Adoption of Rainwater Management Practice in the Blue Nile A Description and Analysis of the IFPRI Farm Survey on Climate Change Noémie Defourny Ms. in Economics, UCL (Belgium)

Internship Timeline ILRI: Static Bio Physical Household level Model Cikeda (Cirad) - IAT (CSIRO) Solutions Feasability in Boneya ILRI-IWMI: NBDC Data set 2005-IFPRI

Nile Basin Development Challenge Objective: to improve the resilience of rural livelihoods in the Ethiopian highlands through a landscape approach to rainwater management. Water scarcity and land degradation – concerns livelihoods of millions households in Sub-Saharan Africa Water for agriculture – crop production to feed population and Animal 70 to 90% of the all water used in the region. Growing populations  Need - to reverse land degradation - to improve water productivity.  Landscape (watershed) approach to rainwater management To better target or ‘match’ promising technologies (or whole strategies) with particular environments.

Nile Basin Development Challenge Objective : creating feasibility maps for rainwater management strategies that include socio-economic constraints. One approach : Mapping Willingness of Adoption Procedure : Define adoption rules Based on census data (=data for the whole basin), simulate “virtual farmers” Run adoption rules on the simulated farmer N3 : on Targeting and scaling out

Integrating socio economic into feasibility maps Bio-physical suitabilityWillingness to adopt Feasibility map

Objective of the internship Aggregate the IFPRI “Climate Change” survey (phase 1) to farm level Describe the dataset in terms of Water Related and Soil Conservation Practices Compute Variables Run first Adoption Models

IFPRI Farm Survey on Climate Change (2005) General features: Geo-referenced (GPS coordinates) 1,000 households (6,168 individuals) 3 Regions: Fogera, Jeldu and Dapo areas. Gender 51.4% of male 48.6% of female Ethnic 40% Oromo Ethnic group 31% Amhara 15% Tigrayan 15% Beninshangul Gumuz 5.00% from SNPP Religion 86.7% Christian 13% are Islamic.

Descriptive Statistics Household level Characteristics 90.10% household heads male Age: head45 years old spouse 35 years Size< 6 persons Farmer’s experience in agriculture 23 years Education5 years of school

Household characteristics (cont’d) AssetsDrought power: 72.2% own oxen 32.4% donkey 12% own horse LaborLabor intensive: Meher, Livestock, Perennials Own labor: Hired labor: Off-farm jobs: seasonal trends Meher > Belg (trader, paid laborers)

Household characteristics (cont’d) Land Land Total 1.9 ha 3 plots/H, 0.79 Ha Water Source Rainfed 95.26%, river 2% Distance to homestead 1.4 km Certification

Household characteristics (cont’d) Fertility 60% moderately 30% plots highly Use Erosion Exposure

Access to Water Distance 31.7 km Source of Water Pump Pump 2.60 % Hh Type Diesel ; 23% Manual Ownership 69.20% HH jointly Purpose Irrigate the crops; garden

Access to Water (cont’d) Water Storage Water Storage 8.50% Hh Type Ownership Purpose

Access to Advice, Market and Credits Access to Advice Access to Advice Access to Market Access to Market Transportation 93.83% On foot 3.34%. Animal 2.43% Motorized vehicle Crop productionLivestock Activities Training 47.10%53.30% Visit 54 InputOutput Distance Km5.66 km5.70 km Travel Distance8.91 hrs11.32 hrs

Access to Advice, Market and Credits (cont’d) Access to Credit Access to Credit 50.00% have at least borrowed money once. Purpose Source Frequency

Shocks and Aid Aid Aid Reason for aid activity’ implementation Type

Crop level Characteristics Fragmentation Type

Crop level Characteristics (cont’d) Irrigation Fertilizer

Crop level Characteristics (cont’d) Soil Conservation techniques Soil Conservation techniques 74.5% households practices at least one type Type

Perennials Type Purpose

Perennials (cont’d) Irrigation 4.10% Hh Furrow at 48.44%, sprinkle 1 pump Fertilizers Only Manure 4.30% of household 6.20% perennials Share 78.2% perennials - 100% plot 20.81% perennials - ≤50% plot Sell 36.9% of household

Livestock 92.3% of household - 3,576 animals. 17.4% of livestock are lost of disease Type

Livestock (cont’d) Source of Feed: Source of Feed: Source of water: Source of water:

Perception of Climate Change (over last 20 yrs) Variation of Rainfall Declined according to 61.53% Hh Variation of Temperature # of Hot Days, 67.72% Hh Perceived Cause of rainfall variation Poor vegetation cover (78.98% ) Variation of vegetation cover 50% Hh unchanged, for 35% decreasing. Major constraints in changing your farming ways

Perception of Climate Change (cont’d) Adjustments made to LT shifts Adjustments made to LT shifts - in temperature - in rainfalls

Willingness to Adopt Model Specification The framework can be estimated with a multivariate PROBIT estimation. Qualitative depend variable  Probit: linear probability model y= α+ β n,i x n,i + β n+1,i x n+1,i +ε i y= Pr(PumpT) Coefficient Estimators are not BLUE R² is not a good measure of equation performance.  Pseudo-r² (goodness-of fit, maximum loglikelihood)

Results: 1.Adoption of Pump as a Water Management Strategy technology Variables computationExplanatory VariablePUMP ADOPTION Marginal Effects of the Average Household Household Size (persons) HHSIZE Distance to Market for Input (hrs) MARKETINPUT Number of Task for which Labor was Hired HHHIRELABOR Whether the household owns or not at least one oxen OXENTBin The number of visit on crop production and livestock VISITAGG Constant CONSTANT If SlopeFlat >0 Spatial Restriction Pseudo-R² Number of obs. 572 First Adoption Models

2. Adopting Planting tree as a SWC technique Variables computationExplanatory Variable Planting Tree for SWC ADOPTION Marginal Effects of the Average Household Age of the Household Leader HHHEADAGE Religion of the Household Head RELIGION Total ha of land ownedLANDHOLDINGTOTAL Squared of Total ha of land owned LANDHOLDINGTOTAL² The average distance from plot to Homestead AVERAGEDISTANCEPL OTHOMESTEAD Whether the household has received visit on crop production or livestock activities along if they have attended a training on either two focus ACCESSTOADVICE CONSTANT ErosionSevereBin>0 or ErosionMildB>0 Spatial Restriction Pseudo-R²0.104 Number of obs.724

3. Adoption of SC Techniques (Soil bunds, Stone bunds, Grass Stripes and Plouhging contour) Variables computationExplanatory Variable SC techniques ADOPTION Marginal Effects of the Average Household Primary Residence with Metal Roof METALROOF Primary Residence concrete stone/bricks CONCRETEHOUSE Number of PlotHHNUMBERPLOT Number of Task for which Labor was Hired HHHIRELABOR Total ha of land ownedLANDHOLDINGTOTAL Has received visit (crop & livestock) VISITBIN Soil conservation aid -binarySOILCONSERVAIDBIN Distance Market Input (hrs)MARKETINPUT CONSTANT SlopeInclinedBinary>0 or SlopeSteepBinary>0 Spatial Restriction Pseudo-R² Number of obs.442

Results PUMPS More productive farmers, closer to market, hiring labour, owning oxen and being informed. TREE PLANTING Plot near homestead, middle size farmers SC TECHNIQUES Poorer farmers, further from market & more likely to received Aid

Thank you for your Attention & Thank you for welcoming me at ILRI/IWMI!