This study analyzed rural households’ food security and its determinants in western Ethiopia. The study used a primary data collected from 276 randomly selected households using interview schedule. In addition, focus group discussion (FGD) and key informants interview were also used to obtain a qualitative primary data. As to the method of data analysis, the study employed descriptive statistics (such as mean, frequency, range and percentage) and binary logit model (BLM). The finding of the study revealed that, in the study area, 59.06% of the sampled households were food insecure and 40.94% of them were food secure. Besides, the study indicated that 86.87% of the households were vulnerable to different shocks, risks, and seasonality’s and 41.67% of the households faced shortage of food. Moreover, the finding indicated that only 31.88% of the households were food self-sufficient from own production. Low productivity, climate related problems and inadequacy of cultivable land were identified as the top three main causes of food shortage and/or food self-insufficiency from own production. The estimated BLM pointed out that sex, age, access to irrigation, off-farm and non-farm income, input cost, access to credit and distance to market were significant in determining household’s food security status. Therefore, policies and actions directed towards improving households’ food security and reducing their vulnerability should focus on the aforementioned factors.
Food insecurity and poverty are crucial and persistent problems facing the world. The number of people who are food insecure and malnourished globally has been an escalation since 2014, reaching an estimated 815 million in 2016 from 777 million during 2015 affecting 11 percent of the global population [
As part of Sub Saharan Africa, Ethiopia is facing with the problems of poverty and food insecurity. A recent study figured out that about 23 million Ethiopian live under the basic poverty line and food insecurity remains a major challenge [
Particularly, Assosa zone is characterized by the prevalence of high incidence of poverty and food insecurity. The food insecurity situation in the study area was worsened by inadequacy of technology innovation, climate change, weed infestation, insect pests, and poor field management [
The rest of the paper is organized as follows: Section 2 provides the methodology employed; Section 3 presents and discusses the results; and Section 4 concludes and infers policy implications.
Assosa zone, the study area, is one of the three administrative zones of Benishangul-Gumuz Region. Administratively, the study area is composed of 7 districts, namely; Assosa district, Homosha district, Bambasi district, Menge district, Kurmuk district, Sherkole district and Odabildi-Guli district. The total population of the zone was 283,707 people, of which 144,616 and 139,091 were male and female, respectively. Furthermore, 86.28% of the population lives in rural areas [
The study employed three-stage simple random sampling method. In the first stage, three districts (namely Assosa, Bambasi and Sherkole) were randomly selected out of 7 districts in the study area. Secondly, 12 peasant associations (PAs) were randomly selected using probability proportional to the size of PAs in each sampled districts. The reason for selecting PAs was that, in the study area almost all the households relied on agriculture and the emphasis of this study was on assessing the food security of households working on agriculture. Finally, 276 sample household heads were randomly selected based on probability proportional to size of the households in the selected PAs. The selection of sample household heads in each PAs was done using lottery method. The sample size for collecting quantitative data for this research was determined by using Yamane formula [
n = N 1 + N ( e ) 2 = 40530 1 + ( 40530 × 0.06 2 ) = 276 (1)
where; n = denotes the sample size; N = denotes total number of estimated household heads (40,530) in the study area; e = denotes margin of error (6%).
The study utilized primary data. Primary data were collected using interview schedule through enumerators and the researchers. Particularly, primary data on the types and quantities of every food item consumed by the household head and his/her family members was collected using Weighed records method for 7 consecutive days from each sample household. The reason for collecting the data from a single household for seven consecutive days was that food security is a sensitive issue that is affected by different unforeseen factors (religious, holidays, etc.) which can be captured by taking weighed data [
The data analysis for this study was conducted using STATA 13 statistical software package. The study employed descriptive statistics and econometric model to analyze the data. Descriptive statistics such as mean, percentage, frequency and standard deviation were used to assess households’ food security status, vulnerability status, causes of food shortage and to provide insight into different socio-economic characteristics of the households. To analyze determinants of food security, the study estimated binary logit model. To estimate the dependence of dichotomous dependent variable on one or more independent variables, studies employ linear probability model [
P i = E ( Y = 1 X i ) = 1 1 + e − ( α + β X i ) = 1 1 + e − ( Z i ) (2)
For ease of exposition, the logit becomes a linear function of different explanatory variables:
L i = ln [ P i 1 − P i ] = Z i = β 0 + β 1 X 1 + β 2 X 2 + … + β i X i (3)
where: Pi denotes probability of being food secure, 1 − Pi denotes probability of being food insecure, Li is the logit, Xi is vector of relevant household characteristics and βi is a vector of parameters to be estimated. It should be noted that the estimated coefficients do not directly indicate the effect of change in the corresponding explanatory variables on probability (P) of the outcome occurring [
We included different demographic, socio-economic, institutional and other variables in our analysis. The selection of those variables is guided by previous empirical studies. Accordingly, Asfir [
Oxen and donkey serve as a source of traction power in many developing countries, thereby significantly affect household’s crop production and enhance households’ access to food items [
For this study, primary data were collected from 276 sampled household heads. About 44% of the sampled household heads were settlers and the rest (56%) were native households. In addition, 89% of sampled farmers were male-headed households, implying that the majority of the sampled households were male-headed households. Besides, 56.43% of the households had access to irrigation, which indicates that more than half of the households had access to irrigation to produce more than once in a year in the study area. Furthermore, 52.89% of the households had access to training on issues related to agricultural production and agricultural productivity improvement during 2016/17 production year. Moreover, only 18.48% of the households had access to credit services. The result also revealed that only 5.07% of the households received remittance and aid, to cope up with the food insecurity and shortage situation in the study area (
Regarding continuous variables, the result indicated that the mean age of the household heads was 43.88 years, with maximum and minimum being 78 and 23 years, respectively. In addition, the mean year of schooling of the households was 3.08. Furthermore, the finding of the study pointed out that the family size
Dummy variables | Frequency (N) | Percent (%) | |||
---|---|---|---|---|---|
Nature of households settlement (settler) | 121 | 43.84 | |||
Sex of household head (male) | 246 | 89.13 | |||
Access to irrigation (Yes) | 156 | 56.43 | |||
Access to training (yes) | 146 | 52.89 | |||
Access to credit (yes) | 51 | 18.48 | |||
Access to remittance and aid (yes) | 14 | 5.07 | |||
Continuous variables | Mean | Std. Dev. | Minimum | Maximum | |
Age of household head (Years) | 43.88 | 10.97 | 23 | 78 | |
Education status of HH heads (years) | 3.08 | 3.62 | 0 | 13 | |
Family size (Number) | 5.58 | 3.05 | 1 | 20 | |
Dependency ratio | 0.96 | 1.07 | 0 | 5.6 | |
Total cultivated land (Hectare) | 0.957 | 0.78 | 0 | 5 | |
Livestock holding excluding oxen and donkey (TLU) | 1.79 | 1.74 | 0 | 8.35 | |
Oxen and donkey ownership (Number) | 0.58 | 0.92 | 0 | 4 | |
Total input cost (Birr) | 1087.41 | 1237.3 | 0 | 8000 | |
Farm income (Birr) | 5160.21 | 6235.05 | 60 | 39,000 | |
Off-farm & non-farm income (Birr) | 1887.07 | 3739.53 | 0 | 24,096 | |
Extension contact (Number) | 8.92 | 8.17 | 0 | 48 | |
Distance from the nearest market (Km) | 6.84 | 6.01 | 0.1 | 20 | |
Source: Authors computation (2017), N = 276.
of the households’ ranges from 1 to 20 member(s) with mean family size of 5.58 members. Moreover, the mean value of the ratio of inactive to active family members was 0.96 with the maximum and minimum being 5.6 and 0, respectively. It also indicated that, the average total cultivated land size of the households was 0.957 ha. Besides, the result revealed that the mean livestock holding (excluding oxen and donkey) of the households in terms of tropical livestock unit (TLU) was 1.79, with the maximum and minimum being 8.35 and 0 TLU, respectively. The study also confirmed that households’ mean holding of oxen and donkey was 0.58. Farm income of the households ranges from Birr 60 to 39,000 with an average of Birr 5160.21 per annum. In addition, households’ income from off-farm activities ranged from Birr 0 to 24.096 with an average of Birr 1887.07 per annum. Besides, the mean input cost incurred in agricultural production in 2016/17 production year was Birr 1087.41. Moreover, the mean frequency of extension contact with the farmers in the study area was 8.92 times per year, with the maximum and minimum being 48 and 0, respectively. The mean distance of households from the nearest market was about 6.84 kms with the minimum and maximum distance being 0.1 and 20 km, respectively (
In this study, household’s calorie intake per adult per day was used to identify the food secure and food insecure households. Data on the type and quantity of food item consumed by the household for seven consecutive days was collected using weighed records method and was converted to kilocalorie (kcal) and then divided to household size measured in AE and number of days. Following this, the amount of energy utilized in kcal by a household was compared with the minimum subsistence requirement per adult per day (i.e. 2200 kcal per adult per day [
healthy and productive life, and it was attributed to the fact that, the study area is among the drought prone areas in the country as the majority of the community rely on agricultural activities to achieve their livelihood goals. Besides, in the 2016/17 production year there was outbreak of pests and other perennial crop diseases that resulted in loss of thousands of tons of crop outputs. In addition, the study area is characterized by the existence of unfavorable condition for livestock production due to the existence of livestock diseases, which ultimately affected households’ food security.
In the study area, households are and have been vulnerable to different types of risks, shocks and seasonality’s and many studies argue that vulnerability to be associated with food security. Accordingly, this study identified that households in the study area were mainly vulnerable to the occurrence of animal and plant diseases as indicated by 50.72% of the households. The result confirms that almost more than half of the households were exposed to the risk of animal and plant diseases and since farming is the main source of livelihood in the study area, it affects the total agricultural production and ultimately affects their food security situation. Besides, the study pointed out that households were exposed to the risk of members’ health problem as indicated by 14.79% of the households and again, this is also another shock that forces the households to divert their income for treating the members rather than utilizing it for food items production and purchasing. Thirdly, 10.14% of the households in the study area indicated that they were vulnerable to drought and famine, which facilitates their exposure to health problems. Shortage of rain and conflict with neighbors’ were another shocks that the households were exposed. Shortage of rain, as it is a key agricultural input, imposes direct impact on total output obtained and indirectly affects households’ food security through its effect on agricultural production. Generally, out of the total sampled households, 86.87% of the households in the study area were vulnerable to various shocks, risks and seasonality (
Vulnerable to: | Frequency (N) | Percent (%) |
---|---|---|
Drought and Famine | 28 | 10.14 |
Members health related problems | 41 | 14.79 |
Animal and plant diseases | 140 | 50.72 |
Shortage of rain | 16 | 5.79 |
Conflict with neighbors | 15 | 5.43 |
Total | 240 | 86.87 |
Source: Estimated result (2017), N = 276; Note that only 86.87% of the households where vulnerable to different shocks in the study area.
livestock diseases especially cattle diseases, plant diseases and pests, household heads and members health problem specially child illness and occurrence of drought and famine due to shortage of rain which results in decline crop and livestock yield”.
Households in the study area were facing shortage of food, which is governed by different factors. This study identified that, in the study area, 41.67% of the respondents pointed that they faced shortage of food in 2016/17 production year and the remaining households (58.33%) reported that they were able to access sufficient food. In addition, the finding also indicated that only 31.88% of the households were self-sufficient from their own production and the rest (68.12%) were self-insufficient from homestead production (
Accordingly, the study identified different causes of food shortage and food self-insufficiency from own production in the study area (
Food shortage occurrence | Total | Self-sufficiency from own production | Total | |||
---|---|---|---|---|---|---|
Response | Yes | No | Yes | No | ||
Frequency (N) | 115 | 161 | 276 | 88 | 188 | 276 |
Percent (%) | 41.67 | 58.33 | 100 | 31.88 | 68.12 | 100 |
Source: Estimated result (2017), N = 276.
Causes of food shortage and/or food self-insufficiency | Frequency (N) | Percent (%) |
---|---|---|
Low productivity | 56 | 20.29 |
Inadequate income from alternative sources | 19 | 6.88 |
Inadequate input usage | 16 | 5.79 |
lack of proper utilization of income | 5 | 1.81 |
Climate related problems | 37 | 13.41 |
Food price inflation | 6 | 2.17 |
Low soil fertility | 17 | 6.16 |
lack of access to irrigation | 11 | 3.98 |
Inadequate cultivable land | 25 | 9.06 |
Total | 192 | 69.55 |
Source: Estimated result (2017), N = 276; note that only 192 households were facing food shortage and/or food self-insufficiency from own production.
finding indicated that 69.55% of the households faced food shortage and/or self-insufficiency from own production in the study area. Low agricultural production and productivity, which is governed by various factors such as climatic, physical, institutional, etc., factors, was identified as major cause of food shortage and/or food self-insufficiency by 20.29% of the households in the study area. Besides, 13.41% of the households indicated climate related problems such as shortage of rain, occurrence of pests and animal disease, flooding, etc., as the major causes of food shortage and self-insufficiency in the study area. Furthermore, 9.06% of the respondents confirmed inadequacy of cultivable land for agricultural production as a cause of food shortage and/or food self-insufficiency in the study area. Moreover, the study figured out that 6.88%, 6.16%, 5.79%, 3.98%, 2.17% and 1.81% of the households reported inadequacy of income from alternative sources, low level of soil fertility, inadequacy of inputs, lack of access to irrigation, food price inflation and improper utilization of income, respectively, as the major causes of shortage of food and/or food self-insufficiency from own production in the study area.
A binary logit model was estimated to analyze determinants of households’ food security. The model estimates indicated that the overall model is significant at 1% (Prob > chi2 = 0.0001) as shown by the likelihood ratio test. The estimated model also revealed that, out of the 18 explanatory variables, 7 variables were statistically significant in determining households’ food security (
Explanatory variables | Coefficients | Std. dev. | P > |t| | ME (dy/dx) |
---|---|---|---|---|
Settlement of the HH head | −0.26 | 0.349 | 0.457 | −0.0615 |
Sex of the HH head | 0.9515* | 0.532 | 0.074 | 0.1991 |
Age of the HH head | −0.0382** | 0.0139 | 0.014 | −0.00813 |
Education status of the HH head | −0.0382 | 0.311 | 0.902 | −0.00909 |
Family size | −0.069 | 0.0607 | 0.253 | −0.0165 |
Dependency ratio | 0.0272 | 0.147 | 0.854 | 0.00647 |
Livestock holding | 0.1126 | 0.10 | 0.264 | 0.0268 |
Number of Oxen and Donkey owned | 0.143 | 0.182 | 0.432 | 0.0341 |
Cultivated land size | −0.0308 | 0.2449 | 0.90 | −0.0073 |
Access to irrigation | 0.825*** | 0.302 | 0.006 | 0.1963 |
Farm income | 0.0000385 | 0.00003 | 0.183 | 9.40e−06 |
Off-farm and non-farm income | −0.000097** | 0.000044 | 0.027 | −0.000023 |
Input cost | −0.000393** | 0.00017 | 0.023 | −0.0000935 |
Access to training | −0.358 | 0.2978 | 0.229 | −0.0774 |
Frequency of extension contact | 0.0141 | 0.021 | 0.503 | 0.00335 |
Access to credit | 1.276*** | 0.379 | 0.001 | 0.3085 |
---|---|---|---|---|
Access to remittance and aid | −1.105 | 0.758 | 0.145 | −0.2207 |
Distance to market | 0.0541** | 0.0273 | 0.048 | 0.0287 |
Constant | −0.1245 | 0.832 | 0.881 | |
Number of observation | 276 | |||
LR chi2 (18) | 58.50 | |||
Log likelihood | −157.50 | |||
Prob > chi2 | 0.0000 | |||
Pseudo R2 | 0.1566 |
***, ** and * significant at 1%, 5% and 10% probability level, respectively.
Sex of household head: as expected, it was found to have positive and significant effect on households’ food security at 10% significance level. From the model result, the marginal effect showed that being male-headed household increases the probability of households’ food security by 19.91%. This implies that male headed households are more likely to be food secure than female headed households. This is due to the fact that, mostly male headed households have better access to different types of resources, which ultimately enables them to produce, purchase and consume diverse and nutritious products. This finding supports the finding of [
Age of the household head: it affected households’ food security negatively and significantly at 1% probability level. The marginal effect of age of household head indicated that a one-year increase in the age of the household head decreases the likelihood of households’ food security by 0.92%. This implies that old aged household heads more likely to be food insecure than younger ones. This is because mostly elder households have less courage to cultivate larger-size farm and become less productive than young ones, which ultimately affects their food security status through restraining production. Nugusse et al. [
Access to irrigation: it was found to have positive and significant effect on households’ food security at 5% significance level. The marginal effect of the variable figured out that having access to irrigation increases households’ probability of food security by 14.4%. This implies that households who had irrigation access are more likely to be food secure than those who had no irrigation access. This is due to the fact that irrigation helps farmers enhance their agricultural production through mitigating water stress and reducing risks of crop failures and obtains more yields; thereby reducing the risk of food insecurity among the households. This finding is in line with the findings of [
Off-farm and non-farm income: In contrary to the expectation, it affected households’ food security negatively and significantly at 10% significance level. The marginal effect confirmed that a one-birr increase in the off-farm and non-farm income of the households decreases the likelihood of households’ food security by 0.002%. This indicates that households with higher off-farm and non-farm income earning are less likely to food secure than low earning households in the study area. The reason is that households who earn higher off and/or non-farm income do not use their income for either food expenditure or production of consumable products rather they prefer to make a saving to improve their future welfare at the expense of today consumption. This finding is in line with the finding of [
Input cost: it determined households’ food security negatively and significantly at 5% significance level. The marginal effect, from the model estimate, revealed that a one-birr increase in investment on inputs decreases the probability of household’s food security by 0.0083%. This indicates that households with higher input cost are less likely to food secure than low cost incurring households. This is because higher cost investment on inputs forces households to decrease their expenditure on food items and thereby expose households to the risk of food insecurity. This result is in conformity with the findings of [
Access to credit: As expected, it was found to have positive and significant effect on households’ food security at 5% significance level. The marginal effect indicated that having access to credit increases households’ probability of food security by 20.3%. This implies that those households who had access to credit service have more chance of being food secure than without access ones. This is due to the fact that access to credit gives the household an opportunity to be involved in income generating activities so that derived revenue increases financial capacity and purchasing power of the household to escape from risk of food insecurity. Moreover, it helps to smooth consumption when household face with temporary food problem [
Distance to market: In contrary to the expectation, it affected food security status of households positively and significantly at 10% probability level in the study area. From the model output, the marginal effect indicated that a one-kilometer increase in the distance from the nearest market center increases the probability of household’s food security by 1.12%. This implies that households living far from the market center are more likely to be food secure than those living near the market center. This is because households living far from the market center obtain little information about the market condition and thereby use their entire production for home consumption rather than bringing it to the market. The finding supports the finding of [
Food insecurity is and has been a persistent problem facing the majority of the Ethiopian population. Food insecurity in the form of both chronic and transitory (seasonal) is severe in the country. To reverse the situation, various studies recommended that planning different programs based on location specific empirical evidence play a key role. Thus, assessing the food security status and its determinants at household level provides basic input. Accordingly, the study finding indicated that more than half (59%) of the rural households were food insecure. This implies that the incidence of food insecurity in the region was high. Furthermore, the households in the study area were vulnerable to different type of shocks and risks such as drought and famine, illness, animal and plant diseases, etc. Moreover, more than one third of the households faced food shortage in the study area―which calls for action to reverse the situation. Besides, food security status of the households was enhanced by sex of the HH heads, family Size, access to irrigation, total farm income, access to credit and distance to market. Unfortunately, age of HH head, total off-farm and non-farm income, and total input cost negatively affected households’ food security in the area.
Therefore, urgent actions aimed at reducing/eliminating the incidence of food insecurity and vulnerability of the households in the study area should focus on:
The authors are grateful to Assosa University for providing both financial and technical assistance in the research work. We also sincerely thank local communities in our research area, Assosa Zone, and all the enumerators for their valuable efforts. Furthermore, we also thank Assosa university Agricultural Economics staff members for their valuable assistance during the study.
The data that support the findings of this study can be obtained from the authors based on request.
The authors declare that they have no conflict of interests.
Sani, S. and Kemaw, B. (2019) Analysis of Rural Households Food Security in Western Ethiopia. Food and Nutrition Sciences, 10, 249-265. https://doi.org/10.4236/fns.2019.103019
kcal: kilocalorie; AE: Adult Equivalent; FAO: Food and Agriculture Organization; CSA: Central Statistical Agency; WFP: World Food Program; AZBARD: Assosa Zone Bureau of Agriculture and Rural Development; BGRDGA: Benishangul Gumuz Region Development Gap Assessment; PA: Peasant Association; HH: Household.