Food security monitoring report in nineteen districts in four regional states of Ethiopia (Tigray, Amhara, Oromia, & South Region)



Rainfall, crop and livestock conditions
Market prices situations
Socio-economic conditions
Nutrition and human health

July 2000
Addis Ababa


The multi agency technical assessment in April 2000 confirmed the failure of Belg crops in Ethiopia this year. As a preliminary estimate (subject to revision after the Belg assessment is completed) the indications are that the total number of relief beneficiaries as of July 2000, will be 10.3 million people. As a result, the resources required will proportionally increase.(Source: DPPC)

The DPPC's estimate of January 2000 also put the food need of WVE operational areas at 73,613 MT for 617,177 affected people. However, the number of affected people and the corresponding food need is expected to grow as the total number of beneficiaries in the country increased.

In response to the drought situation that occurred in Ethiopia this year, WVE launched a relief program for Ethiopia code-named PEG. The program focused on the prevention of famine in WVE ADPs, Emergency Relief food provision for the drought-affected people in Gode Zone and giving Hope for those who have lost their livestock as the result of the drought and the subsequent failure of the Belg rains. WVE planned to distribute food amounting to 44,747 MT, out of which 7,125 MT of food has been distributed to 487,483 people as of June 28, 2000.

WVE is also supporting on a therapeutic-feeding center in a pediatric ward in Gode Hospital with an average in-take of 50 children at a time. So far over 250 children have been treated.

As part of its' overall food security monitoring, WVE conducts two surveys in a year, before and after harvest. WVE conducted its regular Food Security Monitoring Survey in May 2000. The survey covers nineteen districts in four regional states of the country (Tigray, Amhara, Oromia, & South Region).

The sample size was determined using sampling formula.

The survey was conducted in nineteen districts and 25% of the peasant associations in a district were included in the sample and over 700 children were measured using villages as sub cluster. In the seriously affected districts (Saesie Tseda Amba, Tenta, Gera keya, Sodo Zuria and Humbo) every peasant association was included in the survey and 50% of villages were randomly selected for the study. Information on crop, livestock and socio-economic conditions of the areas were collected from the same clusters in the sample peasant associations. Z scores were used to analyze the anthropometric indices.

A total of 16,999 children were sampled for measuring the status of wasting, stunting and underweight in the nineteen districts. Community members were also consulted as key informants in collecting socio-economic, livestock and crop performance information. Secondary data was also collected from line departments.

The levels of wasting in each district are as follows: The highest wasting was reported in Omo Sheleko (23.3%). Saesie Tseda Amba, Kachabira, Kedida Gamilla, Sodo Zuria and Tenta had very high wasting levels (>=15%). Gera Keya, Dawa chefa, Soro and Atsbi womberta had high wasting levels (10-14.9%) . Districts like Antsokia Gemza, Arthuma Jille, Badawacho, Boreda Abaya, Chencha, Kersa Kondaltiti, Adama, Bosset and Humbo, had medium level of wasting (5.0-9.9%) . No districts fall in low level of wasting (<5%) during this survey.

The nutritional status of children in Tenta, Arthuma Jille, Gera Keya, Boreda Abaya, Humbo, Atsbi Womberta, and Saesie Tseda Amba improved as compared to the previous survey results. This was mainly attributed to the relief intervention in the districts.

Other early warning indicators also showed effects similar to the anthropometric analysis. The Belg rain was late by over 8 weeks in WVE operational areas. In most parts of the operational areas it started in the second week of April. The rain was erratic, inadequate and poorly distributed during the season. In some districts it ceased earlier than the normal time.

Consequently, Belg production, which contributes about 40% to the annual crop production failed except for pocket areas in Boreda Abaya, Chencha, Omo Sheleko, Kachabira and Kedida Gamilla districts in Southern Regional State.

Due to late onset of Belg rain the physical condition of the livestock was poor in most of the operational areas except in Boreda Abaya, Atsbi womberta and Tenta districts, which was normal.

Flood disaster was reported in Boreda Abaya district that damaged over 140 hectares of land covered by Belg crops. There were also flood and hailstorm cases in some districts in the Southern Regional State.

Grain prices increased in all districts for the last six months while livestock prices declined as the farmers brought their livestock in great number to markets in order to purchase food grains. Nevertheless, in districts like Tseda Amba, and Atsbi Womberta grain prices declined after the relief food distribution.

Measles epidemic was reported in Dawa Cheffa, Adama and Boset districts. There were also malaria cases in Adama, Boset, Sodo Zuria and Omo Sheleko districts.

In general, consecutive crop failure, depletion of food stock, failure of sweet potato in Southern Regional State, inability of farmers to plant early maturing crops due to late onset of Belg rain, has put most of the operational areas in a critical food shortage situation. Omo sheleko, Kachabira, Kedida Gamilla, Saesie Tseda Amba, Tenta, and Sodo Zuria, districts faced severe food shortage during the survey. Though the degree varies, all other districts fall under transitory food shortage. The food shortage prevailing in the area is expected to continue up to Belg or Meher harvest in the operational areas.


2.1 Methods of data collection

Pre-tested and structured questionnaires were used to collect information on crop, livestock and socio-economic conditions. The information was collected from key informants, which mainly comprises elders, influential individuals and PA leaders. Secondary information was also collected from line offices (Agriculture development and health offices) during the survey.

Though all data collectors were experienced enumerators, Government and ADP staff, appropriate training was conducted on anthropometry/body measurement and other data collection methods prior to the survey.

2.2 Sampling design

About 25% of the peasant associations in a district were included in the sample and over 700 children were measured using villages as sub cluster in each district.

To ensure geographical spread across the survey area, the clusters were first distributed among the agro-ecological zones in proportion to the population size given by the number of households. A systematic random sampling was employed to select sample PAs from a list of all PAs in a district. After the PAs were identified, the list of all villages for each PA was prepared and two clusters were chosen by lottery where anthropometric measurements were taken.

The rest categories of information (crop, livestock and socio-economic conditions) were also collected from the same village/cluster.

The sample size was determined using the following sampling formula; n = Z 2 (P(1-P)D 2

Z = is the error risk and acceptable at 5% (transformed to values 1.96 or 2).
P = is predictive prevalence value depending on latest data. (20% was used)
D = is absolute precision (3% was used)
n = minimum sample size needed

Sampled children were distributed to each sample PA in proportion to their population size (PPS).

In districts such as Saesie Tseda Amba, Tenta, Gera keya, Sodo Zuria and Humbo, every PA was included in the survey and children were measured in about 50% of the PA villages/clusters. These were districts seriously affected by drought and the same method was applied in February 2000 rapid survey for targeting.

2.3 Materials used

A suspended scale of 25kg capacity, graduated by 0.1 kg was used for weighing children 6 to 59 months of age. The reading was taken to the nearest 0.1 kg. The scale was checked after every 10 measurements against 10 or 5Kg for precision. The length of all children was measured lying down on a wooden board having fixed and movable ends. The reading was taken to the nearest 0.5cm. Six to seven enumerators and a supervisor participated in the weight, age and length measurement of a child at a time. Enumerators developed local calendars to know ages of children, in areas where mothers could not tell the exact age of their kids like in Southern region.

2.4 Data processing and statistics

Data entry, editing and analysis were made using EPI INFO version 6.04b and ANTHRO computer software programs. Descriptive statistics like mean and frequency were computed to analyze the data.

Since the calculation of the percent of median doesn't take into account the distribution of the reference population around the median and not consistent across age and height levels, the distribution of indices was expressed in terms of Z-scores. Z-scores have the statistical property of being normally distributed, thus allowing a meaningful average and standard deviation for a population to be calculated. Low Weight-for-Height is considered as an indicator of wasting and is generally associated with failure to gain weight or a loss of weight. Low Height for Age is considered, as an indicator of stunting which is frequently associated with poor overall economic conditions and/or repeated exposure to adverse circumstances. The third index Weight for Age is primarily a combination of Weight for Height and Height for Age and fails to distinguish tall, thin children from short, well-proportioned children. The Z-score cutoff point recommended by WHO to classify low anthropometric levels is 2SD units below the reference median for the three indices for children 6 to 59 months.

The result of this survey could be compared to the following standard prevalence of low anthropometric values (<-2SD). (See table 1). Anthropometric calculations described in this report are based on the growth reference curves recommended by the WHO for international use.

2.5. Limitations of the survey

In South region, majority of the mothers does not know the exact age of their children. Hence, extra time was deployed to develop local calendars with the community and enumerators were trained on how to know the exact ages using local calendars.

In some districts like Tenta, Gera Keya and Soro, data collection was laborious due to inaccessibility of the lowland areas.

(pdf* format)

* Get Adobe Acrobat Viewer (free)