What are the main drivers behind Ethiopian farmers soil and water

conservations practices?

INTERNSHIP REPORT

2011-2012

Defourny NoémieLocal supervisor: Dr. Pfeifer

ILRI-IWMI (CGIAR), Addis Ababa Research Center May -August 2011

PLAN

1. Host Organization Description2. Nile Basin Development Challenge3. Economical Question:

What are the main drivers behind Ethiopian farmers soil and water conservations practices? - Theoretical framework - Descriptive Statistics- Empirical Models- Estimated Results - Conclusions

HOST ORGANIZATION DESCRIPTIONConsultative Group on International Agriculture Research

(CGIAR, 15 Institutes, 1971)

Public Good Provider Non-rival & Excludable; ExternalitiesEconomies of Scope and Scale

Mission Reducing Hunger And Poverty Improve rural livelihoods through agricultural productivity

Addis Ababa Campus (4 CGIAR Institutes, 12 hosted organization)

Int. Water Management Institute (IWMI) - USD30.09 million

Water for a food-secure world: water availability, access, quality, productive uses - Nile Basin Development Challenge

Int. Livestock Research Institute (ILRI)- USD34.4 million

Better lifes through livestock: genetics, livestock market opportunities, crop residues

- Exploring static biophysical models at household level

- Designing Boneya’s meeting aimed at assessing feasibility of research analysis results

& solutions

Nile Basin Development Challenge (NBDC ,2010-) 5 linked projects toward a core objective: improve the resilience of rural livelihoods in the Ethiopian highlands through a landscape approach to rainwater management.

Project 3 : Targeting and scaling out of rainwater management systems by taking a watershed approach: tailoring a set of best bet practices &

technologies to the environmental needs at landscape level.

Steps : Mapping Willingness of Adoption Creating feasibility maps for rainwater management strategies that include socio-economic constraints.

Internship task (2/3 intern work) in order to define adoption rules :- Cleaning, restructuration, description, analysis of NBDC Data set (IFPRI, 2005) - Run first trials of technology adoption models

Biophysical suitability Willingness to adopt SWC technic

Feasibility map

Integrating socio-economic features into feasibility maps

N1: Learning about rainwater

management system N2: Integrating

technologies, policies and institutions

N3: Targeting and scaling out

of rainwater management

systems

N4: Assessing and anticipating

the consequences of innovation in

rainwater management

systems

N5: Catalyzing platforms for

learning, communication

and coordination. 

Identify rainwater management system working best

Identify where in the basin (Map) Analyze best land use system for differents parts of the basin (water productivity, livelihoods & eco benefits

WHAT ARE THE MAIN DRIVERS BEHIND ETHIOPIAN FARMERS SOIL & WATER CONSERVATION (SWC) PRACTICES?

Theoretical Framework based on Literature Review- Selection of adoption model estimation among 3 methods (probit/logit/bivariate probit)

Adoption decision corresponds to a binary dependent variable j = 1, 0 & i= 1, n

Multivariate decision Limited dependent variable (LDV) model Probit (non-linear probability model)Estimation method choice: MLE due to non-linear feature

Interpretational implications : Unsatisfied BLUE criteria; magnitude R² no good measure model fitnessPseudo-r²

- Determining factors affecting soil conservations decisionNo strong foundation in economic theorySWC technique performances are location-specific

Identification regressors in the light of previous studies in Ethiopia [Amsalu and de Graff (2006); Bekele and Drake (2003)]

Factors categories: Personal, physical, socio-economics, institutional factors (Lapar et al., 1999)

Data Description (IFPRI 2005)Cross-sectional data set in Stata 10.0.Attributs 9 agric. and socio-eco. topics, 10 attributes per topic. Sample size 1,000 households / 6,000 plots/ 6,168 individuals.

5 regions, 20 districts, 13 zones and 20 woredas, 50 households per woreda.

Empirical Models Selection of 3 dependent variables of interest among available SWC technics variables.

Control variables : Biophysical restrictions

Technics Adoption Type Description Freq Mean Std. Dev. Min Max

PUMPTOT Binary Whether HH has a pump or not 1000 .026 .1592 0 1

PUMPWATER Continuous Pump used to water livestock 26 .0769 .2717 0 1

PUMPDOM Continuous For domestic use 26 .0769 .2717 0 1PUMPIRRI Continuous Number plots irrigated by a pump 26 1 .4 0 2

SCT SOILBUND Continuous Number plots with soil bunds 1000 1.385 2.384 0 16SCT STONEBU Continuous Number plots with stone bunds 1000 .663 1.7923 0 18

SCT GRASSSTRI Continuous Number Plots with Grass stripes 1000 .15 .81496 0 10SCT WATERW Continuous Number Plots with Waterway 1000 1.688 2.5288 0 13SCT PLANTIN Continuous Number Plots where Tree are planted 1000 .316 1.0757 0 10SCT PLOUGHI Continuous Number Plots Ploughed 1000 .126 .71037 0 7 SCT NONE Continuous Number Plots without SWC practice 1000 1.512 2.5866 0 15

Adopting of PumpSLOPEFLATBin Binary Whether the HH has at least one flat plot 1000 .818 .3860 0 1

Adopting Planting tree

EROSIONSEVE Binary Whether the HH has at least one plot with severe erosion

1000

.22 .4145 0 1

EROSIONMILD Binary Whether the HH has at least one plot with mild erosion 1000 .652 .4766 0 1

Adoption of SC Techniques SLOPEINCL Binary Whether the HH has at least one plot inclined 1000 .642 .4796 0 1

SLOPESTEEP Binary Whether the HH has at least one steep plot 1000 .117 .3216 0 1

Independent Variables

RESULTSEstimated coefficients and their significance for the 3 SWC adoption Probit Equations

.008 .021

.186

RESULTS ANALYSIS Drivers of SWC techniques adoption are specific on technology type.

Adoption of Pump as a Water management strategy technologyGood market access, wealthier or productive peasant.Large supply of intern labor, owning animal for production process, willing to hire labor

and farmer’s knowledge through DA’s advice.

Adopting Planting tree as a SWC techniqueKnowledge access and labor availability.Plots close to homestead, when smaller farm: stronger concern of securing land,

younger farmer,

Adoption of SC techniques (Soil bunds, Stone bunds, Grass Stripes and Plouhging contour)Less wealthy farmersWithout oxen, less likely to hire labor, smaller land the more likely SWC technics will be

implemented, highly fragmented landholding ,further away to the market.

CONCLUSIONS Results compatible with existing literature review but more specifics allowing to

differentiate adoption between variours SWC pratices. IFPRI 2005 dataset ready for further modeling toward NBCD objectives.

Thank you for your attention!