Acting Now for Food Security and Resilient Food Systems: Ethiopia Baseline Report
Most target households in Meskan and Mareko do not have enough food all year: treatment households averaged 8.7 months with enough food. Low literacy and low phone access mean training needs to happen in the field rather than through mobile channels.
Overview
Solidaridad's PMEL team carried out this internal baseline in July 2023 for the Acting Now programme (2023-2025) in Meskan and Mareko districts, around Butajira in SNNP region. The programme targets 20,000 smallholder farmers, with a focus on women and youth, and 13 service providers, and aims to increase food production, incomes and resilience. The baseline sets starting values for the programme KPIs: food sufficiency, income, yield, access to services, training and service-provider support. Treatment households in 22 project kebeles are compared with households in 6 non-project kebeles, and results are broken down by sex.
Key findings
- The average MAHFP was 9.2 months across 730 farmers (treatment 8.7, control 10.6). 10% of treatment households and 43% of control households had enough food in all 12 months.
- On the household hunger scale, 33% of farmers faced little to no hunger, 63% moderate hunger and 3% severe hunger (treatment group: 28%, 69% and 3%).
- 64% of treatment and 63% of control respondents could not read and write, with a significantly higher share among women. 66% of treatment respondents had no access to a phone, and only 2% (treatment) and 5% (control) had mobile internet.
- Maize was the primary crop for about 90% of farmers. Average cultivated land was 1.2 acres in the treatment group and 1.0 acre in the control group.
- Of crop responses, 38% reported lower farm income than the year before, 40% no change and 22% an increase. 43% reported lower yields, mainly because of drought, pests and diseases, and lack of inputs.
- Mean annual production costs were ETB 3,138 per farmer, with fertiliser the largest item (mean ETB 1,592).
- 53% of treatment farmers and 15% of control farmers had received a service in the past year, mostly education and training (41% of services received). 36% (treatment) and 51% (control) belonged to a producer organisation. All 11 service providers interviewed were owned by men.
In figures
10 % of households
43 % of households
Data table
| Item | % of households |
|---|---|
| Project households | 10 |
| Comparison households | 43 |
64 % of project respondents
66 % of project respondents
2 % of project respondents
Data table
| Item | % of project respondents |
|---|---|
| Cannot read and write | 64 |
| No access to a phone | 66 |
| Have mobile internet | 2 |
Recommendations
- Build close working relationships with partners in the project area, such as the Productive Safety Net Project (PSNP), to create synergies.
- Use farmer field days rather than mobile-phone advisory services to share good agricultural practices and new technology.
- Give gender issues much more attention, given the large gaps found.
- Focus interventions on maize, the main crop, together with complementary climate-smart activities.
- Concentrate service support on improved input supply (seeds, fertiliser, agro-chemicals) and aggregation/buying services.
- Work with government to digitise farmer profiling and use the data to link farmers to service providers, markets and extension.
Methodology
Mixed-methods baseline: a household survey of 730-741 smallholder farmers selected by multi-stage purposive and random sampling (project kebeles as treatment, neighbouring non-project kebeles as comparison), interviews with 11 service providers, a transect walk and a review of PSNP documents. Data were collected on KoboToolbox and analysed in Python and R.
Limitations
- The study captures one point in time, and conditions such as climate may change.
- External factors (economic changes, policy shifts, natural disasters) were not accounted for and may affect later comparisons.
- Farmers kept poor or no records of yields and costs, and all data were self-reported, which may introduce bias.
- Consent was hard to obtain because of high expectations or poor past experiences with data collection.
- Service providers could not be split into treatment and control groups, and time was short for data collection, analysis and writing.
