faced, to identify dutys of government and local farmers and helps to get new arrival and trial technologies. This may be also explained by the factors that the message that farmer gain from public meeting help them to initiate to use risk aversion strategies that seek diversification of income within and out of agriculture. This variable has a positive and significant (p<0.01) association with the households food security. Keeping other factors constant; the odds ratio in favor of food security, cetris paribus, and increases by a factor of 1317 as the households public meeting participation trend changed from non- participant to participant in a year.
76

Frequency of extension contact (EXTNS):- This households factor is found to be highly significant to determine households food security in the study area. The objectives of extension is to change farmers outlook towards their difficulties which assists them adapt better solution to their livelihoods. Thus, the information obtained and the knowledge and skill gained from extension office may influence farmers skill and decision making on seeking diversification. The frequent extension contact received will increase the tendency of households to be food security. Frequency of extension contact has a positive relationship with food security and statistically significant at 5 percent probability level. The positive relationship indicates that the odds ratio in favor of the probability of being food security increases with an increase in the frequency of extension contact. Keeping other factors constant; the marginal effect of a change from lacking contact to having contact to extension service provider, computed at mean number of contact with extension service provider in a year by sample households s, increases the probability of food security by 1.1%.

Land quality (LNDQ):- This is because the increase in the fertility of the land is expected to contribute positively towards increase in crop output and consequently increase in farm income. Empirical findings indicate that land fertility problem has a relation with the level of food security. However, Mulugeta Tefera (2002) and Ayalew Yimer (2003) have shown that land fertility problem do not have significant effect on households food security status. On the contrary to their result, this study revealed that this variable affects households food security significantly. The marginal effect of a change of farmers land from poor land quality to good land quality, computed at mean farmers perception on the fertility of their farmland quality by sampled households s, increases the probability of food security by 84.7%. This gigantic response may be achieved after challenging and time consuming task of soil quality improvement in long period of time. It is not an easy task and the response is not automatically, it needs more time and actors commitment to get the above probability percent on food security.

Number of months food purchased (NMFP):- As the number of months food purchased increase for rural farmers, the probability to be food secure decrease due to weak income source stability which forces to purchase low quality food items. Subsistence farmers like in Kamba
77

produce food for home consumption and buy food items when there is food shortage in the family. If the number of months that a household buy food items is more, he has low own production and depend on purchase for his food demand. Number of months food purchased has inverse relationship with food security statues of households. But the logit model revealed that it has positive relationship with food security. The possible reason can be the farmer has chance to buy diversified food items and goat higher calories than own production and frequent consumption of the same food items. The marginal effect of a change from few numbers of months for food purchase to most number of months for food purchase, computed at average number of months food purchased by sample households, increases the probability of food security by 0.5%.

4.7 Major Agricultural Problems Different reasons were given concerning the declining trend in production. The responses of sample farmers on major reasons for the declining trend of crop production are shown in Table 25. Inadequate rain fall, soil infertility, small land holding, lack of credit, inadequate extension service, poor technological diffusion and lack of oxen problems were ranked as a very serious problems of farming households in the study area. Out of total respondents who cited the various problems, about 60% of them mentioned inadequate rain fall, 45% mentioned soil infertility, 41% mentioned small land holding and 39% mentioned lack of credit problem has been very serious problems in the study area. Inadequate rainfall is the most frequently cited agricultural problem. With regard to the proportion of farmers who respond on the major causes of food insecurity problems (Table 25), relatively small numbers of the food secure farmers reported to have these problems as compare to those food insecure groups. For instance, 38% and 44% of food secure and food insecure farmers were cited absence of rainfall. In general, the poor performance traditional farming practice that has greatly affected the sustainability of production and productivity joined with the inadequate and erratic rainfall has made districts rural farm households more vulnerable and food insecure. The responses of sample farmers on trend of food availability in the last five years revealed that 56% confirmed that food availability was declined from year to year.

78

Soil fertility problem is one of the physical factors affecting crop production. The relationship between soil fertility problem and state of food security indicate that soil fertility problem has negative impact on crop production performance, and causes a deterioration of food security status of the households. The proportion of farmers who reported to have soil fertility problem is more for food insecure than secured groups. About 20% of food insecure and 17% of food secure farmers reported to have soil fertility problem in their farm.
Table 24:- Major agricultural problems encountered in the study area

Source of problem N=121 Inadequate rainfall soil infertility small land holding lack of credit inadequate extension service poor technological diffusion lack of oxen Total
Source:-survey result

Food insecure Percent 44% 20% 15% 8% 5% 4% 4% 100% N=79 30 13 10 8 8 6 5 79

Food secure Percent 38% 17% 13% 10% 10% 7% 5% 100%

53 24 18 10 6 5 5 121

4.8 Extension services In a country such as Ethiopia, where the majority of the farmers are illiterate, agricultural extension plays a significant role in assisting them by identifying and analyzing their production problems and by making them aware of opportunities for improvement of food security. Hence, the effectiveness of the other inputs in production partly relies upon the availability of sound agricultural extension services at community levels. The traditional and widely used means of conveying new information to farmers is through the public extension services. In Ethiopia, including the study area, development agents trained in agricultural sciences are assigned in each kebele. The current agricultural policy gives emphasis to the development of private enterprise including development of smallholders agriculture. The extension program requires farmers to
79

use package of new varieties, chemical fertilizer, farm credit etc. However, only 39 % of the sample farm households identified themselves to be beneficiaries of continuous extension services the reaming 61% couldnt get continuous extension service. The number of DA in the woreda increased in number from 48 to 118 since 1994E.C, but the service promotion by them is at infant stage. The percentage of food secure farmers with access to extension services is relatively higher than the percentage of food insecure farmers. This could be because that food secure farmers have more frequency of contact with the extension agents. Both the descriptive and econometric results revealed that contact with development agent and application of improved technologies were influential factors for food security in the study area.

4.9 Coping Strategies Farm households in a vulnerable area like Kamba engage in several activities in order to avoid food shortage and famine. These include adjustments farmers would make to cope with food supply pattern, reduce amount of food consumed ,sale of livestock, labor (including migration) in search for employment opportunities, sales of productive assets and stocks, depending on food aid. Farmers were asked about how they manage food shortage and how they can cope with food insecurity. This section describes the results of the interview and relates the response to the farmers actual activities.

The local coping strategies, which have been practiced during food crisis by groups of sample farmers in Kamba, are presented in Table 26. The principal strategies used by the sample respondents to mitigate food supply shortage include purchasing of grain, diversification of crop production into drought resistant crops such as cassava, 'enset', sorghum, sweet potato haricot been and others during short and erratic rainy seasons to meet their subsistence needs. Almost all (99%) respondents diversified their production by producing drought resistant crops. About 52% of all respondents and 49% of food secure and 54% of food insecure households engaged in offfarm jobs. Even though, there was limited access to off-farm work opportunity in the district, resource poor family migrate(17%) either within woreda or outside woreda for wage earned in kind or cash. Another important coping mechanism considered first by farmers was sale of livestock. Livestock, besides their complimentary relationship with crop production, provide
80

hedging against risk of food insecurity. As a result, when food produced is fully consumed and or no cash reserve is available to purchase more of it, animal products and live animals are sold as ways of getting access to cash income and to buy food for the households. Accordingly, about 74% of all households, 68% of the food secures and 94% of the food insecure households were involved in the sales of animals (mostly small ruminants) to acquire food whenever there is shortfall in food supply. Sales of animals were common for the two groups and these shows that the farm households keep animals as principal assets to manage the shortage. This mechanism is ranked as the first most important coping practice, followed by, reducing food consumed, and involvement in off-farm and food aid. Sales of live animals to purchase food grains during supply shortage have considerable effects on farmers' economy mainly because of sharp decline in livestock prices due to excessive supply from all affected farmers. The proportion of food secure and food insecure households who practiced purchasing grains/food items during food supply shortage were 92% and 95%, respectively. Reduction of consumption in terms of both the number of meals per day and amount of food per meal was identified as means of coping for the largest proportion (40%) of the respondents, 25% of the food secure and 53% of the food insecure sample households during short supply. About 23% of all cases, 7% of the food secures and 30% of the food insecure households reported that they overcome food shortage problems by receiving relief food freely from government and non-government organizations. These and other less frequently mentioned and practiced coping strategies are shown in Table 26. Table 25:- Coping strategies common in Kamba woreda Type of coping Over all (200) Reduced amount of food consumed Sale of livestock Wage work and migration Sale of productive asset and jewelers Food aid Eating wild food(in pastoral kebele) Source:-Survey result 40% 74% 52% 2% 23% 7.5% Insecure (121) 63% 94% 54% 85% 30% 85% Secure (79) 58% 73% 34% 25% 7% 2nd 1st 3rd 5th 4th 3rd Rank

N.B percent does not sum up to 100, due to multiple response.

81

The analyses of the coping mechanism of the sample farmers have shown that, coping mechanisms have different patterns. All farmers were not equally vulnerable to drought or food insecurity; they responded in different ways. Some households implement some coping strategies after all other options have been pursued and exhausted. As the food crisis persist, households are increasingly forced into a greater commitment of resources, just as the households exhaust the strategies that are available in the early stages of food crisis, they begin to arrange key productive assets such as draft oxen, milk cow and assets and jewelers. Accordingly, among the sample households 2% of them sold key productive assets and jewelers as coping mechanism for food insecurity. With respect to the period of severe food shortage that the farm households s practice these coping mechanisms differ area to area, more than 87% of the households in low land and mid- altitude in the study area encountered severe food shortages during the months of April, June and July and highlands in September, October and November. According to woreda agriculture office, farm households in the lowland ecological zone and pastoral areas face severe food shortage more frequently than those in the highlands. Pastorals in the study area begun to eat wild foods as vulnerability increasing, and they shift to the consumption of the cheapest and less quality food items. In general, the proportion of households with local coping strategies implies the extent to which most of the Kamba district's farmers are vulnerable to seasonal shocks and how food security problem is serious. Hence, factors like chronic poverty, rapid population growth, declining per capita food production, poor marketing infrastructure, lack of off-farm job opportunities, small land holding and lack of credit facilities aggravated food security problem and made households more vulnerable.

82

CHAPTER SIX SUMMARY AND RECOMMENDATION


6.1 Summary of findings
Food security problem is the most crucial and persistent problem facing millions of rural households in Ethiopia. Even though the country has considerable agricultural potential, more than half of its rural households are unable to feed themselves throughout the year and yet food availability in the country is largely determined by domestic staple food production by subsistence agriculture. Producing enough food and achieving food security can be made possible through increased agricultural productivity, increased off-farm income and improving the ability of rural households to smoothen and stabilize their income and purchasing power, this problem remained as a top and major challenge than ever met by the Ethiopian government.

This study was conducted in Kamba woreda of Gamo Gofa zone, where food security problem is becoming virtually a continuous concern of most households. Food security problem is now a crucial problem in Kamba woreda. Most of the farming households in the district have difficulties to cope with the situation even during normal seasons. Drought stimulated food security problem has been a recurrent phenomena on exacerbating the vulnerability of the resource poor farming households in the district. This chapter provides the conclusions and recommendations drawn from the data analyzed on how agricultural production could be improved in enhancing small-scale farmers to meet households food security. The chapter looks at the aims of the study that were stated in the introductory chapter (Chapter 1), so that conclusions and recommendations can be drawn.

The study area is not an exception of the above facts. The adverse climate nature of the environment coupled with poor soil fertility, high population growth, lack of sufficient moisture, week technology adoption, lack of infrastructure and traditional way of cultivation pulled back the productivity of agriculture and ultimately resulted in food security problem to many of the rural households. Aware of these problems, the study was carried out with major objective to establish which determinants played a major role in addressing households food security
83

problem in Kamba woreda. Other factors were also put into account in trying to explain which factors are significant in enhancing households food security and estimation of the food security gap and severity of food security problem.

The socio-economic characteristics of both the food secure and insecure households and the livelihood strategies of the rural households in the study area were also considered deeply. To accomplish these objectives primary data on food consumption habit, demographic & socioeconomic characteristics, livestock and crop diversity, access to productive resources, access to service and infrastructure, marketing and credit, technology adoption trend, farm and off-farm income, coping mechanisms etc. were gathered, organized, analyzed and interpreted to come with possible results at the households level from 13 randomly select sample PAs. The analysis employed both descriptive statistics and econometric methods. Descriptive statistics were employed to illustrate weather there was strong relationship between dependant and explanatory variables and existence of mean difference between the two groups with respect to food security status. Binary logistic model was specified and estimated to identify determinants of food security in Kamba woreda. FGT index was used for the computation of incidence and severity of food security problem among sample households.

Based on the survey data, an attempt was made to describe the socio-economic characteristics of the food secure and food insecure sample households groups, i.e., whether there exists mean difference between the two groups with respect to the different socio-economic attributes. Accordingly, the survey result revealed that there was significant difference with respect to mean of family size both in number and adult equivalent, and in the mean of dependent individuals. Large households have more people to feed as compared to small households thus, reducing the calorie intake per households member decreasing the food security in those households. The mean in each case is lower for the food secure households group and higher for food insecure households. On the other hand, farm size, amount of off-farm income received also differentiated the two groups significantly in their mean values. In all the two cases, the mean values were higher for the food secure households group. The total livestock owned (TLU) was found to be significantly and positively related (1% probability level) to the probability of being
84

food secure. The number of livestocks owned in TLU differentiated the two groups significantly in their mean values. Moreover, with regard to education, percent of illiterate households was higher for the food insecure households group than the food secure ones. And there was systematic relationship between education and food security. The use of modern technology (TEC) especially chemical fertilizer and improved seed has come out to be significant (1% probability level) and positive influence on the food security status of the households. The positive sign is an indicator of existence of efficient relationship between technology adoption and food security. Number of oxen owned (NOXEN) is significant and has a positive association (at1% probability) in affecting households food security situation. This implies that the existence of oxen differentiated the two groups significantly in their mean value affects the households food security in such a way that households who owned oxen have better chance to escape serious food insecurity than those who dont owned. Both number of months food item purchased and access to irrigation farm influence the households food security positively and significantly.

Binary logit econometric model was estimated using the survey data to identify the determinants of food security among the rural households in the study area. Accordingly, the estimated coefficient revealed a varied impression. On the one hand, family size, farm size, livestock owned in TLU, total off- farm income, education, technology adoption , adoption of extension service, land quality and participation in public meeting showed theoretically consistent and statistically significant effect while coefficient of number of months food purchased showed theoretically inconsistent and statistically significant effect. On the other hand, amount of food aid received, saving and access to infrastructure were not found to be statistically significant in determining food security.

A closer look at the model result reveals that the variable family size influenced the households food security negatively and significantly. This means the probability for the households becoming food secure decreases as the households size increases.

85

Educational status of the households head also exhibited positive and significant coefficient. This means the variable is directly related with food security. Likewise, farm size and adoption of technology were also other significant variables came out to be positively and significantly related with food security. Total off-farm income, participation in public meeting, land quality and adoption of extension contact were exhibited positive and significant at less than 5 percent. Number of months food purchased on the other hand, as opposite to prior expectation and has a positive and significant coefficient in the estimated model result. This means those farmers who buy food items from market for longer months may have chance to get diversified amount of calories and increased chance of food security.

The other important variable is livestock holding. This variable in agreement with the prior expectation came out to be positively and significantly related with food security. This is so due to the fact that livestock both directly and indirectly contributes to the households food, energy and income requirement. The head count ratio revealed that only 39.5 percent of sampled households are found to be food secure and 60.5 percent of sample households are food insecure. The gap and severity of food insecurity were estimated to be 73 and 3.95 percent respectively. Considering the daily recommended 2100 kcal per adult equivalent, a resource needed to bring all sample households to daily subsistence requirement amounted to 2,248,617 kcal. This shows daily requirements estimate of 607.7 quintals of cereal per day which is equivalent to 221,823 quintals per year.

86

6.2 Recommendations
Hence, after summarizing the findings of this study, the possible policy recommendations that can be made from this study are as follows: 1. Family size and food security are strongly and negatively related. Large households size exerts pressure on consumption than the labour it contributes to production. Rural family size, like in Kamba have high youth dependency ratios(46.1%,CSA 2007) and low life expectancies, it imposes a substantial pull on areas economies by reducing their productive capacity per capita associated with lower rate of savings and investments (as conventionally measured), and therefore slower economic growth and affects directly food security. Recent studies have shown that East Asias rapid demographic transition in the past half century added markedly to the increase in GNP per capita, especially in comparison with regions such as sub-saran Africa where the demographic transition has been delayed (Barry Riley 2002). Relatively poor food productivity and poor public health has probably slowed the demographic transition in the Africa countries. Serious attention has to be given to limit the increasing population in the study area. This can be achieved by creating sufficient awareness to successful family planning through effective extension services and integrated development strategy in the rural households. So family size control is the crucial concern in the rural area like in Kamba otherwise, the ever-shrinking productive resources in the study area coupled with increasing population would hamper any development intervention from achieving its objectives.

2. Productive resources especially land is very limiting and highly binding resource in the study area. The model and descriptive analysis result showed cultivated land size and soil fertility problem was found to be significant and they are positively and negatively related with food security. An increase in land size is likely to increase food security without employing any advanced technologies. Tackling the problem of food security through increasing farm size would not bring any sustainable improvement. So a medium and longer-term food security strategy through increased food production must be introduced. In a medium or shorter term, distribution and
87

allocation of cultivable land, which was not under cultivation, thereby increasing output, should be made. This would give short period relief from the problem; otherwise the amount of return from such a strategy would not be by any means sufficient and sustainable to up-root the problem from the present setting. As a result, strong effort should be made to improve the production and productivity in the agricultural sector in the longer term. The possible measures that can be undertaken to achieve this strategy include, with the limited resources that the farmers have, it could be rational to cultivate smaller pieces of land as in the case of the agricultural production theory (Stage II of the production curve) by applying appropriate conservation measures to improve soil quality and get better turn out.

3. The effect of education on households food security confirms the significant role of the variable in consideration for betterment of living condition. Farmers education helps for improving production and productivity of agriculture. The more households 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. In order to bring food security at the households level the development strategy need to encompass education programmes to the smallholders. So, strengthening both formal and informal education and farmers training centers should be promoted to increase food security in Kamba woreda.

4. The research findings showed that livestock ownership has positive impact on households food security. Livestock sub sector plays a great role in the struggle to eliminate food security problem. Its contribution to the households food energy requirement and total income is significant. In the study area animal disease, existence of local and low productive livestocks and marketing conditions are very crucial. Hence, necessary effort should be made to improve the production and productivity of the sector. This can be done through the provision of adequate veterinary services, improved water supply points, introduction of timely and effective artificial insemination services to up-grade the already existing breeds,
88

launching sustainable and effective forage development program, provision of training for the livestock holders on how to improve their production and productivity, improving the marketing conditions, etc

5. The small size of holdings coupled with poor quality of soil and recurrent droughts do not permit the farmers in Kamba to produce enough for the households without an alternative income source. Rural households in the study area have very limited room for generation of income. Hence, for these households to enhance their welfare in general and food security in particular, they must have diversified access to income alternatives to increase their purchasing power. In the face of this, provision of credit must be taken as a measure, though not the only one, to build the capacity of farmers to invest in the agricultural sector, such as purchase of fertilizer, pesticides, improved seed, live and productive animals. Moreover, development strategies should be able to identify income alternatives other than agriculture. In light of this, nongovernmental organizations that are focusing only on agriculture should also channel their scarce resources to creation of income generating activities, trading, crafting, etc. which would greatly help in strengthening off-farm activities which would enable the households to secure their food through purchase.

6. An increase in per capita aggregate production means that the probabilities of households being food secure increases. Even if farmers experience and interest to use chemical fertilizers and improved seeds need effective awareness creation, yield improvements were feasible through the increased use of chemical fertilizers. Therefore, chemical fertilizers supply, price and application of them have to be adapted to support this possibility. Government, farmer groups or organizations and input suppliers are therefore called to provide agricultural inputs to farming households in communal areas at affordable prices to enable them to increase production and application.

89

7. One area of intervention hypothesized to improve the state of food security at households level is promoting the production of cash crops (coffee and fruits). This implies that efforts has to be made to improve income from cash crops production to ensure food security through promoting and developing small scale and traditional irrigation programs which intern reduce rainfall dependability and enhance the level of households food security. 8. The fact that the climate of the district is dominantly semi-arid and the existence of a serious problem of frequent crop failure caused by drought and erratic rains clearly suggest that one of the intervention options is promoting and increasing crop diversification to reduce crop failure that was happen due to rainfall shortage. Hence, seeds of different variety resistance to moisture stress, pests and adaptability that can increase productivity and yield of crops should be introduced.

9. Food aid is an assistance that comes from outside the system to support and address the immediate food crisis. This is necessary from the viewpoint of saving lives. However, it is not designed to give a lasting solution. Rather it destroys traditional coping mechanisms of the society and puts it in a perpetual position to expect aid whenever there is food shortage. This prompts the society to perpetually depend on donors but, the link with the employment generating system would help both in reducing dependency syndrome and contributing to local development.

10. The study suggested also that absence of market and marketing infrastructure facilities are one of the problems of food security. Farmers close to major roads and market centers are better encouraged to intensify production for their own consumption as well as to produce high value crops for sale. Therefore, in order to solve the problem of farmers in remote areas, attention should be given to the impact of factors like poor marketing infrastructure and transport facilities. In addition, government can improve rural infrastructure to boost households income through the provision of households water, electricity and telecommunications. This could increase the possibility for farmers to get right price.
90

References
Abebaw, S. and Ayalneh, B. (2007),Dimensions of food insecurity and livelihood strategies among rural households s in Dire Dawa, Eastern Ethiopia, Tropical Science Abebaw Shimelese (2003)-Determinants of food insecurity among rural households in Dire Dawa,eastern Ethiopia A. Sen (1981) Poverty and Famines: An Essay on Entitlement and Deprivation. Clarendon Press: Oxford Andersen (1997) 2020 Vision for Food, Agriculture, and the Environment in South Africa, In: Lawrence Haddad (ed.). Achieving Food Security in South Africa New Challenges, New Opportunities, IFPRI. Washington D. C Ayalneh Bogale (2002). Land Degradation, Impoverishment and Livelihood Strategies of Rural households in Ethiopia: Farmers' Perceptions and Policy Implication. Vol. 8. Shaker Verlag, Germany. Bartfeld, J., Dunifon, R. (2006). State-Level Predictors of Food Insecurity among Households with Children, Journal of Policy Analysis and Management

Barly riley (2002) the impact of food aid on food security in Ethiopia
Central Statistical Authority (CSA) (1995), Agricultural Sample Survey Report on Area, Production and Yield of Major crops for Private Holdings, Addis Ababa Central Intelligence Agency (2008) United States Government. World Fact Book: India retrievedfrom<https://www.cia.gov/library/publications/the-worldfactbook/print/in.html> (10 April. Devereux, S, andMaxwell, S. (Eds) (2001), Food security in sub-Saharan Africa London, UK: Intermediate Technology Development Group Publishing EHNRI Food composition table for use in Ethiopia part III Ethiopia food security outlook October (2010 to March 2011) Environmental Degradation, Arid Lands Newsletter Vol- 30 Office of Arid Lands Studies, University of Arizona
91

FEWS Ethiopia: Special report on SNNPR Food and Agricultural Organization of the United Nations (FAO) (2008), The State of Food Insecurity in the World, Rome.Food and Agriculture Organization of the United Nation (2003). Trade Reforms andFood Security: Conceptualizing the Linkages, Rome FAO: (1996) World Food Summit, Corporate Document Repository, FAO, Rome.Food and Agriculture Organization of the United Nation (2002).Food Insecurity, Poverty and Agriculture: ECA Working paper. FAO (1991), households Food Security and Forestry: An Analysis of Socio- Economic Issues. Rome: FAO. Frankenberger, T. R. and D. M. Goldestein. (1990),Food Security, Coping Strategies Frankenberger, T.( 1992) Indicators and Data Collection Methods for Assessing households Food Security, households Food Security: Concepts, Indicators, and Measurements, A Technical Review ,Rome, IFAD/UNICEF. Gebre-Selassie, Samuel (2005). Poverty and Food Security in Ethiopia: Some Evidences from Wollo. In Ethiopian Economic Association (2005): Proceedings of the Second International Conference on the Ethiopian Economy, Vol. II, EEA, Addis Ababa. Getachew Deriba (1991). A Case Study of Habro: Social and Demographic Characteristics. Getachew Deriba (1995). Ethiopia at the Crossroads: Famine and Rural Food Security in Rural Ethiopia. Addis Ababa, Ethiopia. Getachew Deriba (1993). A Case Study of Adama Boset: Resource and Food Security. Getachew Deriba (1995). A Case Study of Kembata Hadya: Housholds and Their Resouces. Getachew Deriba (1995). Economy at the Crossroads: Famine and Food Security in Rural Ethiopia. Gezahegn, K. (1995),Agricultural Marketing Policies and Food Security in Ethiopia. In: Mulat, D.; Wolday, A.; Enui, S. And Tesfaye, Z. (ed.): 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, Ethiopia
92

Gujirati, Domadar N. (1995) Basic Econometrics,Third Edition. MacGraw- Hill, New York Gujirati, Domadar. N. (1995) Basic Econometrics, Fourth Edition. MacGraw- Hill, New York Habtewold, Debebe (2001). Food Aid and Its Impact on Peasant Livelihood strategies,In Forum for Social Studies (2001),Food Security and Sustainable Livelihood in Ethiopia, Proceedings of the Symposium of the Forum for Social Studies, Image Printing Press, Addis Ababa. Hoddinott, J. (ed). (2001). Method for Rural Development Projects, Food Security in Practice, Vol., 1., IFPRI, Washington D.C Holmboe-Ottesen G, Wandel M (1990) households food security: application of a normative concept in research for planning of sustainable development. Forum for tviklingsstudier IFPRI (2008) International Food Policy Research Institute, High Food Prices: The What, Who, and How of Proposed Policy Actions, Policy Brief, May.
IFPRI (2009)-International food policy research institute- Global Hunger Index

IFPRI (2002) Living in the City. Challenges and Options for the Urban Poor, IFPRI Issue BriefNo.9 Washington, D.C IMF (2004) the Federal Democratic Republic of Ethiopia: Poverty Reduction Strategy Paper Annual Progress Report. John Hoddinott (2008) chosing outcome indicators of households food security technical guide number7,J. Hoddinott, Choosing Outcome Indicators of households Food Security. International Food Policy Research Institute: Washington, D.C. (1999). Retrieved from < www.ifpri.org/pubs/pubs_menu.asp> (24 January Kidane, H., Alemu, ZG., Khundhlande, G. (2005). Causes of households Food Insecurity in Koredegaga Peasant Association, Oromiya zone, Ethiopia K. Bagachi & S. Puri, (2002) Food and Nutrition Security and Women. In, Women and Food Security: The Role of Panchayats, Edited by P. Chaturvedi. Concept Publishing Company: New Delhi

93

Maddala, G.S. (1989) Limited Dependent and Qualitative Variable in Econometrics, Cambridge University Press, New York Markos, E., (1997) Demographic responses to ecological degradation and food insecurity in drought prone areas of northern Ethiopia,Thesis Publishers, Prinseneiland 305, 1013 LP Amsterdam the Netherlands Marek T (1992). Ending malnutrition: Why increasing income is not enough, World Bank Africa Technical Department, Technical Working Paper No. 5, Washington, DC: World Bank, October. MEDaC (1999) Poverty Situation in Ethiopia,Welfare Monitoring Unit, Addis Ababa Misery mpuzu sikwela (2008) determinants of households food security in the semi-arid areas of zimbabwe: a case study of irrigation and non-irrigation farmers in lupane and hwange districts MoFED(2006) Ministry of Finance and Economic Development,Ethiopia: Sustainable Development and Poverty Reduction Program. Addis Ababa.

2002.MoFED,Ministry of Finance and Economic Development. Ethiopia: Building on Progress - A Plan for Accelerated and Sustained Development to End Poverty (PASDEP) (2005/06-2009/10) Volume I: Main Text, Addis Ababa Mulugeta Tefera (2002) Determinats of households food security in Eastern Oromia,Ethiopia: The case of Boke district of western Hararghe Zone Reginal baseline survey (2005) Reginal baseline survey (2009) Rome Declaration (1996) on World Food Security R. Gardiner (2002) Towards Earth Summit 2002 Project: Briefing Paper. Earth Summit: London SAH RP (2002).Sustainable Livelihoods and Poverty Alleviation: Case Study in Improving Food Security and Livelihood of Mountain People through Development of Agriculture. A paper presented at the International Seminar on Mountains (ISM), Kathmandu, Agriculture Research Station, Lumle, Kaski, Nepal.
94

Shiferaw T. Feleke,Richard L. Kilmer(2005)Determinants of food security in Southern Ethiopia at the households level. S. Maxwell & M. Smith (1992) households food security: A conceptual review, In households Food Security: Concepts, Indicators, Measurements, edited by S. Maxwell and T. Frankenberger. United Nations Childrens FundInternational Fund for Agricultural Development: New York S. Maxwell and T. Frankenberger (1992) households Food Security: Concepts, Indicators, and Measurements. UNICEF-IFAD: New York Staatz JM, DAgostino (1990) VC, Sundberg S. Measuring food security in Africa: conceptual, empirical, and policy issues. Am J Agric Econ Stephen Ds (2000) Food Insecurity in Ethiopia: A discussion paper for DFID, IDS Sussex Tesfaye Kumbi (2005):- households food insecurity in Dodota-Sire district,Arsi zone:coping strategies and policy options. RDFS Section (2009) the Rural Development and Food Security UN-(2010)-The "Development Strategies that Work" database World Bank (1986) Poverty and hunger: issues and options for food security in developing countries. World Bank Policy Study, xi, 69. Washington, DC: World Bank Yilma Muluken (2005):- Measuring rural households food security status and its determinats in the Benishangul Gumuz region, Ethiopia:The case of Assosa woreda Yohannes Mekonnen (2002) Food security problems and Public Actions in Ethiopia, in Gebrehiwot Ageba et al (eds) 2002, pp 345-365

95

Appendices Appendix 1:- Conversion Factors to Estimate Tropical Livestock Unit equivalents Livestock type TLU (Tropical livestock unit) 1-Calf 2-Heifer 3-Cows/Oxen 4-Horse/Mule 5-Donkey 6-Donkey (young) 7-Sheep/Goat 8-Sheep/Goat (young) 9-Chicken Source: Storck, et at. (1991) Appendix 2:- Conversion Factors Used to Compute Adult-Equivalent (AE)
Age Category (Year) 0---1 1.12 2.13 3.15 5.17 7.110 10.112 12.114 14.116 16.118 18.130 30.160 > 60 Male 0.33 0.46 0.54 0.62 0.74 0.84 0.88 0.96 1.06 1.14 1.04 1.00 0.84 Female 0.33 0.46 0.54 0.62 0.70 0.72 0.78 0.84 0.86 0.86 0.80 0.82 0.74

0.20 0.75 1.00 1.10 0.70 0.35 0.13 0.06 0.013

Source- Source: Storck, et at. (1991)


96

Appendix 3:- Calorie value of food items consumed by sample households Unit Kcal Unit Kcal Food item Food item Boqqollo Duket Boqqollo nifro Boqqollo Kolo Sindye Duket Sindye kolo Nifro ye Sindye Gebs Beso Gebs Kolo Teff Duket Mashila Duket Enset Kocho Bula Irish potato Cassava Yam Taro Sweet Potato Adengwarrye Ater suro Kg Kg Kg Kg Kg Kg Kg Kg Kg Kg Kg Kg Kg Kg Kg Kg Kg Kg Kg Kg 3782 1701 3872 3623 3916 2113 3680 3558 3589 3774 1960 2111 1805 897 1240 850 1083 1342 1703 3522 Baqqella kolo Baqqella nifro Tiqill Gommen Abesha Gommen Qeyy Shinkurt Nech Shinkurt Orange Papaya Muze Mango Yebere siga Yefiyel siga Yebeg Siga Milk Coffee+milk Egg Sugar Shenqura Ageda Qibye Zeyt Kg Kg Kg Kg Kg Kg Kg Kg Kg Kg Kg Kg Kg Lt Lt No Kg Kg Kg Lt 2759 1495 237 401 3761 1383 339 349 878 438 1774 2000 1529 737 180 1529 3850 953 7604 8964

Source-EHNRI Part III (1968-1997)-Food consumption table for use in Ethiopia

97

Appendix4. Summary of the Survey Questionnaire General Information Date of Interview, Name of the enumerator, Name of Keble, Signature, Start and end time PART I. Households roster Name, sex, age, educational level of the households members, marital status, main current activity of the households members, religion SECTION 2: households food consumption What type and quantity of foods were prepared for consumption in this households in the last seven days? We would like to ask you some questions about food consumption in this households in the last seven days. These questions concern to the quantity of foods prepared for consumption Food type Local Unit Local unit estimated to kg Conversion into kilos Adjustment for processing Number of calories available for consumption

Beqqollo Duket Listing all foods commonly eaten in the area Section 3- Food consumption habit How many meals did the adults in your HH ate yesterday? (Adult, above 15 years)? How many meals did the children in your HH ate yesterday (children, under 15 years) What has been the trend of food availability for your households (family) during the last 5 years? Food item Did your house hold consume this food item in the past 7 days Yes---1 No-----0 Wheat Maize Irish potato Sweet potato 98 Number of days the food item was eaten in the last 7 days How much in total did your HHconsume (in kg) or local nit) How much came from purchase How much came from own production How much came from gift

Barely Section 4- Land use information Does the households own agricultural land? ------------If Yes----1, if No-------0 What is the total area of the agricultural land? ------------Hectare----------- Timade----------Other Do you think that your piece of land is enough to support your family? ------ Yes----1, no-----0 What is the quality of your Soil relative to other farmers? ------------If Good-----1 if Poor------0 Section 5- Crop production trend What major crops are produced using your land? (Order by priority and large area allocation) Rank of crops Name of the crop Area allocated in timade 1st 2nd Section 6-Irregation application Do you have irrigation land? ................if Yes---1, if No-----0 If yes how much is it? ------------Hectare or--------------Timade-------------Others (specify). What crops are produced using irrigation? Section 7-Crop production Is your own production enough for your households? -----------If Yes------1, if No--------0 If yes, for how many months do you consume your own production during normal production? For how long can you feed your family financially after now in the future? If you cant produce enough food for your family what are the problems you encountered? Problems encountered Small land holding In adequate rainfall Do you produce both in belg and meher seasons? ----------ifYes------------1,if No-------------0 Which season production gives you the highest product? -------- Belg--------1,Meher------- 0 If yes, what are your major crops you cultivate? Type of annual major Area allocated in In Belg season In meher season crop timade Very serious Serious Less serious Not a problem

If the answer for question is No, what are possible reasons? Possible reasons No enough rainfall 99

Rank the reasons

We fear crop failure A crop does not give enough output Did you plant in your farmland drought resistant crops? ------- If Yes--------1, if No---------0 If yes what are these and their area allocated? Type of drought resistant crop Yes No Area allocated in timade Inset Cassava Sorghum Sweet potato Averagely how many times do you cultivate (weed) your seasonal crops after sawing? Type of annual major crops Number of weeding until harvest Maize Wheat Section 8-Forestry soil and water conservation Is your farm prone (lying face down) to erosion? -----------If Yes-----1, if No-----0 What portion of your farmland is affected by erosion? Portion affected by erosion none All half More than half Has erosion affected your farm severely before? -----------If Yes----1 if No------0 How do you see the level of erosion on your farming plots since you started farming? Level of erosion Yes No Very sever Severe Mino No problem How serious is the decline in soil fertility, on your main plot, since you started farming? 100 Yes No

decline in soil fertility Very severe Severe Minor No problem

Yes

No

Do you think that soil erosion will affect your farm in the future if situations remain Unchanged? Is there any soil and water conservation structure done on your farm land? --- if Yes---1,if No--- If yes, is that done by your own or other body? Who done the structure By you By Government body By NGOS Section 9- Use of modern agricultural inputs Do you use improved seed and chemical fertilizer for your agriculture?- --- If Yes---1, if No0 If yes please tell me the amount? no Type of crop Area cultivated in timade Amount of chemical fertilizer used DAP 1 2 Maize wheat UREA Improved seed used Local in kg fertilizer(compost)used in Kg Yes NO

If you cant apply improved seed what is your main reasons? Reasons Price is high for it dont have enough money to purchase it If you cant apply chemical fertilizer what is your main reasons? Reasons Price is high for them I dont have enough money to buy it Section 10-Agricultural extension services 101

Rank

Rank

Has your households received extension service from any government? --- if Yes ---1. If No-----0 Are there agricultural development agents (DA) in your Peasant Association? - if Yes --1. if No-Has the development agent visited your farm during the year 2002/2003 E.C--if Yes --1. If No--0 Do you use oxen for your farm operation? --------If Yes-----1, if No-----0 If yes, how many oxens do you have in number that helps to plow? ----------------------If yes, to above question how many times? Number of visit Yes No One time Two times What was the purpose of the visit? purpose of the visit To give advice on crop production To give advice on animal production Section11-Access to various services Type of service Yes No

Unit of measurement minute

Total service unit

How far is the market that you can buy and sell food from your km home

What other means do you use to fulfill the food requirements of your family Other than production? Means used to fulfill the food gap yes NO Food purchase Food aid Section 12-Marketing and credit Did your family buy any food from the market this year? ---------- -if Yes----------1ifNo---------0 If yes in which months do you buy more usually, please tell me the months? If you are buying food items from markets, is the price increasing or decreasing? _________ Approximately how much did you spent to buy food items last six months? ------------birr Which food item do you purchase mostly from the market? Type of crop bought Cereals(maize, wheat, barley, sorghum, teff) Root and tubers (cassava, yam, taro, enset) What is your means for transporting goods? Means of transportation Yes pack animals human Is there any credit facility /institute in this area /woreda?------------if Yes-----1, ifNo-----0 102 No

yes

No

This years, have you taken out a loan?-----------if Yes ----1;if No ---- 0 Are you ready to borrow money from credit institutes?----------if Yes----1, if No------0 If no what are the possible answers? Possible reasons Yes I dont have plane what to do I fear any government intervened credits

No

Do you have saving habit? ----- If Yes----1, ifNo---0 if yes amount saved last year ___ birr. Is your family participating in safety net programme? ---------If Yes-----1, if No------0 Is your family included in any food security programme? -------if Yes---1, if No---0 If yes, in which programme? ----------------------------------------How do you like to sustain your family in the future? Means to sustain Cultivation Livestock production Engage in other off- farm activities Are you happy living in this agro-ecology? -------------If Yes---1, if No----0 If not what are the main reason that makes you not happy? reasons Having small land holding decreased soil fertility and low production existence of disease Do you participate always in all public meetings? ---------If yes-----1 if No-----0 Does your family practice in family planning? ----------If Yes------1, if No------0 If yes which family planning methods do you use? Type of family planning Pills Injection Condom rank yes NO

Yes

No

103

If you do not use family planning, reasons are Main reasons Causes health problem

Yes

No

Section13- Livestock production and management Does your family have livestocks? ---------If yes-----1, if No-------0. If yes please give the list. Type of livestock Number currently owned Current market estimate Number not owned but cared for(adera yetkemet) Number owned but away ithothers(adera yetesete) Total owned Total market estimate

Bulls/oxen Young bulls What is area of land allocated for animal feed in Timade? ----------------Which animal disease repeatedly occurs in your area? Rank 1st 2nd Did you spend for veterinary services during the last 12 months? -----If Yes----1, if No-----0 If yes how much was the total cost for veterinary services? -------------------------------birr Is there any animal death in last 6 months from your home? ---------If Yes----1, if No-----0 If yes how many and estimated price in market? Number of animals dead---------Estimated market price in birr-----What is the main mode of feeding for your livestock? Mode of feeding Open grazing Intensive lot feeding What is the main source of feed for your live stokes? Main source of feed Open communal range land Private Grazing land Did you purchase feed for your livestock during the last 12 months? ----if Yes---1,if No---0 Section14 households income Source of income crop sale 104 Unit quantity Unit price Total sale in birr yes No yes No Name of animal disease Time of occurrence in months

crop sale Animal sale Section 15-Off- farm income Has anyone from your households migrated in the past year? -------if Yes...... 1, if No.........0 if yes where did they go? Place of migration yes No Within woreda Outside woreda. In which season did they go mostly? Time of migration Kola harvest season Does your family have any job other than agriculture that increase households income? If yes what are they? Type of activity Hand craft (pottery, Blacksmith ,fencing) Trade in livestock or livestock products Trade in grain and others Selling beverage Wage work Would you please tell total family off-farm income? ---------------birr Section 16 -Copping mechanisms How do your households used to cope during crop failures? Type of cooping Sale of livestock Reduce the number of meals Wage employment Eating wild food Yes No Rank Yes No yes No

105

STATA-10 OUTPUTS
Logistic regression Log pseudolikelihood = -8.8156235 Robust Std. Err. 1 .0 609 32 2 .7 053 32 . 37 287 92 .0 007 82 . 26 026 81 . 01 805 83 1 .7 600 57 . 04 392 36 1 .8 840 22 2 .0 541 63 . 00 063 02 2 .2 808 12 . 37 905 91 4 .5 801 77 Number of obs Wald chi2(13) 1 Prob > chi2 Pseudo R2 = = = = 20 0 53 .3 3 0.0 00 0 0.9 34 3

foodsecuri~c familysize farmsize livestocko~u totalofffa~e educationo~d amountoffo~d technologi~d accesstoro~m participa~1n extensions~t saving landqulity MONTH _cons

Coef. -5 .4 22 71 9 9 .9 12 29 4 1 .6 25 76 3 . 00 24 79 1 . 51 18 66 6 .0 08 01 3 6 .4 75 72 6 - .0 75 28 6 7 .1 83 14 2 5 .2 66 64 6 . 00 10 17 4 8 .3 21 27 2 . 81 68 52 2 -1 1. 50 38 7

z -5. 11 3. 66 4. 36 3. 17 1. 97 0. 44 3. 68 -1. 71 3. 81 2. 56 1. 61 3. 65 2. 15 -2. 51

P>|z| 0. 00 0 0. 00 0 0. 00 0 0. 00 2 0. 04 9 0. 65 7 0. 00 0 0. 08 7 0. 00 0 0. 01 0 0. 10 6 0. 00 0 0. 03 1 0. 01 2

[95% Conf. Interval] -7 .5 02 107 4 .6 09 941 . 89 49 332 . 00 09 464 . 00 17 505 -. 02 73 806 3 .0 26 077 -. 16 13 746 3 .4 90 528 1 .2 40 561 -. 00 02 177 3 .8 50 962 . 07 39 099 -2 0. 48 086 -3 .34 33 3 15 .21 46 5 2. 356 59 3 .0 040 11 8 1. 021 98 3 .0 434 06 5 9. 925 37 4 .0 108 02 7 10 .87 57 6 9. 292 73 1 .0 022 52 5 12 .79 15 8 1. 559 79 4 - 2. 526 89 2

Note: 61 failures and 36 successes completely determined.

VIF test output


. vif Variable livestocko~u farmsize participa~1n familysize accesstoro~m numberofmo~h landqulity saving totalofffa~e technologi~d educationo~d extensions~t amountoffo~d Mean VIF VIF 2 .0 1 1 .8 7 1 .5 6 1 .4 2 1 .4 2 1 .2 9 1 .2 8 1 .2 7 1 .2 6 1 .2 0 1 .1 7 1 .1 6 1 .1 0 1 .3 8 1/VIF 0. 49 66 86 0. 53 59 51 0. 64 18 43 0. 70 24 81 0. 70 48 36 0. 77 81 59 0. 78 41 76 0. 78 67 03 0. 79 05 27 0. 83 67 56 0. 85 58 15 0. 86 12 15 0. 90 91 94

106

Contingency coefficient

landqu~y extens~t parti~1n techno~d foodse~c landqulity extensions~t participa~1n technologi~d foodsecuri~c 1.0 000 -0.0 759 -0.1 437 0.0 750 -0.0 059 1 .00 00 0 .08 57 0 .15 14 -0 .02 93

1. 0000 0. 2321 1.00 00 0. 5833 0.38 60 1. 000 0

Odds Ratio
Logistic regression Log pseudolikelihood = -8.8156235 Robust Std. Err. . 00 4 6 84 2 5 45 8 5 .2 5 1 .8 9 5 08 2 .0 0 0 78 4 .4 3 4 23 2 . 01 8 2 03 5 1 14 2 . 61 2 . 04 0 7 38 2 2 48 1 . 33 2 3 98 . 0 24 8 . 00 0 6 30 8 9 37 5 . 01 4 . 85 7 9 48 8 Number of obs Wald chi2(13) 1 Prob > chi2 Pseudo R2 = = = = 200 5 3. 3 3 0 . 00 0 0 0 . 93 4 3

foodsecuri~c familysize farmsize livestocko~u totalofffa~e educationo~d amountoffo~d technologi~d accesstoro~m participa~1n extensions~t saving landqulity MONTH

Odds Ratio . 0 04 4 15 1 2 0 17 6 .9 1 5 . 08 2 29 5 1 . 00 2 48 2 1 . 66 8 40 3 1 . 00 8 04 5 6 4 9. 1 90 1 . 9 27 4 78 2 1 31 7 .0 4 1 93 . 76 5 1 . 00 1 01 8 4 1 10 . 38 5 2 . 26 3 36 4

z - 5 . 11 3 . 66 4 . 36 3 . 17 1 . 97 0 . 44 3 . 68 - 1 . 71 3 . 81 2 . 56 1 . 61 3 . 65 2 . 15

P>|z| 0 . 00 0 0 . 00 0 0 . 00 0 0 . 00 2 0 . 04 9 0 . 65 7 0 . 00 0 0 . 08 7 0 . 00 0 0 . 01 0 0 . 10 6 0 . 00 0 0 . 03 1

[95% Conf. Interval] . 0 00 5 5 19 1 0 0. 4 7 83 2 . 44 7 1 72 1 . 00 0 9 47 1 . 00 1 7 52 . 9 72 9 9 09 2 0. 6 1 62 . 8 50 9 7 32 3 2 .8 0 3 26 3 . 45 7 5 51 . 9 99 7 8 23 4 7 .0 3 8 32 1 .0 7 6 71 . 03 5 31 9 2 40 5 17 0 1 1 0. 5 54 9 3 1. 0 04 0 2 2 .7 7 86 9 9 1 .0 4 43 6 2 2 04 4 2. 5 6 1 .0 1 08 6 1 5 28 7 8. 7 5 10 8 58 . 8 1 .0 0 22 5 5 3 59 1 80 . 8 4 .7 5 78 4 3

Note: 61 failures and 36 successes completely determined.

107

Marginal effect table


Marginal effects after logistic y = Pr(foodsecuritystatusesecured1insec) (predict) = .0060 565 variable familysize farmsize livestockownedbytlu totalofffarmincome educationofhhhead amountoffoodaidrecived technologicaladoptionusedfertand* accesstoroadeinkm participatininpublicmeetingyes1n* extensionservicegotservice1donot* saving landqulity* MONTH dy/dx -.0326 438 .0596 702 .0097 868 .0000 149 .0030 813 .0000 482 .0475 327 -.0004 532 .1004 428 .010 985 6.12e -06 .8471 405 .0049 173 X 5.97 1.19118 4.74443 617.3 2.34 43.4 .675 18.9825 .595 .885 573.875 .18 2.72

(*) dy/dx is for discrete change of dummy variable from 0 to 1

108

Вам также может понравиться