WOLAITA SODO UNIVERSITY

DIRECTORS OF GRADUATE STUDIES

IMPACT OF CLIMATE CHANGE ADAPTATION STRATEGIES
ON HOUSEHOLDS FOOD SECURITY THE CASE OF BENNA
TSEMAYE DISTRICT, SOUTH OMO ZONE, ETHIOPIA

BY
LEMMA BALI

Advisor: Zemach Lemecha (Ass. professor)

FEBRUARY, 2024
WOLAITA SODO, ETHIOPIA
Outlines of presentation
 INTRODUCTION

 METHODOLOGY

 RESULT AND DISCUSSIONS

 SUMMARY, CONCLUSION AND RECOMMENDATIONS

 INTRODUCTION

Background of the Study

Today’s situation, climate change is inaudible fluctuating, posing

serious environmental challenges and the most alarming global

concern as it affects all biotic and abiotic things and all countries and

regions of the world in general through different climate concerned

factors(Abdo W et al., 2022 ; Beirut, 2018)

These concern stems from its enormous impact on the

environmental, social-economical dimensions of the global.

 Agricultural sector has been primarily affected by climate change

rising temperatures and decreasing rainfall as a result of direct

influences on livelihood of the societies (Abraham et al., 2020).

 As a result, the number of people experiencing hunger worldwide

has been increasing, reaching 9.9% in 2020 as between 720 and

811 million individuals. Since the past six years, moderate to

severe food insecurity has affected more than 30% of the world's

population and for billions of people, a healthy diet is out of

reach (Mukhtar, 2023).

 Statement of the Problems

 Nowadays our planet has been pushing towards irreversible global

warming by greenhouse gases (GHGs) emissions, with its impacts in

worldwide (FAO, 2018)

 The Earth's climate is changing quickly, mostly due to increasing

concentrations of greenhouse gases (GHGs) in the atmosphere

caused by human activities, especially fossil fuel burning,

agriculture, and deforestation (Wigley, 1999; Stern, 2006; IPCC,

 As a result of increasing global climate change, its negatively mark

on agricultural food production and quality reason increase; food

insecurity and drop of livelihoods system of people (Adisu

Mekonnen et al., 2020). This is commonly true in countries where

low income and adaptive capacities (specially Ethiopia) (IPCC,

2018; NMSA, 2001).

 Climate risk impacts both food security and the living standard of

people involved in production structures and their distribution

channels (Belay et al., 2017; FAO, 2016)

 In 2019, “2 billion people, or 25.9% of the world

population, were either suffering from hunger or lacking regular

access to nutritious and sufficient food” (FAO, 2021).

 Despite a large number of Ethiopia's population has been affected by

chronic and temporary food insecurity, which is closely linked to

severe and recurrent food shortages and recurrent drought associated

with climate change (Mota et al., 2019)

  At the country level, the World Bank (2010) estimated that climate
change would reduce Ethiopia's GDP by 2-6% by 2015 and up to
10% by 2045 relative to the baseline.

 Generally, Ethiopian agriculture looks to be locked into a

downward spiral of low and declining productivity, caused by an
adverse combination of agro climatic, demographic, fiscal and
institutional constraints, trends and shocks (Manyazewai, 2000).

 Benna Tsemay district is one of the arid and semi-arid pastoral

and agro-pastoral districts of South nation nationality and peoples
region (SNNPR) and high vulnerability of farmers to climate
change through the year.
  Nevertheless, limited research consideration had been given to
sightseeing the determining factor of climate change adaptation
strategies in the study area.
 The information found in the literature was vague and limited,
however, there were regional and contextual differences in
adaption strategies both within and between communities.
 The review of the variables determining the adaptation approaches
smallholder households in the study's participants choose to survive
climate change had received scant attention in the literature on
climate change. Therefore, the principal motive to embark on the
research would be to fill the aforementioned research gap through the
following objectives.
 Objectives of the Study
General objective of the study

 To analyze impact of climate change adaptation strategies on

household food security in rural area of the Benna Tsemay district

Specific objectives of the study:- Were to:-

 Identify climate change adaptation strategies in the study area.
 Examine determinant factors of adaptation strategies of climate
change in the study area.
 Analyze impact of Climate Change adaptation strategies on food
Security in study area
 RESEARCH METHODOLOGY
Descriptions of the Study Area
 The study was conducted in the Benna Tsemay district of the
South Omo Zone which is one of 13 zones in the SNNPR
province of Ethiopia

 Benna Tsemay is bordered by; Hammer to the south, Selamago to

the west, Bako Dawela ari to the northweast, Mallee to the
northeast, Dirashe Special district to the northeast, Konso Special
district to the east and Oromia region to the southeast.
 Benna-Tsemay (Key Afer) is located about 739 km from Ethiopia's
capital, Addis Ababa, and has a total area of ​3,754 km2 (SOFFDB,
2014).
Data type and source
The study was used:-
 Quantitative and qualitative data type
 Primary and secondary data source

Sampling Techniques and Sample Size

 The study was used multistage sampling technique to select
study sample respondents
 Firstly, Benna-Tsemay district was selected purposely as
classified into lower and upper basis based on geographical area,
 In the second stage, 4 Kebeles were selected purposively based on
the farming system
 Accordingly, the names of kebeles were; Luka and Alduba from
pastoral, and Golidiya and kako from the agro-pastoral farming
system were purposely selected.
 Lastly, 246 sample size was randomly selected by Cochran
(1977) formula. Based on formula:
 Method of Data Collection
Data were collected through
 structured and semi structured questionnaires
 interview
 from sources (annual reports, documents and internet).
 from 12 key informant were interviewed 3 from each kebele,
 from 4 focus group discussions (FGDs) one from each kebele
were generated
Methods of Data Analysis
 Software Package for the Social Sciences (SPSS 26) and
 Stata 14 software and was followed, descriptive, and inferential
statistics with econometric models
 Econometric Model specifications
The study focused on two econometric models
A. Multinomial logit regression model
B. propensity Score matching (PSM) models

Variables
Dependent Variables:- Climate change adaptation strategies
Adapt or not adapt CC:-
Outcome variable:- status of household food security base on kcal
per adult equivalent per day (2200 kcal/adult equivalent/day
Independent Variables:- depend on conceptual framework and on
different reviews
 RESULT AND DISCUSSIONS
Descriptive Results of Sampled Households with dummy/categorical variables
Variable Descriptions Respon % Chi-square
dants test
SEX Male 194 78.86 4.342**

Female 52 21.14

Total 246 100

Marital status Widowed/Widower 45 18.29 3.562**

Single 45 18.29

Married 141 57.32

Divorced 15 6.10

Total 246 100.00

Access to climate information Yes 95 38.62 2.134***

No 151 61.38
 Sources: model output (2023); *, **, *** shows the mean difference
is significant at 10%, 5% and 1% probability level, respectively

 As shown in Table 2 sampled households gain access to climate

information, Access to credit and participation in training were
38.62, 61.38, 44.31 and 23.98 percent, respectively. More than half
of the percent farmers did not gain access to climate information,
Access to credit and agricultural extension or training.
Descriptive Results of Sampled Households with continuous variables
Variable Obs Mean Std. Dev. Min Max t test
Age 246 43.3658 15.38235 20 86 -1.238

Farm experience 246 9.138211 7.164909 1 34 -0.326*

Educational status 246 4.414634 3.329342 0 10 0.5334**

Family size 246 7.654472 3.237302 2 18 -2.342

Farm income 246 7155.26 9267.966 500 70000 -2.381**

Non farm income 246 16973.44 20979.86 543 235600 -3.523

Farm size 246 7.654472 3.237302 2 18 -0.153

Sources: model output
Livestock ownership 246
(2023); *
23.13016
and ** .65
21.3188
shows the mean
213.3 -1.045*
difference is significant at 10% and 5% probability level,
respectively.
Distance to market 246 10.85163 7.845013 1 94 -4.245**
 Climate Change Adaptation Strategies Utilized in the Econometric
analysis(MNL)

Sources: model output (2023)

 Above showed that, the sampled households response to changing

climate adaptation strategies farmers were employed on small scale

irrigation, improved seed varieties, income source diversification,

soil and water conservation practice, crop associations and selling

their asset 7.724%, 9.756%, 15.45%, 12.6%, 10.16% and

8.13%respectively.

 Among others, 36.18% of sampled household did not employ any

adaptation measure as alleviation for climate change effects.

 Table 5. food security status of households
Food security Freq. Percent Chi square (χ
status of 2)
households

food insecure 155 63.01 1.564 ***

food secured 91 36.99
Total 246 100.00
Sources: model output (2023)*** shows the mean difference is
significant at 1% probability level.
The survey result on table 5 above shown that 246 cross-
examined households, around:-

 63.01% (food insecured) of them were consumed less than the

ordinary of adult equivalent calorie intake per day (2200 kcal/day),
whereas

 36.99% (food secured) were consumed more than 2201–2500 kcal/

day which is recommended for healthy life.
 Results of Econometrics Models
 The best method to estimate factors that influence of independent
variables to climate change adaptation strategies on household, the
multinomial logit model was employed. The results of the
multinomial logit model estimations of the marginal effects beside
with the levels of statistical significance and p-values are presented
in Table 6.
Non adaptors (base
outcome)
Dependent Small Improve Income Soil and Crop Selling
Variables Scale d seed source water associati asset
irrigation varieties diversificat conservatio ons
Independent ion n practice
variables
Sex 0.025*** 1.360 −0.586 1.655** 1.689* −0.714*
(0.000) (0.471) (0.169) (0.003) (0.076) (0.037)
Age 0.007 1.143* 0.058* 0.042 0.0332* 0.0006
(0.789) (0.070) (0.021) (0.697) (0.075) (0.678)
Farmer -0.0314 0.0552* 0.226* 0.0062 0.0046* 0.0465
experience (0.675 (0.075) (0.016) (0.481) (0.056) (0.146)

Educational 0.0754** 0. 0.0343 0.1232* 0.00544 0.622

status * 0234** (0.545) (0.064) (0.665) (0.561)
(0.000) (0.007)

Martial status -0.01012 0.0334 0.0453 0.3224 0.00345 0.0465

(0.220) (0.363) (0.483) (0.441) (0.124) (0.234)

Family size −0.324* 0.248** 0.042 0.224 0.446** −0.253

(0.041) (0.006) (0.433) (0.211) (0.001) (0.443)

Farm income −0.0000 1.46e−0 0.000033 6.52e−05 0.00007 0.00006

21 22 (0.456) (0.719) 21 78*
(0.560) (0.237) (0.321) (0.062)

Non farm 0.0003 0.0008* 0.0000 0.0022*** 0.0008* 0.0006

income (0.3445) ** (0.2456) (0.000) ** (0.7889)
(0.000) (0.000)
Farm size 0.0032 0.00456 0.04432 0.00321 0.0076* 0.0004
(0.991) * (0.344) (0.532) (0.023) (0.766)
(0.021)
Livestock 0.245 0.037** 0.034 0.122*** 0.0543 0.0222
ownership (0.333) * (0.256) (0.000) (0.278) (0.156)
(0.000)
Access to credit 0.343* −0.211 0.222 0.332* 0.544* −0.103
(0.022) (0.455) (0.755) (0.023) (0.032) (0.665)
Access to 0.123** 1.556** 0.113 2.334 1.765** 0.566
climate (0.006) (0.003) (0.321) (0.233) (0.002) (0.311)
information
Distance to 0.330 −0.044 −0.236 0.206*** 0.311 0.331
market (0.235) (0.213) (0.779) (0.000) (0.448) (0.556)
Participation in 0.330*** 0.131 0.931 0.289** 0.545 0.223
training (0.000) (0.339) (0.866) (0.006) (0.412) (0.281)
Cons 0.531 −2.673 −4.432 −1.698 −6.731* −4.801
(0.876) (0.377) (0.108) (0.602) (0.071) (0.107)

Sources: model output (2023); *,** and *** shows 10%, 5% and 1%
were statistical significance level of respectively. p-values indicated in
the parenthesis
 The impact of climate change adaptation on farm households’
food security

 A logistic regression model was used to estimate propensity scores

for matching the observed characteristics of the adopters and non-

adopters Table 7.

 Logistic regression model is used a dichotomous dependent

variables with the value of ‘1’ if the farmer adopts and ‘0’

otherwise.

 Out of 14 explanatory variables, 9 explanatory variables were found

to significantly influence outcome variables corresponding of food

 Table 8. Common support region
adopters sampled Obs Mean Std. Dev. Min Max
157 .5943585 .1256429 .2934259 .8688993

non_adopter 89 .6630707 .1196684 .3353508 .9984464

sampled
Sum up 246 .6382114 .1260282 .2934259 .9984464

 The above table tells us the estimated propensity scores vary

between 0.2934 and 0.8689 (mean 0.5944) for adopter
households and between 0.3354 and 0.9984 (mean =0 .6630707)
for non adopter (control) households.
 Based on the minima and maxima criterion the common support
region would then lie between 0.2934 and 0.9984.
Also through the matching process, all the matching methods
employed removed the unmatched non-adapters, and hence, they
lead to the reduction in sample size for the post-matching impact
analysis.
Figure 6: Common support graph

.2 .4 .6 .8 1
.2 .4 Propensity .6
Score .8 1
Propensity Score
Untreated
Untreated Treated
Treated

Sources: model output (2023)

 Choice of matching algorithms
 Dehejia R. H. and Wahba Sadek (2002) state that various factors,
including the equal means test, also known as the balancing test,
pseudo R2, and matched sample size, were taken into consideration
while deciding on a matching estimator.
 The findings indicate that kernel matching with a band width of
(0.1) is the best estimator for the available data Table 9 based on
equal means testing (balance test), a low pseudo-R2 for the overall
balancing test, and a higher number of matched sample sizes and
balanced or insignificant mean differences across all explanatory
variables that was showed in Table 10
 Estimation of the treatment impact of adoption on households
food Security
Matching Matched Matched Treated Controls ATT SE T
algorithm treated controls
Nearest 167 167 926.70 234.56 692.14 118.34 2.33***
neighbor
Kernel 167 167 1020.46 460.68 559.78 135.28 1.32***
Radius 150 167 876.84 356.93 519.91 132.43 1.78***

Sources: model output (2023) * ** mean difference are statically

significant at 1% of probability level

 This implies that adoption of climate change adaptation strategies

has a contributing influence on household food security status.
 SUMMARY, CONCLUSION AND RECOMMENDATIONS
Summary and Conclusion
 The study was conducted to analyzed the impact of climate change
adaptation strategies on households food security in Benna Tsemaye
district, South Omo zone, Ethiopia.
 The data were collected by used household questionnaire surveys,
key informant interviews, and focus group discussions to acquire
first-hand data to analyze the identified objectives.
 The multistage sampling technique was used to select the districts as
well as to select representative 246 household from the total
population and analyzed though descriptive and econometric models.
 The descriptive statistics result showed that in the sampled
households, 63.8% chose to climate change adaptation strategies,
while the remaining 36.2% did not opt for it.

 The study used multinominal logit regression and propensity

matching models to identify the determinant factors that
significantly affects the choice of an adaptation strategy and to
estimate the effect of Climate change adaptation strategies on food
security in study area, respectively.
 The impact estimation results then indicated that there were
significant differences in food calorie intake between treatment
and comparison sample households, which could be attributable to
the participation in adoption. This indicates that, in terms of daily
food intake at the household level, adopters are much better off
than non-adopters.
Recommendations
 Based on the empirical findings of the study, the following
recommendations have emerged for policy-making as well as
government and other stakeholders should support to household
 Enhance the dissemination and accessibility of accurate and timely
climate information to farmers. This can be done through the
establishment of weather monitoring systems, early warning
systems, and the use of mobile technologies to reach remote areas.
 Enhance the capacity of agricultural extension services to provide
farmers with knowledge and skills on climate-smart agricultural
practices. This can include training programs, demonstration plots,
and farmer field schools.
 Facilitate access to credit and financial resources for farmers to
invest in climate change adaptation measures. This can be done
through the establishment of microfinance programs, insurance
schemes, and targeted grants for vulnerable farmers.
 Encourage the adoption of sustainable land and water management
practices, such as conservation agriculture, agroforestry, and water
harvesting techniques. This can help farmers build resilience
against climate risks and improve their productivity.
 Establish social safety nets and support mechanisms to assist
vulnerable farmers during times of climate-related shocks and
stresses. This can include cash transfer programs, crop insurance,
and community-based risk-sharing initiatives.
 Foster collaboration and knowledge-sharing platforms among
farmers, researchers, policymakers, and NGOs to facilitate the
exchange of experiences and best practices in climate change
adaptation.
 The End

Thank you so much