WILLINGNESS TO PAY FOR RAINFALL BASED INSURANCE

BY SMALLHOLDER FARMERS IN CENTRAL RIFT VALLEY OF

ETHIOPIA: THE CASE OF DUGDA AND MIESO WOREDAS

M.Sc. Thesis

Hiwot Teshome Abebe

October 2010
Haramaya University
 WILLINGNESS TO PAY FOR RAINFALL BASED INSURANCE
BY SMALLHOLDER FARMERS IN CENTRAL RIFT VALLEY OF
ETHIOPIA: THE CASE OF DUGDA AND MIESO WOREDAS

A Thesis Submitted to the College of Agriculture

Department of Agricultural Economics, School of Graduates Studies
HARAMAYA UNIVERSITY

In Partial Fulfillment of the Requirement for the Degree of

MASTER OF SCIENCE IN AGRICULTURE
(AGRICULTURAL ECONOMICS)

By
Hiwot Teshome Abebe

October 2010
HARAMAYA UNIVERSITY

ii
 SCHOOL OF GRADUATE STUDIES
Haramaya University

As Thesis Research advisor, I hereby certify that I have read and evaluated this thesis prepared
under my guidance, by HIWOT TESHOME, entitled WILLINGNESS TO PAY FOR RAIN
FALL BASED INSURANC BY SMALLHOLDER FARMERS IN CENTRAL RIFT
VALLEY OF ETHIOPIA: THE CASE OF DUGDA AND MIESO WOREDS. I recommend
that it can be submitted as fulfilling the Thesis requirement.

_________________________ _________________________________________
Major advisor signature date

As member of the Board of Examiners of the M. Sc. Thesis Open Defense Examination, we certify
that we have read, evaluated the Thesis prepared by Hiwot Teshome and examined the candidate.
We recommended that the Thesis be accepted as fulfilling the Thesis requirement for the degree of
Master of Science in Agriculture (Agricultural Economics).

______________________ _____________________ ________

Chairperson Signature Date

__________________ ____________________ ______

Internal examiner Signature Date

_____________________ _______________________ _______

External examiner Signature Date

iii
 STATEMENT OF AUTHOR

I declare that this thesis is my work and that all sources of materials used for this thesis
have been duly acknowledged. This thesis has been submitted in partial fulfillment of the
requirements of M.Sc. Degree at the Haramaya University and is deposited at the
University Library to be made available to borrowers under rules of the Library. I declare
that this thesis is not submitted to any other institution for award of any academic degree,
diploma, or certificate.

Brief quotations from this thesis are allowable without special permission provided
accurate acknowledgement of source is made. Requests for permission for extended
quotation from or reproduction of this manuscript in whole or part may be granted by the
major department or Dean of School of Graduate Studies in his or her judgment the
proposed use of the material is in interest of scholarship. In all other instances, however,
permission must be obtained from the author.

Name: Hiwot Teshome

Signature: ____________________
Place: Haramaya University, Haramaya
Date of submission: 9/12/2010

iv
 ABBREVIATIONS

AAU Addis Ababa University

ADLI Agricultural Development Led-Industrialization
CRWG Commodity Risk Management Group
CSA Central Statistics Agency
CVM Contingent Valuation Method
FAO Food and Agricultural Organization
GDP Gross Domestic Product
HPM Hedonic Price Method
ICRISAT International Crops Research Institute for the Semi-Arid Tropics
ITCZ Inter Tropical Convergence Zone
JJAS June-July-August-September
MARC Melkassa Agricultural Research Center
MPCI Multiple Peril Crop Insurance
NMA National Meteorological Agency
NAPA National Adaptation Program of Action
PA Peasant Association
RP Revealed Preference
SADC Southern African Development Community
SP Stated Preference
TCA Travel Cost Approach
TEV Total Economic Value
VIF Variance Inflation Factor
VLS Village Level Studies
WTA Willingness to Accept
WTP Willingness to Pay

v
 BIOGRAPHICAL SKETCH

The author was born in Amhara regional state Bahir Dar town, in November, 1985. She
attended her high school and preparatory school at Tana Hike secondary and preparatory
school and she completed her high school education in 2003.

She joined Haramaya University in September 2004 and successfully completed her
Bachelor of Science Degree studies in Agricultural Business Management in July 2008.
After graduation, the author joined Haramaya University to pursue her post graduate
studies in the Department of Agricultural Economics in October 2008.

vi
 ACKNOWLEDGMENTS

Above all, I would like to thank the Almighty and Merciful God, for providing me all the
patience and endurance during my study.

My sincere gratitude goes to my advisor Dr. Ayalneh Bogale for his intellectual
stimulation, professional guidance and encouragement in undertaking this study. I have
great respect to my father Ato Teshome Abebe and my mother W/r Keleme Ayisheshm for
their love, encouragement and financial support during my stay at Haramaya university
and research work. I also thank friends in Haramaya University and my families for their
encouragement and valuable support specially my uncle Ato Asres Lake.

I would like to extend appreciation goes to Ato Fresenbet Zeleke the Head of school of
Agricultural Economics and Agribusiness for his collaboration. And I would like to express
my gratitude to Ato Negeso and all other enumerators for their participation in the survey
work. Finally, finally I would like to acknowledge all individuals that directly or indirectly
contributed to the successful completion of this study.

vii
 TABLE OF CONTENTS

STATEMENT OF AUTHOR iv
ABBREVIATIONS v
BIOGRAPHICAL SKETCH vi
ACKNOWLEDGMENTS vii
LIST OF FIGURES xi
LIST OF TABLES IN THE APPENDIX xii
ABSTRACT xiii
1. INTRODUCTION 1
1.1. Background 1
1.2. Statement of the Problem 3
1.3. Objectives of the Study 5
1.4. Significance of the Study 5
1.5. Scope and Limitation of the Study 6
1.6. Organization of the Study 6
2. REVIEW OF LITERATURE 7
2.1. Definitions and Concepts 7
2.2. Risks in Agriculture 7
2.3. Rainfall Based Insurance 8
2.4. The Rural Poor and Risk Coping Strategies 9
2.5. Demand for Weather Based Insurance in Developing Countries 9
2.6. Traditional Crop Insurance versus Weather Index Insurance 10
2.7. Standard Approach to Develop a Weather Insurance Pilot 11
2.9. Methods of Valuation 11
2.9.1. Revealed preference method 11
2.9.2. Stated preference method 12
2.9.3. Willingness to pay (WTP) and willingness to accept (WTA) 15
2.10. Empirical Studies on Demand for Agricultural Insurance 16
2.11. Experience of Ethiopia 18
3. RESARCH METHODOLOGY 19
3.1. Description of the Study Area 19

viii
3.2. Sampling Techniques and Method of data Collection 21
3.4. Method of Data Analysis 22
3.5. Variable Definitions and Hypothesis 26
4. RESULTS AND DISCUSSION 32
4.1. Descriptive Statistics Results 32
4.1.1. Household characteristics 32
4.1.2. Resource ownership 35
4.1.3. Institutional characteristics 36
4.2. Sources of Risk and Management Strategies Practiced 38
4.2.1. Risk management strategies 39
4.2.2. Risk perception of sample households 40
4.3. Willingness to Pay Analysis 41
4.3.1. Estimation of the mean WTP value 42
4.3.2. Estimating total willingness to pay and total revenue 43
4.3.3. Derivation of aggregate demand 45
4.4. Econometric Model Result 47
4.4.1. Econometric test and results 47
4.4.2. Interpretation of the results 48
5. SUMMARY AND POLICY RECOMMENDATIONS 53
5.1. Summary and Conclusion 53
5.2. Policy Implication of the Study 56
7. APPENDIX 61
7.1. Appendix 1. Tables 62
7.2 Appendix 2. Survey Questionnaire 64

ix
 LIST OF TABLES

Table Page

1. Number of household and sample sizes ......................................................................... 21

2.Variables and their measurement included in the mode ................................................. 31
3. Characteristics of sample households for dummy variables .......................................... 34
4. Characteristics of sample households for continuous variables ..................................... 35
5. Economic Characteristics of sample household for continuous variable ...................... 35
6. Institutional characteristics of sample household for dummy variables ........................ 36
7. Summary of descriptive statistics of sample households characteristics between two
woredas for continuous variables. ....................................................................................... 37
8. Summary of descriptive statistics of sample households characteristics between two
woredas for dummy variables. ............................................................................................ 38
9. Major sources of risk and their rank given by sample households ................................ 39
10. Risk management strategies practiced by sample households ...................................... 39
11. Definition of risk by sample households ....................................................................... 40
12. Maximum WTP and percentage distribution of the sample household ........................ 42
13. The Logit model to calculate the mean WTP................................................................ 42
14. Total willingness to pay and total revenue in (Birr) ..................................................... 44
15. Maximum Likelihood estimates of the Tobit model ..................................................... 51
16. Marginal effects of the explanatory variable on the dependent variable ...................... 52

x
 LIST OF FIGURES

Figure Page

1. Location of the study area .............................................................................................. 20

2. Estimated demand curve for rainfall risk insurance service .......................................... 46

xi
 LIST OF TABLES IN THE APPENDIX

Appendix Table Page

Appendix 1. Variance Inflation Factor for continuous variables ........................................ 62

Appendix 2. Contingency Coefficients for dummy variables ............................................. 62
Appendix 3. Conversion factors for tropical livestock unit (TLU)..................................... 63

xii
 WILLINGNESS TO PAY FOR RAINFALL BASED INSURANCE BY SMALL
HOLDER FARMERS IN CENTERAL RIFT VALLEY OF ETHIOPIA: THE CASE
FOR DUGDA AND MIESO WOREDAS.

ABSTRACT

Current climate variability is already imposing significant challenge to Ethiopia.

Therefore, farmers have faced income variability in almost every production season.
Problems associated with dependence on rain fed agriculture are common in Ethiopia.
Smallholder farmers’ vulnerability from such income variability is also common. Over the
years, a range of risk management strategies have been used to reduce, or to assist
farmers to absorb, some of these risks. Since insurance is potentially an important
instrument to transfer part of the risk, this study try to describe the nature of weather
related risks faced by smallholder farmers, assess small holder farmers willingness to pay
for the rainfall risk insurance and examine factors that affect the maximum farmers are
willing to pay for the rainfall risk insurance. The data was collected from 161 sample
households from the two woredas of the study area using closed ended value elicitation
format followed by open ended follow up questions. The study uses Logit model to
estimate the mean willingness to pay in the close ended format in addition with Tobit
model to examine factors that affecting small holder farmer willingness to pay as well as
intensity of payment. The mean willingness to pay values are found to be 129.98 and
183.41 birr per hectare for the open and close ended formats respectively. The total
willingness to pay for the study area was found to be birr 5,740,244 per year. The tobit
model shows six potential explanatory variables affect the willingness to pay value.
Income of household and ownership of radio have positive and significant effect on the
value of willingness to pay, whereas off-farm income, age of household head, number of
livestock owning and availability of public and private gifts have negative and significant
effect on willingness to pay value. If the rainfall risk insurance premium is affordable and
households have enough information about the service they are willing to pay for the
service. Eventually policy makers need to be aware that socio-economic and institutional
characteristics of households influence the willingness to pay for rainfall risk insurance
services.

xiii
 1. INTRODUCTION

1.1. Background

Agricultural producers around the world are exposed to a variety of income uncertainties,
both market related, such as price variations, as well as non-market related, such as
unstable weather patterns. It is well known that such uncertainties induce substantial
income risks, and these can be particularly detrimental to small and/or poor producers in
developing countries (Sarris, 2002).

A number of countries in Africa already face various challenges due to climate variability
and recognize that adaptation is not an option but a necessity (Thornton et al., 2006). It is
also well known that farmers have developed several ways for dealing with the various
risks they face. Because climate change is expected to adversely affect agricultural
production which remains to be the main source of income for most countries (Bryan et
al., 2009).

A large array of adaptation practice is available to improve the resilience of smallholder

agricultural system to uncertain future impact of climate change. Over the years, a range
of risk management strategies have been used to reduce, or to assist farmers to absorb,
some of these risks. These strategies include on-farm measures such as diversification or
selecting less risky production methods, as well as strategies for sharing risk with others.
Risk management strategies in which risks are shared with others include, among others,
farm financing, share-cropping, price pooling arrangements, forward contracting of farm
products, and hedging on future markets. Also insurance is potentially an important
instrument to transfer part of the risks (Anderson, 2001).

Insurance markets are growing rapidly in the developing world, as part of this growth;
innovative new products allow individual smallholder farmers to hedge against
agricultural risks, such as drought, disease and commodity price fluctuations (World Bank,
2005). These financial innovations hold significant promise for rural households. Shocks
to agricultural income, such as a drought-induced harvest failure, generate movements in

1
consumption for households who are not perfectly insured, and at the extreme, may lead to
famine or death.

Interest in developing catastrophic weather insurance products for rural dwellers in

developing countries has grown radically in recent years. This interest has been fueled by
the successful introduction of new products for the management of systematic risks to
international financial markets in the recent years. Rainfall-based index insurance products
for agriculture represent an attractive alternative for managing weather risk (Hellmuth et
al., 2009).

These products include catastrophic bonds and area yield crop insurance options, and their
success suggests that it may be possible to package catastrophic weather and natural event
risks facing developing countries and reallocate them to international markets in a cost
efficient manner, bringing affordable risk management services to rural dwellers in
agriculture dependent countries (Skees, 2001).

The demand for such insurance particularly in developing countries has been increasing
over time, as a result of unpredictable weather conditions. In case of Ethiopia the impact
that climate variability has on predominantly rain-fed agrarian economies is clearly
demonstrated. Current climate variability is already imposing significant challenge to
Ethiopia by affecting food security, water and energy supply, poverty reduction and
sustainable development efforts, as well as by causing natural resource degradation and
natural disasters. In response, the national adaptation program of action (NAPA) for
Ethiopia has been prepared and the basic approach to NAPA preparation was along with
the sustainable development goals and objective of the country where it has recognized
necessity of addressing environmental issues and natural resource management with the
participation of stakeholders (MoWR, 2007).

Agriculture, as in many other developing countries, is the mainstay of Ethiopian economy.

Nearly 85% of the population earns its livelihoods and contributing over 43% of the GDP
(Gross Domestic Product), about 90% of the foreign exchange earnings, a further 10%
earn their living from livestock (CSA 2004) which is almost entirely small-scale and rain-
fed. Both farmers and pastoralists are highly dependent on the climate for their livelihoods
(World Bank, 2000).
2
Ethiopia has recognized climate change as an important issue and attempts are being made
to incorporate potential response measures for reducing impact of climate change in to
over all development planning process. One important constraint that emerged as a result
of stakeholder consultative meetings is the extreme need for agricultural rainfall risk
insurance. It is believed that agricultural rainfall based insurance is seen as one of the
strategies to minimize risk and capitalize on opportunities associated with the variable
climatic conditions.

The high covariance of climatic risks, coupled with the lack of property to be attached as
collateral, makes it difficult for cooperatives, microfinance organizations, or banks to
provide financial services to smallholder farmers unless they have insurance/reinsurance
against weather risk. These conditions in turn keep farming at a subsistence level.
According to Stern (2007), adaptation to climate change and variability will be crucial in
reducing vulnerability and is the only way to cope with the impacts that are inevitable over
the next few decades. This research is therefore an attempt to look in to the possibility of
rainfall based insurance existence in Ethiopia.

1.2. Statement of the Problem

The farming community on the globe in general and that of the least developed countries
in particular is considered to have a risk aversion attitude (Anderson et al., 1977; Dillon
and Hardaker, 1993). Hardaker et al., (1997) and Binswanger, (1980) have also argued
that most smallholder farmers avers to risk-by and large, they are too poor to be otherwise.

Better decisions in risky world can always be made if information about more productive
technology option, marketing opportunities and marketing trends are available. But almost
all small scale farmers in the globe avers to risk because they are poor (Hardaker et al.,
1997), which holds true for the farming community in Ethiopian central rift valley. This is
because of the erratic rainfall in the area. Furthermore, weather related agricultural
production shocks also conspire to keep smallholders within the poverty trap, preventing
the country from reaching its productive potential in the agricultural field (Hess and
Syroka, 2005).

3
Agriculture is often carried out in open air, and always entails the management of
inherently variable living plants and animals which are especially exposed to risk.
Production risk comes from the unpredictable nature of the weather (Hardaker et al.,
1997). And it is probably fair to claim that farmers in developing countries are exposed to
most types of risk, and the low-income farmers, especially in semi-arid areas are the most
exposed (Hazell 1992).

Ethiopia is among famine-prone countries in Africa and has a long history of famine and
food shortage that can be traced back to 250 BC (Assefa and Ramakrishna, 2002). More
than half of the food insecure African population lives in Ethiopia, Chad, Zaire, Uganda,
Zambia and Somalia and the food insecure population in Ethiopia is estimated to be
around 40-50 percent of the total population (Assefa and Ramakrishna, 2002). .

Therefore, farmers have faced income variability in almost every production season.
Problems associated with dependence on rain fed agriculture are common in Ethiopia;
repeated famine, crop failure, human and livestock loss are among the indicators (Assefa
and Ramakrishna, 2002). Smallholder farmers’ vulnerability from such income variability
is also common in Ethiopian central rift valley and the two woredas of the study area,
Dugda and Mieso.

One major constraint to initialize the opportunity to operational risk-covering mechanisms

like rainfall crop/input insurance in Ethiopia is absence of public and/or private
institutions. In order to exploit the advantage associated with good rainfall seasons, risk
financing institutions need to be encouraged to develop operational risk insurance schemes
in the marginal rainfall areas.

This study tries to identify willingness to pay for rainfall based insurance by smallholder
farmers in central rift valley of Ethiopia. The main issues and problems that need to be
researched and analyzed in this study are: to investigate whether smallholder farmers are
willing to pay for rainfall based insurance and identify factors that determine their
maximum willingness to pay for the rainfall based insurance as well as explore the
existing risk insuring mechanisms commonly used by small holder farmers. This study
was, therefore, initiated to fill the current information gap and awareness on the subject.

4
1.3. Objectives of the Study

The objectives of the study are;

1. To describe the nature of risks faced by small holder farmers in the study
area;
2. To assess the willingness to pay for rainfall based insurance by small holder
farmers in the study area;
3. To examine factors that affect the maximum farmers are willing to pay for
rainfall based insurance in the study area.

1.4. Significance of the Study

The National development plan of the country is based on a strategy called Agricultural
Development Led-Industrialization (ADLI), and aims at changing the country’s
subsistence or traditional agricultural to commercial or market oriented one, which in turn
will increase the demand for goods and services and further lead to industrial
development. The Government strategy is aimed at reducing country’s dependency on
food aid. To achieve the intended goals within a short period of time, understanding
smallholder farmers’ participation as well as their willingness to pay for rainfall based
insurance will be vital. Reducing the vulnerability of rainfall dependent communities to
climate change requires building of local institutions to support better adaptation practices
where vulnerability is usually more clearly expressed.

This research looks also in to the willingness to pay for rainfall-index based insurance
contracts that can promote more efficient program of actions in reducing problems of
imperfect information in mitigating farmers’ risks in Ethiopia. Therefore, identifying
smallholder farmer’s willingness to pay for rainfall based insurance is expected to be
useful for policy makers in providing good information, for decision makers to make
informed choices on where and how to intervene and funding agencies, involved in the
development and promotion of weather based insurance. Even though the study was
conducted in the Central rift valley of Ethiopia, the result can be applicable to other parts
of the country which have almost similar climate condition. The outcome of this study is

5
also expected to be useful for governmental and non governmental institutions who are
involved in the weather based insurance service.

1.5. Scope and Limitation of the Study

The scope of this study covers assessing willingness to pay for rainfall based insurance
and examining socio-economic and institutional factors that significantly affect the
amount of money farmers are willing to pay for rainfall based insurance. The proposed
research is confined only to two woredas of the central rift valley, which can somehow
represent other woredas of the Central rift valley of Ethiopia this is because of resource
constraint to undertake the study at broader level. In addition, the data collected for 2010
are a onetime data this might not be enough to generate adequate information because
there are many variables which could be potentially changed from one survey time to the
other survey time. As the research uses contingent valuation method (CVM) the study is
subject to all limitations associations with the method however, efforts have been made to
minimize the limitations of the methodology.

1.6. Organization of the Study

The study is organized in five chapters. Chapter one deals with background, problem
statement, objectives, scope and significance of the study. And the remaining chapter, two
and three deal with review of theoretical and empirical literature related to weather related
insurance (rainfall) and the research methodologies, respectively. Chapter four presents
results and discussion of the study. Finally chapter five summarizes the finding of the
study and gives policy implication and recommendation.

6
 2. REVIEW OF LITERATURE

2.1. Definitions and Concepts

One common distinction between risk and uncertainty is to suggest that risk is imperfect
knowledge where the probabilities of the possible outcomes are known, and uncertainty
exists when these probabilities are not known but the distinction of what risk and
uncertainty for the farmers is theoretically and practically not clearly defined (Hardaker et
al., 1997). Risk is everywhere and is substantially unavoidable. It is often said that, in
business, profit is the reward for the risk bearing, no risk, no gain

According to (Hazall et al., 2010), Risks can be characterized according to a number of

elements, including: Covariance; the degree to which they are correlated across
households within a community or region, ranging from independent (affecting one
person) to highly covariate (affecting everyone at the same time); Frequency; How often
they occur; Types and severity of losses incurred; Shortfalls in seasonal production and
income, damage to assets and loss of life.

2.2. Risks in Agriculture

Agricultural production is a risky business. Farmers have faced a variety of price, yield
and resource risks that make their incomes unstable and unpredictable from year to year.
The friction due to risk may also contribute to a lag in agricultural incomes relative to
those in other sectors of the economy. The people who need to concern themselves with
risk in agriculture include farmers, farm advisors, and commercial firms selling to or
buying from farmers, agricultural research workers, policy makers and planers. According
to (Holden et al., 1991) the greatest risks to family welfare in agriculture are centered in
rural areas, which specialize in annual food crops but which are marginal to the production
of those commodities, it is poverty and even worse alternatives which bring about such
production emphasis. Such strategies are doubly risky because they are often
unsustainable environmentally. In such area, fluctuation in weather and production are
around the critical margin of profitability which in the case of poor countries and people
means at the margin of existence.

7
2.3. Rainfall Based Insurance

The concept of index-based insurance is not new. Proposals for this type of insurance were
first articulated by Halcrow (1948) and Dandekar (1977). The Australian Government
commissioned a feasibility study of rainfall insurance in the mid-1980s, but decided not to
pursue it (IAC, 1986). Index-based insurance is a financial product linked to an index
highly correlated to local yields. Contracts are written against specific perils or events (e.g.
area yield loss, drought, hurricane, flood) that are defined and recorded at regional levels
(e.g. at a local weather station). Indemnifications are triggered by pre-specified patterns of
the index, as opposed to actual yields (Hazell et al., 2010).

Research carried out through the International Crops Research Institutes for the Semi-arid
Tropics - Village Level Studies (ICRISAT VLS) suggested that rainfall lotteries are better
than the crop insurance schemes to diminish rural household income variability in a cost-
effective manner in rain fed areas of India (Walker and Ryan, 1990). There would be a fair
betting system and would be open to all households in the village. For instance, if landless
labor households felt the demand for their labor was markedly reduced in low rainfall
years, they could hedge their future labor income by purchasing tickets on the lowest or
what they perceive to be the most adverse rainfall event.

Identifying weather risk for an agricultural producer involves defining the time period
during which risk is prevalent, and identifying a measurable weather index that is strongly
correlated to farmers losses on a particular crop. This is the most critical process in
designing a weather risk management strategy. A weather index can be constructed using
any combination of measurable weather variables, over any period of time and any
number of weather stations (Walker and Ryan, 1990).

Problems with fixing insurance premium: three type of problems related to insurance
premium are; adverse selection, covariate risk and moral hazard.

adverse selection: This occurs when potential borrowers or insures have hidden
information about their risk exposure that is not available to the lender or insurer, which
then becomes more likely to erroneously assess the risk of the borrower or insure.

8
Covariate risk: Risk that can affect large numbers of people at one time (e.g. widespread
drought, flooding, earthquake).

Moral hazard: This occurs when individuals engage in hidden activities that increase
their exposure to risk as a result of borrowing or purchasing insurance. These hidden
activities can leave the lender or insurer exposed to higher levels of risk than had been
anticipated when interest or premium rates were established.

2.4. The Rural Poor and Risk Coping Strategies

Siegel and Alwang (1999) developed taxonomy of risk-coping strategies for rural
households facing risk. However, many strategies are unavailable or prove ineffective for
the poor, especially when the risks are covariate. Households living on very low incomes
and limited wealth become highly risk averse. Since even a small disruption in income
flows can have devastating effects, such risk aversion retards the development process by
limiting household incentives to adopt productivity-enhancing technologies and to
specialize in activities where comparative advantages exist. Such risks also affect the
credit-worthiness of rural households and constrain credit markets. Farmers, who are more
risk averse with respect to losses, would be more likely to participate in crop and rainfall
insurance programs and would be willing to pay higher premiums and individuals would
include insurance in their risk management strategies if the insurance premium were less
than the cost of other risk responses having the same effect (Patrick, 1988).

2.5. Demand for Weather Based Insurance in Developing Countries

Weather index-based insurance was being discussed in academic papers as an alternative

solution for developing agricultural economies in developing countries. In 2002, donors
began to finance the piloting of these ideas. In particular, the World Bank’s Commodity
Risk Management Group (CRMG) allocated trust funds from the Swiss and the Dutch
governments to pilot weather insurance for farmers to complement its price risk
management work in commodity markets.

Commodity Risk Management Group (CRMG) has been involved in many weather risk
management technical assistance projects to commercial entities in the developing world.

9
CRMG was involved in its first index-based weather risk management transaction in India
in June 2003, the first-ever weather insurance project in the country. Since 2003 there
have been several other pilots around the world, including completed pilots in Ukraine,
Ethiopia, and Malawi, and upcoming pilots in Kenya, Tanzania, Thailand and Central
America. Successes like the market growth in India have had significant demonstration
effects and have proven that weather risk management for farmers in the developing world
is possible through insurance -type instruments (World Bank, 2007).

2.6. Traditional Crop Insurance versus Weather Index Insurance

Traditional multiple-peril crop insurance that indemnifies losses on individual farm basis
is subject to high administrative costs in order to overcome the problems of adverse
selection and moral hazard. It also requires significant investment in monitoring farm
yields to prevent both higher losses than the initial rating and serious actuarial problems.
Furthermore, multiple-peril crop insurance has large correlated risks, so it requires the
extra cost of providing reinsurance. These extra costs can be quite high in an emerging
economy with little or no experience in providing insurance of this type. These conditions
mean that traditional multiple-peril crop insurance is not a workable solution for most of
agriculture in developing countries (Hess and Syroka, 2005).

One form of agricultural insurance that mitigates these added costs is weather insurance.
Payout is determined by an objective parameter such as millimeters of rain, soil moisture,
etc. Weather index insurance was found to be well suited to the agricultural production in
regions in Ukraine where there are wide spread crop losses due to drought and frost (Hess
and Syroka, 2005). The monitoring costs of weather insurance are less as there is no need
to perform farm-level loss adjustments and the balance of information about the weather is
equally shared by the insured and the insurer (unlike with traditional farm-level insurance
where the farmer will always know more about the yield than the insurer). Thus, weather
insurance could be a preferred alternative to crop insurance, as it avoids moral hazard
problems and high administrative costs. Furthermore, the reinsurer is more likely to
provide better terms when the insurance is based upon weather events and not farm-level
losses.

10
2.7. Standard Approach to Develop a Weather Insurance Pilot

The World Bank (2007) has drawn some lessons from its work and begun to develop a
standardized approach to pilot implementation as well as contract design. While this
approach is still evolving, there are seven basic components of pilot program implemen-
tation that need to be undertaken in order to develop a product that is not only technically
sound but is demanded and can be afforded by clients:

1. Identify potential pilot areas and carry out a basic risk assessment,
2. Identify delivery channels for reaching the end users,
3. Design contracts,
4. Determine the marketability of the products,
5. Finalize contracts and insurance,
6. Market the product, and
7. Monitoring the pilot.

2.9. Methods of Valuation

The farmer’s decision to purchase rainfall insurance and the maximum premium he/she is
willing to pay can be considered in the framework of maximizing net benefits from non-
market goods and services. The principles that non-market goods and services are not
efficiently allocated by the market suggests the possibility of improvement in
measurements of benefit and costs.

According to Freeman (2003), the widely used methods of valuation of some non-market
goods and services are revealed and stated preference methods. Revealed preference
methods are based on the actual behavior reflecting utility maximization subject to
constraint.

2.9.1. Revealed preference method

Revealed preference methods are based on the actual behavior reflecting utility
maximization subject to constraint. One type of the revealed preference method is based
on the observed choices in a referendum way. If an individual is offering a fixed quantity

11
of a good price on a take it or leave it or yes or no basis, observation of the choice reveals
only whether the value of the offered to the individual was greater than or less than the
offered price. The other methods for valuation of non-market goods under revealed
preference techniques are the Hedonic Price Method (HPM) or Property Value Method,
where the change in the environmental amenity is reflected in the value attached to the
amenity and Travel Cost Approach (TCA) where it mostly used to capture the recreation
value of sites, such as national parks and sanctuaries. The travel cost approach is applied
to determine the influence of various socioeconomic characteristics and the nature of
demand for recreation site (Marothia, 2001). However, revealed preference models can not
measure existence value or option value. So, firstly they cannot measure total economic
value (TEV) and secondly while RP models measure the household’s WTP, one cannot be
sure that the price captures all the effects.

2.9.2. Stated preference method

Stated Preference Method uses a direct approach to elicit willingness to pay, this method
involved asking people directly about the values they place on non-market services by
creating in effect, a hypothetical market (Freeman, 2003). Among the frequently used
methods of stated preference, the Choice Modeling and Contingent Valuation Method
(CVM) are the commonly used ones Choice Modeling do not ask questions directly;
instead they ask people to rank alternatives, whereas, CVM is used when market do not
exist for environmental resources by asking questions directly (Mitchle and Carson, 1989;
Hausman, 1993).

The valuation is done based on hypothetical or non-existing market. The valuation task is
therefore, to determine how much better or worse off individuals will be as a result of
change in non-market goods. Among the commonly used methods of the stated preference
contingent valuation method is widely used.

2.9.2.1. Contingent valuation method

Contingent valuation method as one of the stated preference methods, is basically uses a
survey based approach. The decision to use willingness to pay (WTP) or willingness to
accept (WTA) depends on, among other things, individuals’ perception as to who has the

12
property right over their source in question (Carson et al., 2001). This is computed by
asking how much people are willing to pay for a non-market goods (WTP) or how much
they are willing to give up having a specified non-market goods quality improvement
happen (Freeman, 2003).

When market data are unavailable or unreliable, economists can use alternative estimation
methods that rely on hypothetical market conditions. Such methods typically use surveys
to inquire about individuals’ willingness to pay (WTP) for some environmental policy
initiative. This survey approach to benefit estimation is known as the contingent valuation
method (CVM) because the results are dependent up on the hypothetical market devised.

In general, CVM helps researchers to capture the total value of the good both use and
non-use values and its flexibility facilitate valuation of a wide range of non-marketed
goods. As a result, this method is becoming the most preferred valuation method at
present. The major problems with this approach have largely to do with the specification
of the “scenario” or the “benchmark” against which the agent is supposed to compare the
current situation, and express a monetary value for what it is worth to him/her to move to
the new situation, or avoid a bad one

There are number of different elicitation methods used in CVM. Dichotomous and open
ended are among the methods used for obtaining the WTP. The open ended question asks
the respondent how much he or she is willing to pay for given change in the status quo.
This means individuals are asked for their maximum willingness to pay with no value
being suggested to them. The other method is dichotomous choice question whereby a
respondent is asked if he or she is willing to pay a specific amount of money for a pre
specified change (Bateman et al., 2000)

The use of specified format has got the advantage over the open-ended format question in
eliciting WTP because of the simplicity for respondents and reduced incentives for
strategic responses (Bateman et al., 2000). In the dichotomous method, if the first bid
given to the respondent is accepted, a second somewhat higher offer price is made. If the
first bid is refused, the second bid price offers is somewhat lower, the bid levels offered in
the follow up question will be greater than that offered in the initial payment if the answer

13
to the initial payment question is “yes” and vice versa. Finally the dichotomous choice
question is followed by an open-ended follow up question (Alberini and Cooper, 2000).

Biases in CVM: criticism on CVM is, since individuals are being given a hypothetical
market their responses could be far from reality. Thus there will be biases, which can
systematically understate or overstate true values. There are a number of types of biases
indicated, some of them are:

Strategic Bias: This occurs when the respondent tries to understate or overstate the bid
value so as to influence the outcome. For instance, if the CVM requires payment of a tax
the respondent may strategically understate the bid value to influence the outcome. Using
the take-it-or-leave it method, Mitchell and Carson (1989) suggests that, deleting protest
bids and remove all outliers are the ways to tackle this bias.

Hypothetical Bias: This arises due to the hypothetical nature of the market in CVM
surveys which can render respondents’ answers meaningless if their declared intentions
cannot be taken as accurate guides of their actual behavior. Experimental trials suggest
that this problem is less when one uses WTP format instead of WTA format.

Information Bias: The quality of information given in a hypothetical market scenario

almost certainly affects the responses in a CVM Survey. Inadequate or improper
presentation of information on the good or service to be valued can bias the quality of the
CVM study. Besides by making respondents feel that the hypothetical market is realistic,
and avoiding WTA format can remove information bias.

Starting Point Bias: The suggestion of an initial starting point in a bidding game can
significantly influence the final bid. For example choosing a low (high) starting point
leads to a low (high) mean WTP.

Interviewer and Respondent Bias: The interviewer’s conduct and interviews can
influence responses. Though this kind of bias can be minimized by using mail or telephone
surveys, this will result in less information forthcoming and also give rise to hypothetical
bias. Respondents may not give correct answers or give the questions proper

14
consideration. Therefore, to minimize this problem, professional interviewers should be
used or well trained interviewers to reduce this type of bias.
.
2.9.3. Willingness to pay (WTP) and willingness to accept (WTA)

Willingness to pay and willingness to accept are two methods for elicitation of values.
WTP is the amount that must be taken away from the person’s income while keeping his
utility constant in the same manner, WTA for a good is defined as the amount of money
that must be given to an individual experiencing deterioration in environmental quality to
keep his utility constant. The decision to use willingness to pay (WTP) or willingness to
accept (WTA) depends on, among other things, individuals’ perception as to who has the
property right over their source in question (Carson et al., 2001). This is computed by asking
how much people are willing to pay for a non-market goods (WTP) or how much they are
willing to give up having a specified non-market goods quality improvement happen
(Freeman, 2003).

In theory, when WTP is a small fraction of income, WTP and WTA for a given
commodity should be approximately equal. However, a number of CV studies have found
that WTA is often much larger than WTP for the same commodity. One explanation is that
the difference between WTP and WTA depends on the elasticity of substitution between
the commodity to be valued (a public good) and private substitutes. The lower the
elasticity, the fewer will be the available substitutes and the greater the difference between
WTP and WTA (Hanemann, 1991).

Another explanation - the theory of prospects - is that individuals value losses more
heavily than gains. It is also possible that individuals react to their perception of who has
the property rights over the commodity in question. If the proposed policy contradicts their
perception of the existing property rights, individuals might express their rejection of the
scenario through high WTA values. Carson (1991), suggests that WTP should be used
whenever the individual might incur benefits from the proposed policy, and Mitchell and
Carson (1989), offer ways to frame the payment question to elicit WTP.

However, even when the individual might incur benefits from the proposed policy, there
are some scenarios under which the respondent may not overstate WTA values (Cooper

15
and Osborn, 1998). Moreover problem with direct WTP studies involves the fact that
reported values are likely to be influenced by recent experiences. For instance, farmers are
more likely to express high demand for drought insurance if weather in recent periods has
been adverse. There are also several technical issues concerning the method of deriving
the WTP from either direct expression of values, or contingent rankings of alternative
choices, but these seem to have been largely resolved (Hanemann and Kanninen, 1998).

2.10. Empirical Studies on Demand for Agricultural Insurance

The review presented in this section shows the logical reasons forwarded by different
researchers about the demand for agricultural insurance by farm households, factors
including socio-cultural, economical and institutional.

There are very few studies relevant to agricultural insurance that use the CV approach.
Patrick (1988), analyzed producers’ demand for a multiple peril crop insurance (MPCI)
program with indemnities based on actual yields, and a rainfall insurance program with
indemnities based on area rainfall. Tobit regression analysis was used to estimate
responses utilizing information from the participants and non-participants in the
hypothetical programs. He found that expected wheat yield had a negative effect on the
premium, and suggesting that an area crop insurance program might encounter difficulties
of adverse selection. Area in wheat had positive relation to premium paid for crop
insurance whereas age has negative relation to premium paid for crop insurance. Farmers
who are legume producers and those who are averse to risk would be willing to pay higher
crop insurance premium.

The finding of this study is showed that the participation in the crop and area rainfall
insurance programs would be limited. One quarter of the producers would participate in
the crop insurance program and over one half would not participate in the rainfall
insurance program. Twenty percent or less of the producers would be willing to pay the
estimated full costs of the insurance programs and the author suggests very limited
potential for commercial establishment for programs under current circumstances of
drought assistance.

16
In developing country context, the study by McCarthy (2003) found considerable demand
for weather-based wheat insurance in Morocco farmers. The indirect methods of
estimating WTP involve first the specification of a model of the random income or other
variable of direct relevance to the farmer’s welfare (e.g. consumption), the information
from formal sources e.g. radio, television in fact has a negative impact on demand,
indicating that those who keep better informed of rainfall at the station are less likely to
prefer any insurance counteract.

The result showed that explanatory variables had ambiguous impacts differed both
quantitative and qualitative across and within the region. The author concluded that
demand for insurance, however, appears to be quite distinct across the different areas,
which indicates the need for larger data sets to satisfactorily estimate the determinants of
the willingness to pay. The author also expresses the WTP as the amount of money that
would equate the expected utilities of the relevant variable with and without the insurance.
This amount of money (the premium) is then estimated for objectively estimated values of
the risks with and without the insurance, and for a range of relevant utilities, or relevant
parameters (such as degrees of risk aversion) from a given class of utilities.

The study by Gautam et al. (1994), where the farm household’s behavior is assumed to be
described by the maximization of the expected value of inter temporal utility function. The
production, saving, labor allocation, diversification, borrowing, and insurance decisions
are assumed to be endogenous. The equilibrium conditions of the optimization problem
are manipulated to infer the production and diversification decisions of the household as
functions of both standard variables as well as a variable that measures the relative
preference of the household for risky versus non-risky income.

The same approach is essentially followed by Sakurai and Reardon (1997), who utilized
panel data for Burkina Faso. The additional feature of this study is that the researchers
regress their estimates of farm level demands for drought insurance on a set of variables,
so as to identify variables that increase or decrease such demand. They found as expected,
that the demand for drought insurance depends on the perceived probabilities of droughts,
and is higher for regions with higher such probabilities. They also found that variables
such as the size of cultivated area, and the age of household head significantly affect
positively the demand for insurance, while the amount of off-farm income, the availability
17
of public aid and private gifts, and the size of household significantly affect negatively the
demand for insurance.

2.11. Experience of Ethiopia

In recent years Nyala insurance has provided two types of crop insurance: multiple-peril
crop insurance (MPCI) and index-based weather insurance, each designed to meet the
needs of different farmers. Nyala’s MPCI is a double-trigger scheme that insures farmers
against a number of different shocks both natural and human caused that affect crop
yields, including shortages of rainfall, excess rainfall, fire, and transit risks. Because MPCI
insures against a number of perils, it is better suited to farmers who face a number of
sources of risk to crop yields than it is to farmers whose predominant source of risk is
rainfall variability. Since 2009, Nyala Insurance Corporation introduces and sells weather
index insurance through farmer cooperatives, taking advantage of low-cost automatic
weather stations owned by the National Meteorological Agency (Hazell 2010) this is
specifically to protect smallholder farmers against weather risk. This product was provided
in cooperation with Oxfam-America, mainly using satellite data and a weather index
product was designed in collaboration with the World Food Program around the rainfall
requirements of different crops. Currently, Nyala insurance has found that farmers’ unions
serve as effective delivery channels for the weather insurance products.

18
 3. RESARCH METHODOLOGY

3.1. Description of the Study Area

This study is under taken Dugda and Mieso woredas. These woredas are among the
central rift valley woredas of Ethiopia. Dugda woreda is located in eastern Shoa zone of
Oromia region. Dugda woreda, the capital is Meki, is located about 175km south of Addis
Ababa. It has a total population of 144,849 (CSA, 2008). The altitude ranges from 1610-
2020 m. a. s .l. Because of its location in semi-arid type of ago-ecology, the woreda has a
bimodal and erratic type of rainfall with high variation between and within years. The
woreda has a total of 36 Peasant Associations (PAs).

Crop-livestock mixed farming system characterizes agriculture in the woreda. Cattle,

goats, sheep and chickens are important livestock species reared by farmers. Maize, teff,
wheat and haricot bean are major crops grown by farmers. Besides cereals and pulses
farmers in the woreda produce significant amount of horticultural crops particularly
vegetables. Onion, tomato, pepper and cabbage are the most widely grown vegetable crops

Meiso Woreda is located 300km east of Addis Ababa, and at about 200 km east of Adama
town. It is located west and is one of Somalia region woredas in Oromia where agro
pastoral farming system is practiced. The woreda has a total number of 37 rural kebeles
and four town dwellers’ associations. The total human population of the woreda is
estimated at 145,775, and is composed of 22,012 agricultural rural households and 6785
urban households. The total rural population is 115,568, out of which 58,612 (51%) are
males. Of the total rural households, 17,495 (80%) are male-headed households.

The woreda has a total area of 2573.44 km2 (about 257,344 ha) and is situated between
4009”30’ E and 40056”44’ E; and: 8048”12’ N and 9019”52’ N. the woreda’s attitude
ranges between 900-1600 masl. The mean annual temperature varies between 240C-280C.
The mean annual rainfall ranges from 400 to 900 mm, with an average of about 790 mm
(IPMS 2006). Agro-ecologically, the woreda is classified as lowland (Kolla). The area
receives a bimodal rainfall where the small rains are between March and April while the
main rains are between July and September. During the small rains, are unpredictable and
erratic, and as a result, crops fail in most years due to lack of even distribution of rainfall.

19
Recurrent drought is a major problem, and is making relief aid a regular source of
livelihood for many rural families. A total land area of 22,487 ha (about 12% of the
woreda) is considered suitable for crop production.

Dugda
Meiso
East Shewa
West Hararge
Oromia
Etregions
N

W E
400 0 400 800 Kilometers
S

Figure 1. Location of the study area

20
3.2. Sampling Techniques and Method of data Collection

3.2.1 Sampling techniques

A multi stage sampling technique was used to select 161 sample households. In the first
step of the sampling, out of the woredas in the central rift valley that have almost similar
climate condition, Mieso and Dugda woredas were purposively selected because these
areas are most drought prone areas. In the second stage, out of the 36 PAs in Dugda
woreda 4 PAs were selected randomly and of 36 PAs in Mieso woreda 3 PAs were
selected. In the third stage the total numbers of households in each PAs were listed and
finally a total numbers of 161 sample households were selected and interviewed based on
the proportional to sample households.

Table 1. Number of household and sample sizes

Farming No of HH in
Woreda No of PAs system Name of PAs PAs Sample HH
Dugda 4 Mixed B/Gusaa 344 18
Mixed Jawe Bofo 332 16
Mixed Odd Bokota 496 24
Mixed S/Wakalee 474 23
Mieso 3 Agropastoral Buri Mulu 584 29
Agropastoral Chobi 500 24
Crop-livestock Husemandhera 551 27
Total 7 3281 161

3.2.2. Method of data collection

Data were gathered from primary and secondary sources. The primary data were collected
from sample households through a structured questionnaire using face to face interview as
well as CVM was employed to collect willingness to pay data. The secondary data were
collected from the existing government line departments and offices, records of non-
governmental organizations.

21
In addition a structured questionnaire focus group desiccation was done with a group of
farmers in each woredas PAs to know major sources of risk and management strategies
practiced. Six enumerators who speak the local language were recruited from the study
area and trained on interviewing techniques and how to manage CV questions.

3.2.3. Method of WTP data collection

Willingness to pay is defined as the amount that must be taken away from household’s
income. The willingness data is collected through CV method, this method is also suited to
solicit consumers’ willingness to pay for a product that is not yet on the market. CVM is
now increasingly used in developing countries (Alberini and Cooper, 2000). In this
method, the researcher creates a hypothetical market in a non-market or new good. The
values which are generated through this hypothetical market are treated as estimates of the
value of new good.

After designing the draft questionnaire pre test was conducted with 26 randomly selected
sample households. An open ended question was used for the elicitation of the
respondents’ maximum amount they are willing to pay for the insurance service per
hectare. This is due to make some modifications in the designed questionnaire of the
survey and to obtain starting bid values. Based on this elicitation some values were
selected as the starting bid values for the survey questionnaire. The bid values were
distributed randomly through 161 sample households and the respondents were asked are
you willing to pay this amount if the respondent says yes or no, finally the single bounded
dichotomous choice question is followed up by an open-ended follow up question. To
avoid biases which were mentioned in the literature part, hypothetical scenario were
presented in the Appendix part 2.

3.4. Method of Data Analysis

The data that had been collected through contingent valuation method has been analyzed
using both descriptive statistics and econometric model. Descriptive statistics such as
mean, percentage, standard deviation and frequency of appearance was used, whereas on
the econometric approach adopted the Tobit model.

22
3.4.1. The Tobit model

A very common problem in microeconomic data may stem from conditions in which the
researcher had information only on the regressors but not on the regressand (amount of
the respondent is willing to pay for rainfall insurance) for some observations. A sample in
which information on the regressand is available only for some observations is known as a
censored sample. When data are censored, the distribution that applies to the sample data
is a mixture of discrete and conditional distribution and the most appropriate model to
analyze such distribution is the Tobit model.

This model is also commonly known as censored normal regression model (Greene,
2003). It assumes that many variables have a lower or upper limit that is known as
threshold value and take on this limiting value for a substantial number of respondents.
For the remaining sample respondents the variable takes on a wide range of values above
the limit. The explanatory variables in the model may influence both the probability of
limit responses and the size of non-limit. The two parts correspond to the classical
regression for the non limit (continuous) observations and the relevant probabilities for
the limit (zero) observations, respectively. Based on the above behavior of the model,
Tobit analysis is appropriate for this study and the formula for the Tobit model is given as
follows:

Following Long (1997), the structural equation of tobit model censored from below can be
expressed as:

. ……………………………………………….. (1)

Where,
Yi = the observed dependent variable, in this case the maximum willingness to pay
the respondent is willing to pay in Birr.
Yi*= the latent variable which is not observable.
Xi = vector of factor affecting willingness to pay.

23
 βi = vector of unknown parameters to be estimated .
= residuals that are independently and normally distributed with mean zero and
constant variance δ2.

The model parameters can be estimated by maximizing the tobit likelihood function of the
following form (Maddala, 1997);

1 Y − β X  −β X 
L = ∏ f  i i i  ∏ F i i  ……………………………………………. (2)
Yi∗ >0 σ  σ  Yi∗ ≤0  σ 

Where f and F are respectively, the density functions and cumulative distribution function
of > 0 means the product over those i for which > 0, and < 0 means the
product over those i for which < 0.

Maddala (1997) proposed the following techniques to decompose the effects of

explanatory variables into the decision to pay and intensity effects. Thus, a change in X
(explanatory variables) has two effects. It affects the conditional mean of Yi* in the
positive part of the distribution, and it affects the probability that the observation will fall
in that part of the distribution. Similar approach will be used in this study.

• The marginal effect of an explanatory variable on the expected value of the

dependent variable is:
……………………………………………… (3)
Where, is denoted by z, and F is cumulative distribution.

• The change in the probability of willingness to pay as independent variable Xi

changes is:
………………………………………….. (4)

• The changes in the amount of money respondent are WTP with respect to a unit
change in an explanatory variable among those who are willingness to pay are:

24
 .................................. (5)

Where;
F (z) = is the cumulative normal distribution of z,
f (z) = is the value of the derivative of the normal curve at a given point (unit normal
density),

z = is the z score for the area under normal curve,

β = is a vector of Tobit Maximum Likelihood estimates and
σ = is the standard error of the error term.

3.4.2. The logit model

In the logit model of single bounded dichotomous format, households are given initial bid
value in which they may accept or reject. In the logit model the dependent variable is
dummy variable yes/no.The purpose of the Logit model is to estimate the mean WTP.
Following Gujarati, (1999) the Logit model is expressed as follows:

Logit (P(x)) = ……………………………………………. (6)

Where:
= probability that a given household is willingness to pay
= Constant term
= regression coefficient to be estimated or Logit parameter
= initial bid value
= error term of the Logit regression

One of the main objectives of estimating an empirical WTP model based on the CV survey
responses is to drive a central value or mean of the WTP distribution Hanemann et al
(1991). According to Gujrati (1999) both probit and logit models provide similar results
thus, for comparative computational simplicity logit model was used for the estimation.
And the mean willingness is formulated as:

25
E (WTP) = ……………………………….. (7)

Where:
= bid coefficient

= constant term

3.5. Variable Definitions and Hypothesis

Dependent variable

The amount of money the respondent is willing to pay for rainfall based insurance service
per hectare was taken as the dependent variable.

The independent variables

It is very important to identify the potential explanatory variables and describe their
measurements in a model. Therefore, based on review of theoretical and empirical works,
Socio-economic characteristics of the households and institutional factors were considered
as in the model.

Age of household (AGE): Age is continuous variable defined as the age of the head of
farm household at the time of interview measured in years. According to the study by
Patrick (1988) the age of the household has negative effect on the demand for insurance.
The other study which has almost similar result with Patrick is Gine et al. (2007), who
found that young farmers are more likely to purchase insurance than elders. Therefore, in
this study it is hypothesized that young farmers are more likely to purchase insurance than
elders.

Sex of household (SEHH): This is measured as a dummy variable taking the value of 1 for
male headed household and 0 otherwise. The sex of the household head was included to
differentiate between male and female household heads in their participation of making a
decision on income distribution. In this study it is hypothesized that male head households are

26
likely to purchase the insurance service than female head households. Therefore, it is expected
to affect willingness to pay for rainfall based insurance positively.

Marital status of household (MRST): Marriage is social engagement to support each

other both socially and economically. Married households put aside some of resources for
unforeseen circumstances to smoothen their life, pool their resources and reduce cost that
would have been spent separately. In this study marriage and willingness to pay for
rainfall based insurance premium are hypothesized to be related positively.

Location of the study area (NAWO): This is dummy variable taking 1, if the study area
is located in Dugda and take 0, if the area is located in Mieso woreda.

Income from crop (FINC): It is a continuous variable expressed in Birr and shows the
amount of income that the household head earned from crop production activities. The
increase in demand for insurance associated with income and it appears that an increase in
an income may create pressure on the household to purchase additional insurance. This is
based on economic theory, which states that individual’s demand for most commodities or
services depend on income (Mbata, 2006). Vince and Joyce (1994) have found that
income of the household has positive impact on the demand for rainfall based insurance.
That is financial security for households with greater income may warrant additional
protection. In this study income from crop is expected to have positive influence on
farmers’ willingness to pay for rainfall based insurance.

Off-farm income (OFINC): It is income from other non farming activities like basketry,
roping etc. It is a continuous variable measured in Birr. A study conducted by Sukurai and
Readon (1997) showed that respondents who received high amount of income from other
non-farm activities are not interested in participating in drought insurance. Therefore
households who have less amount of off-farm income are expected to be more willing to
pay for rainfall based insurance.

Family Size (FSIZE): It is a continuous variable measured in number of people living

under one roof. Higher family size is accompanied with larger household expenditure,
which consequently depletes household cash resources. Sukurai and Readon (1997) have
shown that as size of household increase, demand for insurance decrease. In this study size

27
of household is expected to have negative effect on the willingness to pay for the rainfall
risk insurance.

Dependency ratio (DEPR): This is a continuous variable measured in ratio. It refers to an

increase in working-ratio that reduces the ability to meet subsistence need and also
increase the personal rate of time preference. It also tells us the proportion of household
members who are dependent on the economically active members of the family. The more
dependency ratio in the household, the less active labor force the family would have.
Paulos (2002), has found that it is negatively related to willingness on decision of the
farmers to participate in soil and conservation practices. Therefore in this study it is
expected to affect the expected willingness to pay for rainfall based insurance service
negatively.

Education of household head (EDUC): It is dummy variable taking 1 if the respondent

is literate and 0 if the respondent is illiterate. Education may increase farmers’ ability to use
information as well as practice. Education has been shown to be positively related to farmers’
willingness to pay for willingness decision of the farmers to participate in soil and
conservation practices (paulos 2002). Therefore, it is hypothesized to have a positive
influence on farmers’ willingness to pay for rainfall based insurance.

Availability of public and private aid (PAPA): Gifts may be in kind or in cash from
governmental and other nongovernmental organizations. This is a dummy variable takes
the value 1, if households have gift from different sources 0, otherwise. The result from
the study by Sukurai and Readon (1997) showed that as the availability of public and
private aid is high, participation and willingness to pay for insurance is low. Therefore in
this study availability of such aid expected to have a negative influence on the willingness
to pay for rainfall based insurance.

Credit constraint (CREDIT): It is dummy variable which takes the value 1, if the
household has high credit constrained and 0, if less constrained. A study conducted by
Gine et al. (2007) indicates that insurance participation is higher when households are less
credit constrained. In this study credit constraint is expected to have a negative effect on
the demand for insurance and willingness to pay for it.

28
Extension service access (EXTENTION): It is a dummy variable which takes a value of
1, if the farmer has access to extension service and 0 otherwise. Access to extension
service indicates to the availability and existence of technical advices to stallholder
farmers in the study area. Extension service widens the farmer knowledge with regard to
use of improved seed and agricultural technologies. And has positive impact on household
farm and decision for willingness to pay for rainfall based insurance. (Paulos, 2002) have
found Extension access to farmers influenced the application of soil conservation
technologies positively. In this study it is hypothesized that expected to affect willingness
to pay positively.

Initial bid value (BID): This is continuous variable measured in Birr and included in the
regression analysis to check weather starting bid bias exist or not. If this variable is
significant and positive there is a bias on the starting bid value otherwise not.

Livestock holding (TLU): It is a continuous variable which represent livestock holding of

the respondent in tropical livestock unit. It is expected to influence the willingness to pay
of the household head either positively or negatively. This is because of the fact that the
income from sale of livestock as well as production of livestock have positive influence on
income and in turn income has positive influence for willingness to pay. Therefore, On the
other hand, it may have negative impact on willingness to pay if the farmers believe that,
as the willing to pay amount increases they might have to shift their attention from crop
production to livestock production activity. In this study, it is expected to have negative
influence on the willingness to pay for rainfall based insurance.

Ownership of radio of the household (RADIO): This variable is a dummy variable,

which takes the value of 1 if household has radio and 0 otherwise. Radio is a source of
information and can enhance the ability of farmers’ access to different sources of
information such as extension service, credit service, use of new technologies, improved
seed varieties, input price, output price, crop protection, post harvest handling techniques
than those farmers don’t possess radio. Thus, farmers who have radio might be able to
understand those information’s earlier than those who do not have. Therefore, in this study
it is hypothesized that owner of radio will be positively related to willingness to pay for
rainfall based insurance service.

29
House type of the households (HOUSE): It is a dummy variable that takes the value 1 if
the household has iron roofed house and takes the value 0, if they have grass roofed house.
Since type of house is a proxy for wealth status it may have positive influence on the
farmers’ willingness to pay for rainfall based insurance. In this study, it is expected to
have positive influence on the WTP.

30
Table 2.Variables and their measurement included in the mode

Variable Code Type of variable Definition and Measurement

Age of the household AGE Continuous Age of household head in years
Sex of the household SEHH Dummy Sex of household head 1, if male 0, otherwise
Marital status MRST Discrete Marital status 1, married 2, single 3, divorced
Location of the study area NAWO Dummy Location of the study area 1, if in Dugda 0, in mieso
Family size FSIZE Continuous Number of family members
Income from crop FINC Continuous Total annual income of the households from crops in Birr
Education of the household EDUC Dummy Education status of HHH, 1 if literate 0, otherwise
Off-farm income OFINC Continuous Total off-farm income measured in Birr
Initial Bid value BID Discrete Initial bid value offered in Birr per hectare
Maximum willingness to pay MWTP Continuous Maximum WTP in Birr per hectare
Credit constraint CREDIT Dummy 1 if the household is highly credit constrained 0, otherwise
Extension service EXTENTION Dummy 1 service user 0, otherwise
Livestock holding TLU Continuous measured in tropical unit
Dependency ratio DEPR Continuous measured in Birr
House type HOUSE Dummy 1 if iron roofed 0, otherwise
Owning radio RADIO Dummy 1 if owning radio 0, otherwise
Availability of public and private aids PAPA Dummy 1 if household head has aid 0, otherwise
 4. RESULTS AND DISCUSSION

The number of sample households included in this study was 161. Both descriptive and
econometric analyses were used in analyzing the data obtained from the survey.

4.1. Descriptive Statistics Results

Descriptive statistics such as mean, minimum and maximum values, range and standard
deviations were used to describe the major factors explaining farmers’ willingness to pay
for rainfall risk insurance. In addition, mean difference for continuous variables and
frequency of discrete variables were tested using t-test and chi-square test respectively.

4.1.1. Household characteristics

From the total surveyed respondents 144 (89.4%) were willing to pay for rain fall based
insurance where as the rest 17 (10.6%) were not-willing to pay for the service. Based on
the survey result, of the interviewed households 155 (96.3%) were male respondents while
the remaining 6 (3.7%) were female respondents. Out of willing respondents, 141 (97.9%)
were male respondents and 3 (2.1%) were female respondents, while out of non-willing
respondents 14 (82.4%) were males and 3 (17.6%) were female respondents. The result of
chi-square test shows that there is statistically significant difference in sex of household
heads between willing and non-willing groups (p<0.01).

Of the total respondents, 147 (91.3%) were married, 7 (4.3%) were single, 5 (3.1%) were
divorced and 2 (1.2%) were widowed. Out of the willing respondents, 133 (92.4%), 7
(4.9%), 3 (2.1%), and 1 (0.7%) were married, single, divorced and widowed, respectively.
While out of the non-willing respondents 14 (82.4%) were married, 2 (11.8%) were
divorced and 1 (5.9%) were widowed. There is statistical significant difference in marital
status between willing and non-willing groups (p<0.05).

Of the total household surveyed 42.2% have iron roofed house and the rest 57.8% have
grass roofed house. There is statistically significant difference between willing and non-
willing households in terms of their housing type. The education level of the sample
respondents was categorized in to those who can read and write as literate and those who
cannot read and write as illiterate. Based on this, illiterate respondents constituted 73
(45.3%) of the total respondents and the literate groups constituted 88 (54.7%). Out of the
willing respondents 62 (43.1%) were illiterate and 82 (56.9%) were literate and the chi-
square test depicted that there was statistically (p<0.1) significant difference in education
of the household head. Sample respondents who have their own radio were 132(82%) of
the total respondents. Out of this 128(88.9%) are from willing to pay group, and 4(23.5%)
are from non-willing group. There is also statistically significant difference between the
two groups (p< 0.01). The summery of the result is shown below in Table 3.

33
Table 3. Characteristics of sample household heads by willing and not-willing groups for
(dummy variables).

Variable Willing Not-willing Total

to pay to pay
N % N % N %
SEHH Male 141 97.9 14 82.4 155 96.3 10.266***
Female 3 2.1 3 17.6 6 3.7
MRST Married 133 92.4 14 82.4 147 91.3 8.876**
Single 7 4.9 - - 7 4.3
Divorced 3 2.1 2 11.8 5 3.1
Widowed 1 0.7 1 5.9 2 1.2
HOUSE Iron roof 65 45.1 3 17.6 68 42.2 4.710**
Grass roof 79 54.9 14 82.4 93 57.8
EDUC Literate 82 56.9 6 35.3 88 54.7 2.876*
Illiterate 62 43.1 11 64.7 73 45.3
RADIO Yes 128 88.9 4 23.5 132 82 43.983***
No 16 11.1 13 76.5 29 18
Source: own survey, 2010
***,**,* Statistically significant at 1%, 5% and 10% probability levels respectively.

Regarding the continuous variables mean age of the respondent was found to be 39.84
with the minimum 20 and maximum of 71 years. The mean age for willing respondents
was found to be 40.03 with 20 minimum and maximum of 71 years while that of the non-
willing was 38.29 with minimum of 22 and maximum of 70 years respectively. There is no
statistically significant difference between willing and non-willing respondents. The
average family size was found to be 5.86 with a minimum of 1 and a maximum of 11
family members. The average family sizes of the willing respondents and non-willing
respondents were 6.05 and 4.24, respectively. The result indicates that there is statistically
significant difference at 1% significance level between willing and non-willing
respondents in their family sizes (p<0.01).

34
Table 4. Characteristics of sample household heads by willing and not-willing groups for
(continuous variables).
Willing to Non-willing
pay to pay Total t-value
Variables Mean Std. Mean Std. Mean Std.
AGE 40.03 11.94 38.29 14.79 39.84 12.2 0.552
FSIZE 6.05 2.306 4.24 1.2 5.86 2.28 5.198***
Source: own survey, 2010
*** Statistically significant at 1% probability level

4.1.2. Resource ownership

The survey result showed that income from farm is the main source of subsistence for the
majority of the surveyed households. The mean income of the respondents was 7035.40
birr/year with minimum 700 birr/year and maximum of 19,000 birr/year. The mean
income of willing and not-willing respondents was 7237.50 birr/year and 5323.53
birr/year, respectively. The result shows that there is statistically significant mean
difference between the two groups of the respondents. On the other hand, the mean annual
off-farm income of the respondents from different activities was 773.29 birr/year. The
corresponding figure for willing respondents was 570.14 birr/year and the mean of non-
willing respondents was 2494.12 birr/year. There is statistically significant difference
between two groups (p<0.01).

Table 5. Economic Characteristics of sample household by willing and not-willing groups

for (continuous variables)

Willing to Not-willing Total t-value

Pay to pay
Variable Mean Std. Mean Std. Mean Std.
FINC 7237.5 4679.1 5323.5 4534.4 7035. 5468.7 1.600
OFINC 570.14 1376.1 2494.1 2237.6 773.1 1595.3 -3.469***
TLU 7.3141 4.4872 1.0906 3.4249 7.693 4.5161 -3.19***
Source: own survey, 2010
***, Statistically significant at 1% level respectively

35
4.1.3. Institutional characteristics

Households’ institutional characteristics have important effect on the households preferred

status with respect to willingness to pay for rainfall risk insurance. The important
institutional factors included in the study are: agricultural extension, credit facility, and
availability of public and private gifts.

Of the total households surveyed only 88.8 % had contact with extension agents (Table 6).
There was statistically significant difference between the willing and non-willing
households in their access to extension services (p<0.1). On the other hand, 28%, the
respondents reported to have obtained different public and private aids and the rest 72%
did not have this opportunity. There was also statistically significant difference between
willing and non- willing households (p<0.01). About 59% of the household were reported
that they were credit constrained. The result from chi-square test shows there is
statistically significant difference between willing and non-willing households in their
credit access (p<0.05).

Table 6. Institutional characteristics of sample households by willing and not-willing

groups for (dummy variables)

Willingness to pay Not-willing to pay Total

Variable Yes (%) Yes (%) Yes (%)

PAPA 22.9 70.6 28 17.159***
CREDIT 43.8 17.6 41 4.283**
EXTENTION 90.3 76.5 88.8 2.919*
Source: own survey
***,**,*Statistically significantly at 1%, 5% and 10% level respectively.

An attempt has also been made to compare respondents grouped based on the two survey
woredas Dugda and Mieso among dummy and continuous variables. Table 7 presents the
summary of the descriptive statistics.

36
Table 7. Summary of descriptive statistics of sample households’ characteristics by
woreda for (continuous variables).
Dugda Misso t - value
Variables mean Std. mean Std.
AGE 41 13.1 38 10.5 1.639**
FINC 6764 4643 7480 4763.9 -0.940
OFINC 721 1475.8 859 1783.5 -0.531
FSIZE 6.25 2.5 5.21 1.704 3.125***
TLU 7.76 4.52 7.58 4.545 0.243
Source: survey result (2010)
***,** Statistical significant at 1% and 5% probability level.

The purpose of these comparisons using descriptive statistics by woreda was to examine
whether there is significant difference between the two woredas in terms of household
characteristics, institutional factors and resource ownership. The above Table (Table 7)
presents the mean comparison of continuous variables between two woredas. Age of the
household head and family size were found to be statistically significant at 5% and 1%
significant levels respectively. On the other hand no statistically significance difference
was observed between the two woredas in terms of total income from crop, total off-farm
income and total livestock holding.

Summary of descriptive statistics for dummy variables presented in Table 8 also depicts
that there is statistically significant difference at less than 1% significant level in two
woredas of the study area includes: sex of the household head, marital status of the
household head, education of the household head, housing type, access to extension
services, owning radio and credit constraint. But there is no statistically significant
difference in availability of public and private aid in between two woredas.

37
Table 8. Summary of descriptive statistics of sample household characteristics by woredas
for (dummy variables).
Dugda Mieso

Variable N N
SEHH Male 100 55 10.217***
Female - 6
MART Married 94 53 13.343***
Single 6 1
Divorced - 5
Widowed - 2
EDUC Literate 68 20 18.957***
Illiterate 32 41
HOUSE Iron roofed 55 14 17.625***
Grass roofed 45 48
EXTENTION Yes 95 48 10.152***
No 5 13
PAPA Yes 29 16 0.144
No 79 45
RADIO Yes 89 43 8.788***
No 11 18
CREDIT Yes 58 8 31.557***
No 42 53
________________________________________________________________________
Source: survey result (2010)
*** Statistically Significant at 1% Probability level.

4.2. Sources of Risk and Management Strategies Practiced

Farmers were highly affected by many sources of risk, but they were trying to cope and
live with these risks. Households in the study areas were also practicing different
mechanisms in order to make their living. Among many sources of risk, the following
were identified as the major ones by respondents. Households were asked to list the most
important, second most important and the third most important sources of risk that they
faced. Responses were classified into the categories listed below.

38
Table 9. Major sources of risk as perceived by sample respondents and their rank given by
sample households

Rank (=161)
Sources of risk 1st 2nd 3rd
Drought/erratic rainfall 156 10 18
Crop disease 1 128 66
Loss of livestock 1 6 2
Loss of fertility of the soil 2 8 71
Price variability 1 7 2
Low market demand - 1 2
Fire - - -
Flood - 1 -
Source: own survey, 2010

Table 9 shows, clearly that the most important source of risk identified by the respondents
was drought or erratic rainfall. The second was crop failure due to crop diseases and the
third reason was loss of fertility of the soil. Therefore, Drought was the major source of
risk in the study area.

4.2.1. Risk management strategies

In order to cope with sources of risks below in table 10, rural households have developed
through time various risk management strategies which only differ from place to place,
and among the farmers. Farmers in the study area practice sale of livestock as a major risk
coping strategy. Diversification, use of improved technology, delay in sale of crop and
intercropping were also strategies used by farmers. There is statistically significant
difference in coping strategies between willing and non-willing respondents in terms of
diversification, off-farm employment and use of improved technologies. But there is no
statistically significant difference in terms of intercropping, go for credit, delay in sale of
crop and sale of livestock.

39
Table 10. Risk management strategies practiced by sample households

Willingness to Not-willing to
pay pay
Management strategies Yes (%) Yes (%)
Intercropping 66.7 76.5 0.668
Diversification 81.9 64.7 2.837***
Off-farm income 11.8 70.6 35.58***
Go for credit 75.5 58.8 0.279
Delay sale of crop 2.1 - 2.244
Contract sale 80.6 52.9 0.3461
Use of improved technology 85.4 88.2 6.679***
Sale of livestock 85.4 29.4 0.099
Source: own survey, 2010
***, significant at 1%

4.2.2. Risk perception of sample households

Households in the study area perceive that they are exposed to different types of
substantial risks from different sources. Therefore, based on the results obtained from
formal survey questionnaire, households define risk in three ways: year when rainfall
delays, year when rainfall is inadequate, year when rainfall is high. The summary of the
result are presented and discussed below in Table 11.

Table 11. Definition of risk by sample households

________________________________________________________________________________________________

Description Willing Not-willing____

N % N %
Year when rainfall delays 27 18.7 - -
Year when rainfall is low 116 80.6 17 100
Year when rainfall is high 1 0.7 - -
________________________________________________________________________
Total 144 100 17 100
Source: survey data, 2010

40
Out of the total households surveyed 80.6 % define risk as a situation where the expected
rainfall is low. The others 18.7% and 0.7% percent of the sample households define it as
the situation when the rainfall delays and the expected rainfall is high, respectively.
Similarly, when the households were asked which was the most risky year in the ten years
preceding survey, 105 (65.2%) of the sample households reported the year 2001 the others
47 (29.2%), 3(1.9%), 2(1.2%) identified the years 1995, 2000 and 1994, respectively.

Out of the willing respondents 131 (91%) showed their interest to pay the premium in cash
and the remaining 13 (9%) in kind. The respondents also discussed about the time of
insurance premium payment. Of the willing respondents, 129 (89.6%) reported that the
preferred to pay after the time of harvesting. Their main reason stated for this preference
was the prospects of earning money after harvest by selling what they produced. Ninety
four (65.28%) of the respondent choose the indemnity to be paid in cash, because they
need the money to buy better improved varieties, water pump (hand pump), implement
and engaged in fattening activities etc. The other fifty (34.72%) stated that they would
opposite for the payment in kind because they are afraid of losing the money without any
activity, so that they prefer only the lost product or grain.

4.3. Willingness to Pay Analysis

Before implementing the final survey, the pilot survey was conducted using open-ended
elicitation format to set up starting bid. The bid values were 50, 100, 150 and 200 based on
the pilot survey. The follow-up question was open-ended; if the respondent answered “no”
to the randomly aligned initial bid, he/she was then asked how much he/she would pay for
the service. If the respondent answered “yes” to the randomly aligned initial bid, he/she
was then asked what was the most he/she would pay for the service.

The total sample households were randomly distributed to the four initial bid value groups
and each contains 41, 47, 36 and 37 respectively. Out of the total sample respondents
17(27.2%) responded “no” to the initial bid value. The main reason they have stated
includes mainly they couldn’t afford it, and they didn’t trust the service. But the rest
144(72.8%) show their interest to contribute and gave the “yes” or “no” response to the
initial bid value then follow up values.

41
Table 12. Maximum WTP and percentage distribution of the sample household
Maximum WTP Number of Percentage of the
Birr/hectare respondents Respondents
0-50 39 24.3
51-100 54 33.6
101-150 11 6.8
151-200 37 23
201-250 7 4.3
251-300 7 4.3
301-350 5 3.1
351-400 1 0.6
Total 161 100
Source: own survey, 2010

4.3.1. Estimation of the mean WTP value

The initial bid value was regressed with the dependent dummy variable, the result of the
coefficients were presented in Table 13, and mean willingness to pay for the single
bounded dichotomous format is as follows.

Table 13. The Logit model to calculate the mean WTP

Variables Coefficient St. d t-value p-value
CONST 3.993626 0.6619955 6.03 0.0000
BID -0.021873 0.0042091 -5.20 0.0000

E (WTP) =

Where:
= 0.021873

= 3.993626

= 183.41

42
Thus the mean willingness to pay calculated from the single bounded dichotomous format
is 183.41 birr per hectare. However, the mean WTP is 129.93 birr per hectare from
responses to the open-ended CV survey questions, which is lower than the mean value
obtained from the closed-ended Logit model estimates. Thus the result showed that the
respondents were willingness to pay between the ranges of 129.93–183.41 Birr per hectare
for the proposed rainfall based insurance service.

4.3.2. Estimating total willingness to pay and total revenue

In this section the total willingness to pay and total revenue at different prices that
households in the seven PAs of the two woredas (Dugda and Mieso) were willing to pay
as computed. The sampled seven PAs namely, (B/Gusaa, Odd Bokota, Jawe Bofo,
S/wakalee, Huse mandhera, Chobi, Burimulu) have a total of 3281 households with a total
population of 49,966 households with a total population of 275,307 and an average family
size of 5.86. Based on this information and the distribution of WTP amount by the
respondents, it would be possible to estimate the expected total willingness to pay and
total revenue for the study area. Table 14 provides the procedure and results from this
analysis.

43
Table 14. Total willingness to pay and total revenue in (Birr)

Class bound. Class Sample Total no of Total WTP in Birr Sample HHs WTP Total HHs Total Revenue
For WTP mark for District of HHs at least that WTP at
amount WTP HHs amount least that
amount amount
N % N %
0-50 25 39 24.223 12,103.3 302,582 161 100 49,966 1,249,150
51-100 76 54 33.54 16,759.6 1,273,729 122 75.78 37,864 2,877,664
101-150 126 11 6.832 3,414.18 430,186 68 42.24 21,105 2,659,230
151-200 176 37 22.981 11482.7 2,020,953 57 35.40 17,688 3,113,088
201-250 226 7 4.347 2,172.02 490,877 13 8.07 4032 911,232
251-300 276 7 4.347 2,172.02 599,478 11 6.83 3413 941,988
301-350 326 5 3.105 1551.44 505,771 6 3.73 1864 607,664
351-400 376 1 0.621 310.29 116,669 1 0.62 310 116,560
Total 161 100 49,966 5,740,244
Source: Own survey, 2010
The first column shows the maximum willingness to pay interval, and the second is class
mark for willingness to pay (the mid willingness to pay amount) of the first column. The
third and the fourth columns show the number and the percentage of sample households
whose willingness to pay amount falls within the given interval.

The total number of households in two woredas of the study area has been multiplied by
the proportion of sample households falling in each category to obtain the total number of
households whose willingness to pay amount lies in each boundary (column fifth). And
total willingness to pay (column sixth) has been obtained by multiplying the mid
willingness to pay amount by total number of households willingness to pay that amount.
The total household of 49,966 in two woredas of the study area were expected to pay birr
5,740,244 per year, if every household insures one hectare of his land.

Therefore, the result of the survey indicates that the average insurance premium payment
of household was birr 114.88 per hectare per household if the proposed insurance service
is implemented. This result is almost similar with the average willingness to pay of 129.93
birr per household per hectare. A column seven and eight indicates the number and the
percentage of sample household willingness to pay at least the amount in each interval.
Similarly, column nine shows total number of households willing to pay at least the
amount in each interval and it falls as the mid willingness to pay amount rises (column
ten). Total revenue has been obtained by multiplying the mid willingness to pay amount
(column two) by the corresponding total number of households’ willingness to pay at least
that amount, (column nine).

4.3.3. Derivation of aggregate demand

The aggregate demand for this study has been derived from the above willingness to pay
scenario (Table 14). The aggregate demand curve is derived using the mid willingness to
pay amount along the vertical axis and the number of households’ willingness to pay at
least that mid value per hectare along the horizontal axis, (Figure 2). The figure shows the
aggregate demand curve for the rainfall based insurance using the observations in the
study. Any point on the curve shows all the households that prefer the insurance service
but do not bid more than the corresponding value on the mid willingness axis.
As shown in Figure 2, the demand curve is negatively sloped, indicating the fall of the
demand for the insurance service as the premium increases, like most other non-market
goods other things remaining constant. The area under demand curve represents the gross
value of consumers’ surplus if the service is available for free or zero.

Figure 2. Estimated demand curve for rainfall based insurance service

46
4.4. Econometric Model Result

Econometric software called Limited dependant (Limdep 7) was employed to estimate the
Tobit model. In the tobit model the main objective is to identify factors affecting the
willingness to pay for rainfall risk insurance and its intensity in relation to socio-economic and
demographic variables.

4.4.1. Econometric test and results

Before taking explanatory variables in to the analysis, it was necessary to check the
existence of multicolinarity among the continuous and dummy variables. Variance
inflation factor (VIF) was used to detect mulitcollinarity among continuous variables and
contingency coefficient (C) was used to detect the degree of association among dummy
variables. According to Maddala (1992), VIF can be defined as:

VIF =
Where, R2 is the squared multiple correlation coefficient between xi and other explanatory
variables. A statistical package known as SPSS 16 was used to compute the VIF values.
As a rule of thumb, when the VIF exceeds 10, there is muliticollinarity problem. VIF
values shown in the Appendix Table 1 indicate that there was no serious muliticolinarity
problem. Similarly, contingency coefficients were used to check the existence of
muliticolinarity. Contingency coefficient is computed as follows:

C =

Where:
C = is coefficients of contingency
= chi-square random variable
N= total sample size

The rule of thumb for Contingency coefficient is that when its value approach as 1 and
greater than 0.75 there is muliticolinarity problem between dummy explanatory variables.

47
But the result shown in the Appendix Table 2 revealed that all values were less than 0.75
and there was no serious muliticolinarity problem among dummy variables.

The assumption in regression analysis is that the errors terms, Ui has a constant variance
δ2. If the error term doesn’t have a constant variance, there is problem of hetroscedastics
(Maddala 1992). In the general linear model, OLS estimates are consistence but not
efficient when the disturbance terms are hetroscedastic. In the case of limited dependent
variable models (such as Tobit), the estimates of the corresponding regression coefficient
is upward biased in the presence of hetroscedasticity. But nothing can be said about the
direction of bias. It is more practicable to make some reasonable assumption about the
nature of hetetroscedasticity and estimate the model to say that the maximum likelihood
estimates are inconsistent if hetroscedasticity is ignored (Maddal 1992). The test for the
presence of heteroscedasticity problem in the model was also done by using Breusch-
Pagen test and the result was ρ= 1.250, this shows that there is no heteroscedasticity
problem in the model.

4.4.2. Interpretation of the results

The estimates of the parameters of the variables that were expected to affect the
households’ willingness to pay for rainfall based insurance are shown in Table 15. The
dependent variable was a continuous variable that household response as maximum
willingness measured in birr. Out of the 16 hypothesized explanatory variables, six were
found to be statistically significant, four of them were continuous and the rest two were
dummy variables. The variables were age of the household head (AGE), total income from
farm (FINC), total off-farm income (OFINC), livestock holding (TLU), owning radio
(RADIO), and availability of public and private aid (PAPA). Moreover, the sign of the
estimated coefficients were consistent with the expected signs.

The result has shown that age of household (AGE) is an important factor that influences
the respondent’s willingness to pay negatively and it is statistically significant (p<0.05).
Earlier studies by Patrick (1988) and Gine et al. (2007) have found similar results. As the
age of household head increases, the willingness to pay amount decreases significantly.
Therefore, younger household heads are more likely to be willing to pay for rainfall based
insurance compared to older household heads. This may be explained by the fact that
48
younger household heads have less long life experience on predicting weather conditions
and they are also sensitive to the new technologies than elders. The result shows that for
each additional year in age of the respondent, the probability of the willingness to pay for
rainfall based insurance decreases by 0.548%. The marginal effect result also shows that
as the age of a respondent increase by one year, the amount of cash s/he is willing to pay
for rainfall based insurance decreases by 1.5159 Birr.

Household income from crop (FINC): This variable is found to have a positive impact on
the probability of willingness to pay as hypothesized and the effect is statistically
significant at 1% probability level. Those household heads that generate high income from
crop production would be more willing to pay for rainfall based insurance. When the
income of the household increases by one birr, the probability of the household to be
willingness to pay for rainfall based insurance increases by 0.002%. The marginal effect
result shows that when the income level of the household increase by one Birr, the amount
of cash the household could pay for rainfall based insurance increases by 0.0055 Birr,
other factors held constant at their mean values.

Ownership of radio by the household (RADIO) is another important factor which is

positively and significantly (p<0.05) related to farmers’ willingness to pay for rainfall
based insurance and its amount. Information from radio enhances the ability of farmers’
access to improved technologies and risk management strategies. Farmers that own radio
may get different information on extension service, credit service, improved seed variety,
input prices and output prices than those farmers who do not have radio. This variable also
shows that farmers that own radio have 15.218% more probability of paying for rainfall
based insurance than those farmers who do not possess. Also, the marginal effect of this
variable shows farmers that own radio, would pay Birr 42.0637 more than those farmers
that do not have radio.

As expected the availability of off-farm income (OFINC) is negatively and significantly

related to willingness to pay (p<0.05). Households engaged in off-farm activities reduce
the probability of willingness to pay for rainfall based insurance by 0.003%. Households
who are engaged in off-farm activity are expected to have less attention to farm activity.
Sukurai and Reandon (1997) have found similar result. The marginal effect of this variable
also shows when off-farm income increases by one Birr the amount of cash households
49
would be willing to pay for rainfall based insurance decreases by 0.0098 Birr, other factors
held constant.
Public and private gift (PAPA): This is another important factor which affects the
dependent variable negatively and significantly (p<0.01). Availability of public and
private aid decreases the willingness to pay by 16.233%. Sukurai and Reandon (1997)
have found a negative effect on the dependent variable when farmers have aid from
governmental or other non-governmental organizations, either in kind or in cash, and this
may be explained by the fact that as households become more dependent and less active,
and their willingness to pay tends to be less. The marginal effect of the variable shows that
those household who have some kind of public and private gifts decrease willingness to
pay amount by 44.8686 Birr than those who don’t have the gift, other variables held
constant.

Livestock Holding (TLU): Number of livestock owned by households is found to have

negative and significant (p<0.01) effect on willingness to pay. Each additional unit of
livestock (TLU) decreases the willingness to pay by 1.618%. This implies that income
from livestock may encourage farmers to depend more on livestock than farming and
results in less attention being give to the crop production. The marginal effect shows that
for each additional TLU that the household possess the willingness to pay amount
decreases by 4.4728 Birr, other variables held constant.

50
Table 15. Maximum Likelihood estimates of the Tobit model

Variables Estimated coefficient Standard error t-ratio

Constant 3.70112 103.639 0.0357
NAWO -2.77379 18.8725 -0.14697
SEHH 105.47 62.4742 1.68821
AGE -1.72656 0.71336 -2.42032**
EDUC 7.89432 15.9354 0.49539
MRST 22.2316 22.1423 1.00403
FINC 0.00556 0.00155 3.59743***
OFINC -0.01032 0.00487 -2.12119**
DEPR -0.66475 8.84637 -0.07514
FSIZE 5.18168 3.7561 1.37954
TLU -5.01802 1.64755 -3.04575***
CREDIT -9.49309 15.8865 -0.59756
EXTENTION 24.4899 22.8428 1.0721
PAPA -49.3245 16.0296 -3.0771***
BID 0.05866 0.12318 0.47623
RADIO 33.4783 20.6775 1.61906**
HOUSE 0.68939 14.8145 0.04653
Number of observation = 161
Log likelihood = -854.3120
Threshold value for the model: Lower = 0.0000 Upper = + infinity
δ = 83.4523
Z= 1.48 ф (z) = 0.2859 Ф(z) = 0.9307
________________________________________________________________________
***, **, * indicate significance at 1%, 5%, and 10% levels, respectively
Source: model result, 2010

51
Table 16. Marginal effects of the explanatory variable on the dependent variable

Change in probability Change among willing Total change

Variables
NAWO -1.66800 -2.63030 -2.58250
SEHH 63.4230 100.0140 98.1955
AGE -0.00548 -1.51590 -1.4900
EDUC 0.04564 12.6173 12.4022
MRST 13.3687 21.0861 20.6982
FINC 0.00002 0.0055 0.0054
OFINC -0.00003 -0.0098 -0.0096
DEPR 0.01727 -4.7743 -4.6929
FSIZE 0.01691 4.6763 4.5966
TLU -0.01618 -4.4728 -4.3966
CREDIT 0.02439 -6.7416 -6.6269
EXTENTION 0.09502 26.2641 25.8164
PAPG -0.16233 -44.8686 -46.723
BID 0.00085 0.0235 0.0231
RADIO 0.15218 42.0637 41.4367
HOUSE 0.01919 5.3063 5.2159
Source: based on model out put

52
 5. SUMMARY AND POLICY RECOMMENDATIONS

5.1. Summary and Conclusion

The main objective of this study was to identify factors affecting smallholder farmers’
willingness to pay for rainfall based insurance in Central rift valley of Ethiopia. The study
was designed to identify the variables, which determine farmers’ willingness to pay for
rainfall based insurance.

This study tried to look in to socio economic, institutional and physical and other related
factors which can affect farmers’ willingness to pay for rainfall based insurance. Data
were collected from 161 farm households drawn randomly from Dugda and Mieso
Woredas. The primary data were collected using semi structural questionnaire and the
secondary data were obtained from woredas agricultural office and other non-
governmental organizations around the woredas. Both descriptive statistics and
econometric model were employed to analyze the data. Contingent Valuation Method
(CVM) was employed to elicit farmers WTP for rainfall based insurance. The responses
from the survey were analyzed by using economic software Limdep version 7.

Descriptive statistics were also used to describe risks faced by smallholders’ farmers and
management strategies practiced by the respondents. It also shows that there were
significant differences between willing and non-willing households with respect to some
variables of interest which include: sex, marital status, house type, education, owning
radio, off-farm income, income, family size, credit constraint, availability of public and
private aid and access to extension service at different significant levels.

The result from descriptive statistics also revealed that households suffered from drought
or erratic rainfall, crop disease and loss of fertility of soil respectively. Therefore, they
have practiced different types of coping strategies which included: intercropping,
diversification, off-farm employment, go for credit and delay in sale of crop.

The study used CVM technique to elicit farmers’ willingness to pay for the proposed
rainfall based insurance service. The sampled households were asked questions, related to

53
their socio economic, demographic, institutional characteristics and some general
questions. They were also asked dichotomous question and this were followed by open
ended question to elicit households’ willingness to pay for the proposed rainfall based
insurance service. Of the total sample households 144 (89.4%) were willing to participate
and the rest 17 (10.6%) were not willing to participate. The following bid values 50, 100,
150 and 200 were found from the first open-ended questions. The total willingness to pay
amount for the total of 49,966 households is estimated to be birr 5, 740,244 per hectare per
year.

Sixteen potential explanatory variables were hypothesized to explain farmers’ willingness

to pay for rainfall risk insurance, and they were measured based on the model output. The
result of Tobit model revealed that only six potential explanatory variables were used to
identify willingness to pay among selected sample households at different significant
levels. Among the six potential explanatory variables, three were significant at 1%
probability level and the other three were found to be significant at 5% probability level to
willingness to pay.

Age of the household was found to have a negative and significant impact on farmers’
willingness to pay for rainfall risk insurance at (p< 0.05) level it implying that aged
farmers’ have confident by their own weather condition prediction trend from their long
life experience and they may not trust the insurance service. But younger farmers could
easily decide to take part willingly in proposed insurance service.

Income from crop production was another important, highly significant and positively
related variable that affect willingness to pay for the proposed insurance service at 1%
probability level. Households with more income from their crop production could be
willing to contribute more of their income for the proposed rainfall based insurance
service. This means that income is an important variable affecting the willingness to pay.
Households’ off-farm income is also another significant and important variable which is
found to be negatively related to the willingness to pay for the rainfall based insurance 5%
probability level. The result revealed that households with radio have more information
access on different aspects of both agricultural and non-agricultural sectors. This variable
was positively related to willingness to pay at 5% probability level. This means

54
willingness to pay for the rainfall risk insurance is more related to both information access
and awareness.

Availability of public and private aid was found to be highly significant at 1% probability
level and it was negatively related to the willingness to pay. Households who are more
dependent on governmental and non-governmental organizations aids are not willing to
pay for the proposed rainfall based insurance service payment. The last important variable
is households’ livestock holding it was found to be negatively related to willingness to pay
at 1% probability level. This is also because of households with more livestock number
depending more on the livestock production as their primary activity and may have less
time to care the crop production activity.

55
5.2. Policy Implication of the Study

The overall understanding of factors affecting smallholder farmers’ willingness to pay for
rainfall based insurance would help policy makers and development workers to design and
implement the rainfall based insurance service in sustainable and in effective manner.
Based on the findings of the study, the following points are suggested to be considered as
an important element in order to implement the service and enhance farmers’ rainfall
based insurance utilization and effectiveness in the country.

The strong negative relation between availability of public and private aid and
willingness to pay for the proposed rainfall based insurance revealed that in order
to increase the willingness of household, development policies focused on
sustainable development rather than giving some aid at the time when disaster is
occurred. This can be through organizing farmers in to saving and credit
cooperatives so that they can increase their income and can be self sufficient.

Household income from crop and willingness to pay for rainfall based insurance
were positively related, development policies should target at increasing income of
households that address specially the low income members of the smallholder
farmers. This can also be through facilitating and forming small business groups,
educating and giving awareness to involve in credit and saving cooperatives.

Household off-farm income and willingness to pay for rainfall based insurance
were negatively related. Household willingness to pay for rainfall based insurance
were relatively less if they have off-farm income from other non-farm related
activities this is because of off-farm income is another option for households’
livelihood.

Households who have radio have an access to information as well as they have
awareness on different agricultural activities such as market price, post harvest
management strategies, and information on weather conditions. Development
policies focused on the different Medias to create awareness and understanding
among farmers.

56
 6. REFERENCES

Addis Ababa University, 2001. Introductory geography of Ethiopia Geo,101. Teaching

text, Department of Geography.

Alberini, A., and Cooper, J., 2000. Application of contingent valuation method in
developing countries. Economic and social development paper. No. 146, FAO, Rome.

Anderson, J. R., Dillon,J. L., and Hardaker, J.R., 1977. Agricultural decision analysis:
Lowa State University press.

Anderson, D.R., 2001. All Risks Rating Within a Catastrophe Insurance System. Journal
of Risk and Insurance. 43: 629-651.
Assefa Demeke and G. Ramakrishna, 2002. An empirical analysis of food security in
Ethiopia, the case of North Wollo. African Development Journal. 27 (1&2):127-144.

Bateman, I. J. H. Langford, A. P. Junes, G. B. Kerr and R. Scarpa, 2000. Bound and path
effects in double and triple bounded dichotomous choice contingent valuation. A paper
presented at the tenth annual conference of the European Association of Environmental
Resources Economists (EAERE), Great Britain.

Binswanger, H.P., 1980. Attitudes towards risk. Experimental measurement in rural India.
American Journal of Agricultural Economics. (62): 395-407.

Bryan, E., T. Deressa, G. Gbetibouo and C. Ringler (2009). Adaptation to climate change
in Ethiopia and South Africa: options and constraints. Environmental Science and Policy,
12: 413-426.
Carson, R.T., N.E. Flores and N.F. Meade, 2001. Contingent Valuation: Controversies and
Evidence. Environmental and Resource Economics, 19(2): 173-210.
Carson, R.T., 1991. Constructed Markets. In: J.Barden and C.Kolstad (eds). Measuring the
Demand for Environmental Commodities. Amsterdam, North Holland.
Cooper, J. and T. Osborn, 1998. The Effect of Rental Rates on the Extension of
Conservation Reserve Program Contracts. American Journal of Agricultural Economics
80, 184-194.
CSA (Central Statistics Authority), 2004. Agricultural sample survey report Federal
Democratic Republic of Ethiopia, Addis Ababa, Ethiopia.

Dandekar, V.M. 1977. Crop insurance for developing countries. Teaching and Research
Forum Paper No. 10. New York: Agricultural Development Council.

Dennis Holden, Peter Hazell, Anthony Pritchard, 1991. Risk in Agriculture. Proceedings
of the Tenth Agriculture Sector Symposium. World Bank Washington, D.C. USA.

Dillon J.L and J.B. Hardaker, 1993. Farm management research for small farmers
development. FAO Farm Management Series. FAO, Roma Italy. (6):302

57
Freeman, A.M., 2003. The Measurement of Environmental and Resource Values: Theory
and Methods. Washington, D.C.

Gautam, M., Hazell, P., and H. Alderman, 1994. Rural demand for drought insurance,
world bank Policy Research Working Paper No. 1383.

Gine Xavier, Robert Towensend and James Vickery, 2007. Patterns of rainfall insurance
participation in rural India. Policy research working paper. 4408.

Greene, W. H., 2003. Econometric Analysis. Fourth Edition, Prentice Hall International,
Inc. USA.

Gujarti, D., 1999. Essential of Econometrics. Second edition, Mc-Graw hill companies.

Halcrow, H.G. 1948. The theory of crop insurance. Ph.D. diss. Chicago: University of
Chicago.

Hanemann, W.M., 1991. Willingness to accept and willingness to pay and how much can
they differ? American Economic Review, 81(3): 635-647.

Hanemman, W.M & Kanninen. B., 1998. The statistical analysis of discrete response
data. Department of Agricultural, Resource Economics and Policy, Working paper No,
798 University of California, Berkeley.

Hardaker, J. Banri., Ruud B.M. Hurnie and Jock R.Anderson, 1997. Coping with risk in
agriculture. CAB International London.1-10p.

Hauseman, J., 1993. Contingent Valuation: A critical Assessment, North Holland

Publishing Co., Amsterdam.

Hazell, P.B.R., 1992. The appropriate role of agricultural insurance in developing

countries. Journal of International Development. (4):567-581.

Hazell, P, J. Anderson, N.Balzer, A.Hustarp Celmmensen, U.Hess and F.Rispoli. 2010.

Potential for scale and sustainability in weather index insurance for agricultural and
rural livelihood. International Fund for Agriculture Development and World Food
Program, Roma.

IAC. 1986. Crop and rainfall insurance. Report No. 393. Canberra: published for the
Industries Assistance Commission by the Australian Government Publishing Service.

IPMS (Improving Productivity and Market Success), 2006. Pilot searing site of Mieso
woreda. IPMS, Adiss Ababa, Ethiopia. http//www. IPMS-Ethiopia.org.

Hellmuth M.E., Osgood D.E., Hess U., Moorhead A. and Bhojwani H (eds), 2009. Index
insurance and climate risk: Prospects for development and disaster management. Climate
and Society No. 2. International Research Institute for Climate and Society (IRI),
Colombia University, New York, USA.

58
Long, S., 1997. Regression Models for Categorical and Limited Dependent Variables.
Saga Publications.

Maddala, G.S., 1992. Introduction on Econometric. Second edition . Macmillan Publishing

Company, New York.

Maddala, G.S., 1997. Limited Dependent and Qualitative Variables in Econometrics.

Cambridge University Press, Cambridge.

Marothia, D.K., 2001. Valuation of Day Use Recreation site: Application of Alternative
Estimation Techniques. Indian journal of Agricultural Economics. University of Gulbarga.
56(3) 313-324.

Mbata, J., 2006. Estimating Household Willingness for Water Service in Rural Economy:
the case of Kenya, Southern Botswana. Development of Southern Africa, 23 (1):29-43.

McCarthy, 2003. Demand for rainfall-index based insurance: A case study from Morocco
International Food Policy Research Institute, U.S.A.

Mithcele, R.C. and T.R. Carson, 1989. Using survey to value public Goods: The
contingent valuation method. Resources for the future, Washington, D.C.; The Johns
HopkinsUniversity Press, Baltimore, U.S.A.

MoWR (Ministry of Water Resource of the Federal Democratic Republic of Ethiopia)

2007. Climate Change Adaptation Program of Action (NAPA) of Ethiopia. Addis Ababa,
Ethiopia.

Patrick G. F., 1988. Mallee Wheat farmer’s demand for crop and rainfall insurance,
Australia Journal of Agricultural Economics, 32 (1): 37-49.

Paulos, A., 2002. Determinants for Farmers' willingness to participate in soil conservation
practices in the highlands of Bale: The case of Dinsho Farming System Area. M.Sc. Thesis
Presented to the School of Graduate Studies of Alemaya Universty.

Sakurai, T and T. Reardon., 1997. Potential demand for drought insurance in Burkinafaso
and its determinants. American Journal of Agricultural Economics. 79 (4):1193-1207.

Sarris, A. 2002. The demand for commodity insurance by developing country agricultural
producers: Theory and an application to cocoa in Ghana. Working Paper, DECRG and
RDV. Washington, D.C.

Siegel, P.B. and J. Alwang., 1999. An asset based approach to social risk management: A
conceptual approach, Social Protection Discussion Paper 9926, World Bank,
Washington DC.

Skees, Jerry R. 2001, The Bad Harvest, Regulation, the CATO Review of Business and
Government, 24: 16-21.

Stern, N.H., 2007. The economics of climate change: the. Cambridge University Press.

59
Storck, H., Bezabih Emana, Berhanu Adnew , Borowiecki, A.A and Shimelis w/ Hawariat,
1991. Farming systems and Farm Management Practices of Small-holders in the Hararghe
Highlands. Farming System and Resource Economics in the Tropics.11: Wissenschafts
Varlag Vauk Kiel KG, Germany

Thornton, P.K., P.G. Jones, T.M. Owiyo, R.L. Kruska, M. Herero, P. Kristijanson, A.
Notenbaert, N. Bekele, 2006. Mapping Climate Vulnerablity and Poverty in Africa. Report
to the Department for International Development, ILRI, Nairobi.

Ulrich Hess and Joanna Syroka, 2005. Weather based insurance in Southern Africa,
Malawi. Agricultural and rural development discussion paper 13.

Vince E. Showers, and Joyce. Shotick, 1994. The effect of household characteristics on
demand for insurance; a tobit analysis. The Journal of Risk and insurance, Vol 61, (3):
492-502.

Walker, T.S and James G.Ryan, 1990. Village and household economies in India’s semi-
arid tropics. Baltimore, Maryland, USA: The Johns Hopkins University Press.

World Bank, 2000. Mexico Rural Finance, Saving Mobilization Potential and Deposit
Instruments in Marginal areas, Washington DC.

World Bank, 2005. Managing agricultural production risk innovation in developing

countries. Report No. 32727-GLB. The International Bank for Reconstruction and
Development. Washington , D.C.

World Bank, 2007. UN Department of Economic and social affairs division for sustainable
development innovation briefs. Developing Index-based Insurance for Agriculture in
Developing Countries, New York.

60
7. APPENDIX

61
7.1. Appendix 1. Tables

Appendix 1. Variance Inflation Factor for continuous variables

Variables R2 VIF
AGE 0.32 1.47
TINC 0.12 1.14
OFINC 0.12 1.14
DEPR 0.12 1.14
FSIZE 0.33 1.51
CULAND 0.33 1.49
TLU 0.19 1.24
BID 0.06 1.06

Appendix 2. Contingency Coefficients for dummy variables

________________________________________________________________________
EDUC CREDIT EXTENTION PAPG RADIO HOUSE
EDUC 1
CREDIT 0.29 1
EXTENTION 0.033 0.211 1
PAPG 0.067 0.044 0.089 1
RADIO 0.092 0.094 0.09 0.241 1
HOUSE 0.194 0.203 0.103 0.056 0.073 1
Source own survey, 2010

62
Appendix 3. Conversion factors for tropical livestock unit (TLU)

Livestock TLU(Tropical Livestock Unit

Camel 1.25
Cow 1
Ox 1
Donkey 0.7
Heifer 0.75
Calf 0.25
Goat 0.13
Sheep 0.13
Chicken 0.013
Source: Storck et.al., 1991

63
7.2 Appendix 2. Survey Questionnaire

General information

Name of the woreda _____________________

Name of peasant association _____________________
Name of the village_____________________________
Name of the interviewers________________________

1. household characteristics
1. Name of household_______________
2. Sex a) male b) female
3. Age___________
4. Educational level__________
5. Marital status
a) Married
b) Single
c) Divorced
6. Number of years that the has been living in this region_________
7. Do you have some social position in the community so far? a) Yes b) no
8. If yes, type of responsibility_______________
9. When did you start your own farming ___________years ago?
10. Do you have any vocational training on agriculture a) yes b) no
11. If yes, type of training______________
12. How much is your off-farm income in Birr/ year __________

64
13. Yield of major crops grown in 2001/2002 production year
Crop type
Maize
Wheat
Teff
Sorghum
Barely
Beans
Vegetable

14. Age, sex and education of family members

No Name Sex Age Education level

2. resource ownership and tenure

15. Ownership of resources

Resource 1=yes 0=no

Type of house owned 1=grass roofed

2=iron roofed

Plowing tools (mofer,kenber,maresha)

Animal cart

65
16. How much is your total farm in hectare

Particulars Land holding (Current season) (in ha)

Total Cultivated

Land Land

1. Self owned

3. Rented in

4. Rented out

5. Share cropped

6. Other (eg. Family land,)

17. Livestock holding pattern

LIVESTOCK number owned (0 = none)

Cow

Ox

Calf

Heifer

Bull

Donkeys

Goat

Sheep

Camel

Chickens

66
18. Other materials
Does your household own 1= Yes , 0= No

Radio

TV

Fan

Land-line phone

Mobile phone

Bicycle

Motorbike

Water pump

Car/truck/tractor

3. Production pattern of major crops

19. Production pattern of major crops ( ha)

Last season Good rain season Bad rain year

crops
(on average) (on average)

Area Production Area Production Area Production

(ha) (quintal) (ha) (quintal) (ha) (quintal)

Maize

sorghum

Teff

Beans

Wheat

Barely

Vegetable
(onion, pepper,
tomato,cabbage)

67
 4. Access to rural services
20. Source of technical advice on
Type of advice Source of advice

1.crop production technology(seed, fertilizer)

2.Crop management(pest, farm handling, agronomy

,etc)

3.post harvest

4. policy

1. Mass media, 2.Self (market), 3. Trades/ buyers 4. Research centers 5. NGOs 6. others

21. Are you credit recipient a) yes, b) no

If yes, specify the type, amount and source of credit

Source of Form of credit (in Amount of Usage of credit(the The

credit cash or in kind) credit(in birr or purpose the credit accessibility of
unit) is used to do) credit

1=NGO 2=local money lender 3= farmer association, 4= other ___________

22.Are there NGOs that work with you?

a. Yes, b. No

23.If the answer for Q 17 is “a”,

1. What is their name___________?

2. What is their role (what are the services they provide)? _______________

68
24. What are the major sources of risk that you face? (Rank as 1st, 2nd, and 3rd)

Risks 1st reason 2nd reason 3rd reason

Drought

Crop failure

Crop disease

Loss of livestock/ disease

Price changes

Fires

Flood

Moisture stress

Low market demand

Loss of fertility of soil

25. What is your management strategies used to reduce risks and risk sources?

Management strategies followed Did you use it?

(yes or no)

Intercropping

grow more type of crops

reduce the type of crops

grow only one type crop

off-farm employment

Go for credit

delay sales

contract sales

improved technologies

sale of livestock

69
26. How you define risky or bad year (s) to your livelihood?

Year when there was no/low rainfall,_____ b) Year where there is delay in rainfall,___

Year when there is high rainfall,_____ d) other

(specify)_________________________________________________________________

Which years during the last 10 years were bad or risky years for your agricultural
production in E.C.? (circle)

a) 1992 b) 1993 c) 1994 d) 1995 e) 1996 f) 1997 g) 1998 h) 1999 i) 2000, j) 2001

Hypothetical Scenario

Climate change and variability will likely bring drier condition and shorter, more intense
rainfall events. Because of this crop production in your area is not as expected In order to
reduce vulnerability of households in rural areas from the impacts of such events,
introduction and adoption of an innovative rainfall based insurance designed to
compensate farmers in case of deficient rainfall is important. We will ask you a question if
you are willing to participate in rainfall based insurance service. Implementing of this
service needs initial investment. The service will be managed by governmental and
nongovernmental organizations collaboration with farmers’ cooperative and development
agent. The amount of money to be paid is decided by the insurance companies. We want
to know if you are willing to pay some money so that the implementation of the service
will be undertaken in your area. And we would now like you to answer the following
questions on the amount of money you are willing to pay for proposed insurance service.

70
 5. Willingness to pay questions for rainfall based insurance

Do you know what agricultural insurance is? 1= Yes 0= No

If Yes explain_________________________________________________

Do you need other agricultural insurance mechanisms if it is? 1=yes 0= no

Would you participate in such insurance scheme if it is introduced 1=yes 0= no

Are you willing to pay? _____ birr/year per hectare? 1=yes 0=no

What is the maximum amount that you are willing to pay to insure a hectare of crop land?
Birr ________

In what way are you able to pay the premium 1) in cash 2) in kind

At what time are you able to pay the premium?

a) Right after harvest

b) At the time of sowing

c) After sowing

d) Others

Why do you choose the time of

payment?__________________________________________________________

___________________________________________________________________

In what way you need to be insured in the time of rainfall failure?

1) In kind or form (the lost commodity)

2) In cash.

3) Others (specify)

If you selected any of the above options, why do you select it?
__________________________________________________________________

Any general comment on rainfall based Insurance?

________________________________________________________________________

________________________________________________________________________
________________________________________________________________________

Thank you

71