DETERMINANTS OF FARMERS’ PARTICIPATION IN COFFEE PRODUCTION AND MARKETING (THE CASE OF OYDA WOREDA IN GAMMOGOFA ZONE SOUTHERN NATIONS NATIONALITIES AND PEOPLES REGIONAL STATE)
MSc THESIS
TENKIR TENKA
JUNE, 2016
ARBA MINCH, ETHIOPIA
1
DETRMINANTS OF FARMERS’ PARTICIPATION IN COFFEE PRODUCTION AND MARKETING (THE CASE OF OYDA WOREDA IN GAMMOGOFA ZONE SOUTHERN NATIONS NATIONALITIES AND PEOPLES REGIONAL STATE)
TENKIR TENKA
A THESIS SUBMITTED TO THE
DEPARTMENT OF ECONOMICS, COLLEGE OF BUSSINESS AND ECONOMICS, SCHOOL OF GRADUATE STUDIES, ARBA MINCH UNIVERSITY IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN ECONOMICS (ECONOMIC POLICY ANALYSIS)
JANUARY, 2016
ARBA MINCH ETHIOPIA
2
DECLARATION
I hereby declare that this M.Sc. thesis is my original work and has not been presented for a degree in any other university, and all sources of material used for this thesis have been duly acknowledged. Name: TENKIR TENKA MAMO Signature: ______
Date: June, 2016
i
ADVISORS’ THESIS SUBMISSION APPROVAL SHEET PAGE
SCHOOL OF GRADUATE STUDIES
ARBA MINCH UNIVERSITY
This is to certify that the thesis entitled “Determinants of Farmers’ Participation in coffee Production and Marketing (The case of OydaWoreda in GammoGofa Zone Southern Nations Nationalities and Peoples Regional State)” submitted in partial fulfillment of the requirements for the degree of Master’s with specialization in Economic Policy Analysis, the Graduate Program of the Department of Economics and has been carried out by TenkirTenkaId. No RMSc/161/06, under our supervision. Therefore, we recommend that the student has fulfilled the requirements and hence hereby can submit the thesis to the department for defense.
______
Name of Principal advisor Signature Date
______
Name of co-advisor Signature Date
ii
EXAMINERS’ THESIS APPROVAL SHEET
SCHOOL OF GRADUATE STUDIES
ARBA MINCH UNIVERSITY
We, the undersigned, members of the Board of Examiners of the final open defense by TenkirTenka have read and evaluated his thesis entitled “Determinants of Farmers’ Participation in coffee Production and Marketing (The case of OydaWoreda in GammoGofa Zone Southern Nations Nationalities and Peoples Regional State” and examined the candidate’s oral presentation. This is, therefore, to certify that the thesis has been accepted in partial fulfillment of the requirements for the degree of Master of Science in Economics.
______
Name of the Chairperson Signature Date
______. ______
Name of Principal Advisor Signature Date
______. ______
Name of Co-Advisor Signature Date
______
Name of External examiner Signature Date
______
Name of internal examiner Signature Date
______
SGS Approval Signature Date
iii
ACKNOWLEDGEMENTS
Above all, I would like to forward my deepest gratitude to almighty God and his mother, St., Virgin Merry, for helping me to accomplish my will. No word of thanks and gratitude is sufficient to appreciate them have done for me. By their decree, this paper came out as a result of the contribution and support of many individuals whom I am greatly indebted to.
I really want to express my greatest thanks to my advisor Dr. Tora Abebe for his patience and constructive advice throughout the development of this thesis without which this paper would have been lost. Besides, my special gratitude goes to my Co-advisor Mr. Sileshi Abebe for his remarkable advices and encouragements throughout the course of the study.
I would like to warmly acknowledge my sponsor, Oydaworeda Chief Administration Office, for its full sponsorship. I want to extend my deepest gratitude to Oydaworeda Finance and Development Office, Agricultural Development Office, and Educational office for material support.
Lastly, I would like to thank my family and friends: To my wife Hawa Oumer for her support and love throughout the year; my daughters and son thank you so much for always being willing to help me out and for your love.
iv
Acronyms
AD After Death
CC Contingency Coefficient
DHM Double Hurdle Model
ECX Ethiopian Commodities Exchange
FAO Food and Agriculture Organization
FOB Free On Board
GDP Gross Domestic Product
GGZADR Gamo Goffa Zone Agricultural Department Report
GTP Growth and Transformation Plan
MARD Ministry of Agriculture and Rural Development
MOFED Ministry Of Finance and Economic Development
MOT Ministry Of Trade
NGO Non Governmental Organization
OWAO OydaWoreda Agricultural Office
PAERT Policy Analysis and Economic Research Team PASDEP Plan for Accelerated and Sustained Development to End Poverty PLCTC Primary Level Coffee Transaction Centers
SNNPR South Nation Nationalities and People Regional state
SSA Sub-Saharan Africa
USAID United States Agency for International Development
VAT Value Add Tax
VIF Variance Inflator Factor
v
Table of Contents
ACKNOWLEDGEMENTS ...... iv
Acronyms ...... v
Table of Contents ...... vi
LIST OF TABLES ...... viii
LIST OF FIGURES ...... ix
ABSTRACT ...... x
CHAPTER ONE ...... 1
INTRODUCTION ...... 1
1.1 Background of the study ...... 1
1.2 Statement of the Problem ...... 4
1.3. Objective of the study ...... 6
1.4. Hypothesis of the Study: ...... 6
1.5. Significance of the Study...... 7
1.6. Scope and Limitation of the Study ...... 7
1.7. Organization of the Study ...... 7
CHAPTER TWO ...... 8
LITERATURE REVIEW ...... 8
2.1. Theoretical Literature ...... 8
2.1.1 The definition of Cash crops and its Cropping by Smallholder Farmers in Developing Countries ...... 8
2.1.2 Coffee Production and Marketing in Ethiopia ...... 10
2.1.3. Coffee marketing in Ethiopia ...... 13
2.2 Empirical Literature...... 16
2.2.1. Determinants of Farmers’ Participation in Production and Marketing of coffee...... 16
CHAPTER THREE ...... 21
METHODOLOGY ...... 21
3.1. Description of the Study Area ...... 21
vi
3.2. Data Type and Source ...... 21
3.3 Population of the Study ...... 23
3.4 Sampling Techniques ...... 23
3.5 Sample Size ...... 23
3.6 Data Collection Techniques and Instruments ...... 24
3.7 Model Specification...... 25
3.7.1. Method of Data Analysis and Respective Empirical Models ...... 25
3.8. Statistical and Specification Tests ...... 29
CHAPTER FOUR ...... 38
RESULTS AND DISCUSSIONS ...... 38
4.1. Descriptive Results ...... 38
4.1.1. Socio-demographic characteristics of households ...... 38
4.1.2: Land ownership status of farmers...... 39
4.1.3. Livestock ownership of households ...... 40
4.1.4. Coffee production and associated problems ...... 41
4.1.5. Income sources of households ...... 41
4.1.6. Coffee Marketing Practices in Oyda ...... 42
4.1.7. Institutional Issues on Coffee Production in Oyda ...... 43
4.2. Econometric Results ...... 44
4.2.1. Production Participation (Probit regression) ...... 44
4.2.2. Factors Determining the Extent of Coffee Production Participation in Oyda ...... 48
4.2.3. Factors Affecting Coffee Marketing in Oyda ...... 52
CHAPTER FIVE ...... 56
CONCLUSIONS AND POLICY IMPLICATIONS ...... 56
5.1. Conclusions ...... 56
5.2. Policy Implications ...... 57
6. References ...... 59
APPENDICES ...... 65
vii
LIST OF TABLES
TABLES PAGES
Table 3.1: Sample size of each Kebele…………………………………………………..24
Table 3.2: The description and expected sign of farmer`s participation in production and marketing of coffee cash crop is summarized in the following table ……………36
Table4.1.1: Demographic characteristics of sampled farmers ………………………..…38 Table 4.1.2: Educational level of sampled households ………………………………….39 Table 4.1.3: Land ownership of the respondents …………………………………….….39. Table 4.1.4: Oxen and donkey ownership of sampled farmers…………………………..40 Table 4.1.5: Perception on major problems associated with coffee…………………...…41
Table 4.1.6: Summary of income from coffee sells………………………………..…….42
Table 4.2.1: Determinants of coffee production participation (Probit regression)………47
Table 4.2.2: Determinants of the extent of Coffee production participation……….……51
Table 4.2.3: Factors affecting income earned from coffee sale in the study area……...... 54
viii
LIST OF FIGURES
FIGURES PAGE
3.1 Location Map of the Study Area 22
4.1 Selling channel 42
4.2 Time of selling coffee 43
4.3 Farmers member ship states of cooperatives. 44
ix
ABSTRACT
Agriculture in Ethiopia remains the key sector that provides lion share of foreign exchange earnings and the largest labor force employer. Out of total agricultural output about 95% was covered by smallholder agriculture sub-sector. However, a number of factors limit farmers from participating in coffee production and marketing. The main objective of this paper was to identify household specific factors determining coffee production and marketing in OydaWoreda. A cross-sectional quantitative study was conducted in a sampled population by taking 214 sample sizes using systematic sampling method. The tools used in the study were structured interview, focus group discussion and observation. To examine the determinants of farmers’ decision to participate in the production activity and level of participation, Double hurdle model were used. In the first stage of double hurdle model, probit regression was used to examine farmers’ decision to participate in production. In the second stage of double hurdle model truncated regression were used to analyze level of participation and income generation from sell of coffee. The study indicated that farm size, family labor, number of oxen owned, access to credit, availability of family food, and distance to extension service significantly explain the decision to produce coffee. On other hand, the number of oxen owned, farmers experience on coffee production, number of working family members, and access to credit service determine the level of coffee production participation considerably. Furthermore, the study verified that in addition to the quantity of coffee marketed, market price, selling channels, selling time, travelling time from the nearest market and market price significantly determines the level of income earned from coffee sale. The implication is that livelihood improvement could be assisted through better participation of farmers in coffee production and marketing in the area.
Key Words: double hurdle model, smallholder farming, coffee, production, marketing
x
CHAPTER ONE
INTRODUCTION
1.1 Background of the study
Farmers in developing countries are subsistence oriented, focusing on growing enough food to feed themselves and their families. In the world, around 80% foods in developing countries are produced under smallholding farms (FAO, 2011). However, in recent years smallholder farmers are also taking part in market to sell some portion of their products. Smallholder farming has an important role in transforming agriculture from subsistence to market oriented produces. This commercialization of agriculture is important in the economic development of developing countries (Md. AtaulGaniOsmani, Md. Khairul Islam, Bikash, Ch. & Md. Elias, H., 2014).
Through commercialization, farmers can earn better profit to increase their family income and promote standard of living. Because of commercialization agriculture is not only just making a shift from subsistence to market oriented farming but also making better welfare outcomes for farmers in the form of increasing consumption of basic and high valued food. Moreover, higher expenditure on education, healthcare, non-food consumption and durable goods for the farmers can be achieved by commercializing agriculture (Gebreselassie and Sharp, 2008).
Smallholder farmers’ agriculture continues to play a key role in African agriculture. East African countries like Kenya, Ethiopia, Uganda and Tanzania have large number of people and land size. Especially in Ethiopia, 11.7 million smallholder households account for approximately 95 per cent of agricultural GDP and 85 per cent of employment (FAO, 2011).
According to Ministry of Agriculture and Rural Development (2010), nearly 55 percent of all farmers at country level, 53%, of farmers at SNNPR, 51% of farmers’ at Gamo Goff zone and 50.45% of farmers at OydaWereda operate on one hectare or less. That means above half of the total farmers in Ethiopia are smallholderfarmers. This confirms the dominant contribution of marginal and small farmers to the overall agricultural growth in the country (MOARD, 2010). In short, as the overall economy of Ethiopia
1 depends on agriculture sector development, the entire movement of the agriculture sector depends on what is happening in smallholder sub-sector.
The government of Ethiopia through its different policy documents such as Growth and Transformation Plan (GTP), positions farmers as a principal source of agricultural growth; and agriculture as the driving source of overall economic growth. For example, commercialization of smallholder farming received high government policy priority through GTP (MoFED, 2010). In this regard, the major effort was placed to support the intensification of marketable farm products both for domestic and export markets by the small and large scale farmers. Such fundamental strategy involves an enhancement of producing high value crops paying a special focus on high potential areas to do so.
Empirical record suggests that cash crops can provide higher returns to land and labor than food grains and thus present major opportunities to promote smallholders income growth, food security, and national foreign exchange generation(Jayne, 1994; Poulton et al., 2001, Lukanu et al., 2004; Poulton et al., 2006, Schneider and K.Gugerty, 2010). According to Chauvin (2012), cash crops are a major source of export revenue for a large number of sub-Saharan African countries and the livelihood basis for millions of rural households who grow those crops.
Coffee is the second most traded commodity after petroleum and determines the livelihoods of 25 million poor families in the world. Its status as a major export for many countries and therefore a determinant of the wellbeing of national economies, gives it significant importance in the global economy. However, coffee also disproportionately affects small-scale farmers as coffee is one of the few internationally traded commodities that is still produced mainly on smallholdings farmed by peasant households, with almost 70 per cent of production coming from producers who farm less than ten acres (4 hectare) of land (Kendra, 2009).
Coffee, which grows on a bush, is grown around the world between the latitudes of 230North and230South.There is two main types of coffee cultivated Arabica and Robusta, with Arabica considered being the better quality of the two. However, it is also more difficult to grow. Arabica coffee prefers to be grown in shade and indeed the resulting coffee bean tastes better when allowed to ripen slowly. Arabica coffee bushes are
2 relatively weak and need to be picked by hand. Robusta, in contrast, can be grown in full sun and can be picked by mechanically. Therefore, Robusta coffee can be grown on large, plantations while Arabica is generally grown by small-scale producers who can more easily tend to the plants and provide a richer, shadier habitat (Kendra, 2009).
Ethiopia is the largest producer of coffee in Sub-Saharan Africa, about 15 million people directly or indirectly deriving their livelihoods from coffee and is the fifth largest coffee producer in the world next to Brazil, Vietnam, Colombia, and Indonesia, contributing about 7 to 10% of total world coffee production. Ethiopia is origin of coffee and produces mostly Arabica coffee. It has economical, environmental as well as social significance to the country. At the momentCoffeeis growing through at the country, but, largely in two regions of the country namely: Oromia and Southern Nations, Nationalities and People Regions(SNNPR). In GammoGoffa zone the known coffee producing woredas are Melokoza, DembaGoffa, GezeGoffa,Oyda,Kamba,Bonke,Boreda,Arba Minch Zuriya and kucha(GGZADR,2014/15). In the Oydaworeda twelve kebeles are producing coffee, those kebeles are Shefit 01,Uba dama,Ubayambala,Ubaganchila,Kamo,Lame, Garda, Markala, Shefite 02,Kalamalo, Gemtgocho and Shallabarind (OWAO,2015).95%of Ethiopia’s coffee is produced by smallholder farmers while the remaining five percent is grown on modern commercial farms (Abu, 2012).
Marketing in oydaworeda is conducted local coffee markets or primary transactions center collecting coffee from scattered small scale farmers. Most of the coffee farmers are not bringing their produces to the coffee markets, in steady; they are supplying their coffee to the collectors. Two main reasons could be mentioned for this. First, majority of the coffee farmers are small-scale farmers that their produce is little and the cost of round trip transportation and other contingent costs they incur made it economically undesirable. Secondly, there are many farmers who are physically weak that they can’t transport their produce to these markets who have only one option which is to deliver to the collectors (OWAO, 2015/16).
Therefore, given the agriculture based economy of Ethiopia and the dominance of smallholder sub-sector, it is imperative to conduct a study which focuses on identifying factors determining farmers’ participation in production and marketing of coffee crop.
3
Thus, analyzing determinants of farmer’s participation in production, level of production and marketing of coffee in Oydaworeda is the main concern of the current study.
1.2 Statement of the Problem
The coffee sub-sector is important to the Ethiopian economy; in 2005coffee export generated 41% of foreign exchange earnings and provides income for approximately 8 million smallholder households in Ethiopia. Policy attention to the sector was always imminent, and its importance has been renewed in the latest Poverty Reduction Strategy, the Plan for Accelerated and Sustained Development to End Poverty (PASDEP) (MOFED, 2006). Coffee particularly is the backbone of the Ethiopia economy. Coffee has always been Ethiopia’s most important cash crop and largest export commodity, which account 90 percent of exports and 80 percent of total employment. By its very nature, coffee is highly labor-intensive production activity. Thus very significant part of the population derives its livelihood from coffee. Coffee thus has a significant impact on the socio-economy life of the Ethiopian farmers and economic development of the country (PAERT, 2008).Coffee production also has a multiplier effect that could lead to increased demand for consumption of basic and high value food, non food consumption, durable goods and services in the local economy leading to higher levels of monetization and its better integration into the wider economy (Samuel & Eva, 2008).
Despite its importance as cash crop and export major export item in Ethiopia the production and controlling systems affects the amount of productivity that decreases income of farmers (Kifle, 2015).According to Alemseged (2013),factors reducing coffee production in Ethiopia were weak farm management systems, the agronomic practice are traditional, extension services provided to smallholder farmers are inadequate, lack of the necessary technical skills and knowledge in using agricultural technologies, poor extension and credit services, low rate of technological adoption and poor infrastructure.
Coffee is cultivated by over 4 million primarily smallholder farming households (CSA, 2013) and grows in Ethiopia under diverse environmental conditions ranging from 550 meters to 2600 meters above sea level, with annual rainfall from 1000-2000 mm, temperature (minimum and maximum from 8-150C, and 24-310C, respectively), requires deep, well drained, loamy and slightly acidic soils (Paulos and Tesfaye, 2000). The estimated area of land covered by coffee is about 600,000 hectares, whereas the estimated annual national production of clean coffee is about 350,000 tons (Alemayehuet al., 2008). 4
Oydaworeda is one of coffee growing Woredas in the GammoGofa zone south Regional State, which has a total area of 672 hectares of coffee land (OWAO, 2015). Currently, the total area of land covered by coffee in the Oydaworeda is about 217 hectares. The numbers of household farmers in coffee producing Kebeles were 1735. But the total numbers of participant house hold farmers are only 937 in numbers. According to OydaWoreda Agricultural office (OWAO) annual report (2015/2016) the capability of coffee production was 12thousand tons per year. However, the achieved productivity of coffee was 7 thousand tons per year which shows us the level of productivity was far below from available potential.
Different researches indicate that there is huge potential to grow coffee in the country and there is high market demand at local and international levels (McMillan, Assefa, Kibre and Amdissa, 2003). However, in addition to the limited availability of agro-ecologically suitable areas for coffee production and productivity in the country, farmers’ production and marketing participation is not as such satisfactory (MOFED, 2010). That is, even in the agro-ecologically suitable areas for coffee production, farmers’ participation is far below the potential. For example, in Oydaworeda, there are suitable agronomic conditions and large land size for growing coffee. Despite the available potentials and opportunities, majority of farmers are not participating in coffee production and marketing in this area.
This indicates that there are external and internal (household specific) factors that constrain some households from participation. In addition, the extent to which the participant farmers participate varies significantly and the overall participation is unmatched with the available potential. Similarly, producer farmers’ face a number of marketing problems, which influences the income these farmers could derive from coffee sale. Due to these factors, smallholder farmers in Oydaworeda are differently responding to the available potential and thus obtain different welfare benefits from the available opportunities.
The existing studies conducted by (Alemayehu,2010, PAERT,2008 andAnwar,2010)more concerned on performance of the coffee export sector in Ethiopia, technical efficiency of coffee producers using stochastic frontier analysis, constraints and dissemination of improved coffee varieties, coffee production, utilization and marketing in Ethiopia. All
5 the studies used Logit and Tobit models .These models are not perfectly convenient to assess the participation decision and level of participation, because Logit model analyze only decision to participate in the activities or not and Tobit model estimate the participation decision and level of participation in coffee production determined by the same variables and the same sign. The researcher additional used variables such as experience of farmers in coffee production, food sufficiency for the whole year and coffee selling time to fill the variable gaps. The present researcher used double hurdle model with probit and truncated regression to better address the research problem.
Studies conducted on coffee production and marketing in Ethiopia (Kendra, 2009, Samual&Eva, 2008, and Abu, 2012) have considered the common coffee production related problems, ignoring factors affecting production participation decisions at individual household levels. This study examines factors affecting production and marketing participation decisions at individual household levels.
1.3. Objective of the study
General Objective
The general objective of this study is to analyze factors that influence farmers’ participation decisions in production and marketing of coffee.
Specific objectives
To examine factors affecting farmers’ participation decision in coffee Production. To identify factors affecting level of coffee production. To analyze factors affecting marketing of coffee in the study area
1.4. Hypothesis of the Study:
Farmers’ participation in coffee production has positive correlation with farmers own farm size. Level of coffee production participation is positively correlated with credit access, number of active family labor, number of oxen, and other related variables. Income generation from coffee sale has positive correlation with quantity of coffee marketed
6
1.5. Significance of the Study
In Ethiopia majority of farmers participate in production of staple crops to sustain food security and then to market the surplus if any. In addition to providing staple crops for domestic and international markets, farmers also produce large shares of traditional export cash crops. Coffee is one of the Ethiopian export crops and is the major cash crop cultivated by farmer’s in the study area, even though it is cultivated below available potential and opportunities.
However, this study will have a significant contribution to the farmers in the study area and can be used as an input for policy making and researchers in the area. It also shed some light on the problems of farmers less participation in production and market of cash crops especially in coffee.
1.6. Scope and Limitation of the Study
In any research, there would always be certain limitations. The primary limitation of this study is its limited scope of being in a single Wereda. That is said because, this study is designed to identify demographic, socioeconomic, physical and institutional factors explaining the participation status of smallholder farmers in production and marketing of coffee in OydaWereda alone. However, the issue of coffee production and marketing by farmers would have been better understood in the country if the process dimension is studied through time, and at least cover an additional potential Woredas under the investigation.
1.7. Organization of the Study
The rest of this research work organized as follows; second chapter presents review of related literatures. Chapter three deals with the methodology part that introduces data type and source, sample size determination, sampling techniques, data collection instrument, method of data analysis and econometrics model specifications. Chapter four contains the descriptive and the econometric analysis of factors affecting farmers’ participation decision; factors determine the extent of coffee production and factors affecting coffee marketing. Finally, chapter five deals about conclusion and policy implications based on the empirical results.
7
CHAPTER TWO
LITERATURE REVIEW
This chapter reviews some relevant literature regarding smallholder cash cropping. The first section reviews the theoretical issue of smallholder farmer’s cash cropping in developing countries context, coffee production, marketing and consumption in Ethiopia. Then in the next section, we present some relevant empirical literatures on factors affecting smallholder farmers’ cash crop production participations and marketing.
2.1. Theoretical Literature
2.1.1 The definition of Cash crops and its Cropping by Smallholder Farmers in Developing Countries
One of the common forms in which farmers` commercialization occurs in developing countries is through production of cash crops in addition to staple crops. In almost all these countries, when any one talks about agriculture, the issue of farmers’ commercialization comes first. A cash crop is a crop that is primarily produced for market and largely sold, thus generating income for the farming households (Lukanu et al., 2004). In theory, it is generally believable that the basic motivation of cash crop is higher returns to used resources for its production. In this regard, many recorded literatures reflect the importance of cash cropping in developing countries as it can be defined in terms of land use, employment, output, income or export at household, village, regional or national levels (Von Braun & Kennedy, 1994,&Poulton, 2001).
However, the issue of cash crop production is strongly hot in many ways, especially in Africa. According to de Janvry et al., 1991& Jayne, 1994, for one thing, cash cropping are favored from their potential contribution to growth, employment and external balances. These authors further explained that, the expansion of cash cropping is recommended to use comparative advantage and provide the basis for industrial development through internal linkages. According to the authors, on the other hand, cash crops are opposed by those who disagree with these benefits and point out to additional drawbacks, especially in the spheres of food security. This part of literature argues that, this contrary view is particularly associated with the sustained evaluate of the food-first tendency. These bodies of critics describe cash crops as the enemy of food security. Of
8 course, the main argue in this case was comes from the fact that cash and food crop productions competes for farm household resources (especially in developing countries where these resources are scarce and limited). And this competition is severe particularly under missing or imperfect food markets in which households prefer to produce their own food crops to secure household consumption at the expense of higher returns from cash crop production (de Janvry et al., 1991; Jayne, 1994).
Despite these arguments, many household level studies show the complementary nature of food and cash crop productions at household levels (Von Braun & Kennedy, 1994, Poulton et al., 2001; Schneider &K.Gugerty, 2010). Their argument bases itself on the income and financial linkages between the two types of crops. The researchers argued that income from cash crops might be used either to purchase food crops from a market, which permits allocating most household resources to cash crop production, or to purchase external inputs for the production of food crops that enhance food crop productivity. Cash cropping necessarily never associated with declining of food production at either the household or national levels. Similarly, Poulton et al., (2001); argue that although food and cash crop productions often seen as mutually exclusive alternatives, increased cash crop production need not reduce food production at household levels. They reason out this that, income from cash cropping may enable households to invest in lumpy assets such as animal traction and helps to use more modern production inputs such as fertilizers and others that increases productivity of the food production. The study presented by Von Braun and Kennedy (1994) also suggests that households participation in cash cropping need not reduce own food production or nutritional status.
Additionally, many different studies indicate that, in sub-Saharan Africa, cash cropping remains the most important income sources for farmers and governments (through exports). In this regard, Chauvin (2012) suggested that cash crops are the major source of export revenue for a large number of Sub-Saharan African countries and the livelihood basis for millions of rural households who grow those crops (Chauvin, 2012). The author recommended that poor farmers in the cash crop sector should stand a better chance to rise out of poverty on the back of export market prices which normally bring better returns.
9
In line with this, Poulton et al. (2001) have listed some trends which will encourage the move toward cash cropping across a wide range of developing countries. For example, the increasing high demands for cash (e.g. for schooling, health, high cost of production inputs, etc) encourage participation of smallholder farmers in cash cropping for those whom crop sales are the major source of income. In addition, these authors argues that, the exchange rate policy (e.g. real devaluation) of a county make production of internationally tradable crops relatively more profitable than production of crops sold only on local markets, hence enhances smallholder cash crop production participations in those countries. Furthermore, Poulton et al (2001) suggested that long-term changes in the relative prices (on international markets) encourages those households who grow these crops for cash and may result in greater market-orientation of rural households. This indicates that, cash cropping contributes to growth through production linkage effects; in which it permit diversification away from the subsistence farming to somewhat market- oriented behaviors( de Janvry et al. 1991; Jayne, 1994).
2.1.2 Coffee Production and Marketing in Ethiopia
2.1.2.1 Coffee production in Ethiopia
Ethiopia is the birth place of coffee and it discovered earlier in the world. More than 1,000 years ago, coffee was produced in Ethiopian southwestern highlands. David Beatty discovered the Ethiopian area where they first blossom Kaffa gave its name to coffee. Nobody is sure, exactly how coffee was originally discovered as a beverage plant; it believed that its cultivation and use began as early as the 9th century in Ethiopia. It cultivated Yemen earlier, around AD 575. While, it originated in Ethiopia, from where it traveled to the Yemen about 600 years ago, and from Arabia began its journey around the world. Among the many legends, Kaldi, an Abyssinian goatherd, who lived around AD 850 found the origin of coffee. It is vital to the cultural and socio-economic life of Ethiopians and contributes 25%-30% of the country's foreign exchange, 50% of GDP, 85% of total employments in the country and part of the culture; about 50 % of the produced coffee is consumed domestically(Alemayehu, 2014)
Coffee grows well under the large indigenous trees such as the CordiaAbyssinicaand the Acacia species, in two regions of the country Oromiya and southern nation nationality and people regional state. In our country smallholder farmers on less than two hectares of
10 land produces and supply Ninety-five percent of Ethiopia’s coffee produces, while the remaining five percent grown on modern commercial farms (Taye, 2013 & USAID,2010).
According to (USAID, 2010) Coffee production systems in Ethiopia generally categorized into four areas i.e. forest coffee, semi - forest coffee, garden coffee, and plantation coffee. Forest coffee is a wild coffee grown under the shade of natural forest trees and it does not have a defined owner. Semi-forest coffee farming is a system where farmers select forest trees to let sufficient sunlight to the coffee trees and to provide adequate shade. A farmer who prunes and weeds the forest area once a year claims to be the owner of the semi forest coffee. Garden coffee normally found in the vicinity (near) of a farmer’s residence. It normally fertilized with organic material and usually inter- cropped with other crops. The government or private investors for export purposes plant Plantation coffee. Fertilizers and herbicides usually used in the coffee plantation farming system.
As (Sentayhu, 2013) Forest coffee accounts 10%, Semi forest coffee accounts 30%, Garden coffee accounts 50 % and Plantation coffees accounts 10% and according to (Taye, 2013) the forest coffee production accounts 8-10%, semi-forest coffee accounts 30-35%, garden coffee accounts 50-55% and Plantation coffee accounts 5-8% of its total production respectively. Ethiopia Small-scale holdings equal to or greater than 95% of total coffee production. According to (Alemseged&Getaneh, 2013) Ethiopia is the world’s fifth largest coffee producer and Africa’s top producer, with estimated coffee production of more than 450,000 tons and marketable supply of 334,000 metric tons in farm year 2012/13. Half of the coffee produced consumed locally and the country leads the African Continent in domestic consumption. It has been used income generation for that about 20 percent of the populations, directly or indirectly, depend for a living on coffee production and trading.
As (Anwar, 2010) coffee is the most important crop in the national economy of Ethiopia and the leading export commodity. Ethiopia is well known not only for being the home of Arabica coffee, but also for it is very fine quality coffee acclaimed for its smell and flavor characteristics.
11
Ethiopia encompasses a potential opportunity to increase coffee production. It is endowed with suitable elevation, temperature, and soil fertility, indigenous quality planting materials, and sufficient rainfall in coffee growing belts of the country. Coffee is a shade- loving tree. Forest coffee yield is low as considered to garden and semi-forest coffee because resource ownrity belongs to communal and poor management.
2.1.2.2 Coffee Consumption in Ethiopia
Ethiopians are heavy coffee drinkers, ranked as one of the largest coffee consumers in Sub Saharan Africa. Nearly half of Ethiopia’s coffee produce have locally consumed. Coffee in Ethiopia has both social and cultural value. It mainly consumed during social events such as family gatherings, spiritual celebrations, and at times of sadness. Coffee supplied and traded in the local market usually has a lower quality. Coffee on the local market is mainly coffee destined for export through the Ethiopian Commodities Exchange (ECX) market but rejected for failing to meet ECX’s quality standards (Abu & Teddy, 2013)
.
An interesting new development in Ethiopian major cities regarding coffee consumption is the emergence of small roadside stalls selling coffee to passer by customers. The small roadside stalls serve coffee in a traditional manner. They have emerged and flourished in Ethiopia’s major towns, growing very popular among coffee consumers who are frustrated by the escalating price of coffee and the deteriorating quality of coffee served in cafes and coffee shops. The exorbitant local coffee prices have also pushed some consumers, particularly those residing in non coffee growing areas, to boil and drink the skin of a coffee grain as a substitute for normal coffee (Abu & Teddy, 2013).
2.1.2.3 Coffee Fair Trade
According to ToraBäckman (2009), fair trade is a trading initiative based on equity that claims to contribute to development by increasing farmers’ profits and empowerment in communities. Ethiopia has grown coffee for a thousand years, is heavily dependent on export of coffee beans, and has recently started to export Fair trade certified coffee.
As ToraBäckman conventional coffee is collected from the individual farmers, processed and shipped to the auction in Addis Ababa or Dire Dawa, and exported through the port
12 in Djibouti. Fair trade coffee has received permission from the Ethiopian Coffee and Tea Authority to bypass the auction and be directly exported through Djibouti, with the benefit of avoiding middlemen to get a higher FOB price.
According to Ethiopian coffee annual report 2013, coffee is still Ethiopia’s number one export item. It accounts for 45 to 50% of Ethiopia’s total export earnings but its share of total export earnings has gradually declined in recent years because of increased exports of other commodities such as gold, flowers, chat, textiles, and leather products.
2.1.2.4 Coffee Price Volatility
Global coffee production varies from year to year according to weather conditions, disease and other factors, resulting in a coffee market that is inherently unstable and characterized by wide fluctuations in price. This price volatility has significant consequences for those who depend on coffee for their livelihood, making it difficult for growers to predict their income for the coming season and budget for their household and farming needs. When prices are low, farmers have neither the incentive nor resources to invest in good maintenance of their farms by applying fertilizers and pesticides or replacing old trees. When prices fall below the costs of production, farmers struggle to put adequate food on the table and pay medical bills and school fees a major reason for children taken out of school to contribute to the family income by working on the farm or in the informal sector. Therefore the volatility of coffee markets in combination with poor production infrastructure and services have sunk the majority of coffee producers in developing countries in low-input-low-output cycles and structural poverty (ToraBäckman, 2009).
2.1.3. Coffee marketing in Ethiopia
2.1.3.1 Primary Level Coffee Transaction Centers (PLCTC).
Place where coffee farmers and suppliers transact coffee. They are located near to the coffee farms. Currently there are about 979 primary coffee marketing centers in the country (Taye, 2013).
2.1.3.2 Ethiopian Commodity Exchange (ECX).
The secondary level where coffee transact in Ethiopia. Currently ECX warehouses are located at 8 different parts of the country. The centers are in DireDawa, Hawassa, Dilla,
13
Sodo, Bonga, Djimmah, Bedele and Gimbi. The coffee transact in Addis Ababa in open outcry or protest.
The ECX is entrusted with broad objective of modernizing the Ethiopian agricultural market and thereby attaining overall economic growth. Specifically, for which the ECX is established for the following main purposes. Firstly, ECX is established to provide a centralized marketing mechanism in which transactions are carried out publicly through a physical trading floor or electronic system or both. Secondly, it is devised for creating an efficient, transparent, and orderly marketing system which addresses the interest of all stakeholders including buyers, sellers and intermediaries and small scale producers. Thirdly, it is there to gather and monitor and disseminate timely information concerning the market and exchange transactions to the general public. Fourthly, ECX is established to conduct trading based on product grade certificates, warehouse receipts, and standardized and grade specific contracts. Fifthly, ECX is come to reality with a view to do clearing and settling of transactions the Exchange itself to minimize default risks. Finally, it is there to provide a dispute settlement forum; undertake market surveillance activities to maintain the integrity of the market and of the members, and avoiding contingent risks by employing modern risk management tools (ECEP, 2007).
2.1.3.3 International coffee market:
The third level where Ethiopian coffee transacts takes place. In this level, the Exporters sell coffee to importers. In Ethiopia, only the citizens export green coffee. According to (Getue, 2011) Coffee improvement opportunities related to market growth of specialty coffee industry and wide range of market options, diverse coffee consumers preference, modern marketing system, trade marking and licensing initiative, natural resource richest Arabica coffee gene pool, diverse agro ecology with unique quality profile, associations Active role cooperatives coop, Proclamation updates on coffee quality and marketing systems. In addition good investment policy in the country specialization, Example Jimma University Control Institutes coffee research center, Promising capacity-building efforts (graduates studies on coffee tea and spice Nongovernmental organizations newly emerging development intervention on coffee by NGOs.
14
2.1.3.4. Coffee Production and Marketing Value chain in Ethiopia.
The producers under this stage in the coffee value chain of Ethiopia include small-scale farmers, private owned farmers and state firms. The major portion inters of volume of products mobilized, value adding functions, market share and capital owned in coffee value chain of the country is under the hands of producers especially the large-scale private coffee plantations and state farms of coffee plantations. After the coffee is grown and matured, the following value adding activities in the value chain performed by those producers are collecting coffee chary and transporting to processing areas (USAID, 2010)
Coffee cherry collecting and transporting activities in Ethiopia in which except loading and unloading, mostly performed by women groups of farmers. Most of the farm products including coffee are raw in nature and need to process before consumption. This increases the cost of marketing service, which adds value and price on farm products. Under this main activity the sub tasks performed in processing the coffee are pulping, washing coffee, drying, sorting, sacking/ packing, loading, and transporting then finally unloading to the warehouse. The small scale coffee producers are always sell the red cherry coffee on their farm as it is without harvesting, drying, and hulling to the coffee collectors. However, some small-scale farmers in country grow, harvest, dry, hull and sell their dry cherry coffee to collectors (legal and illegal collectors). While, household farmers were mostly sell red cherry coffee. The large scale-private farmers and state farms harvest coffee chary and use pulping machine (dry or wet pulping machines) add more value on the coffee products. The pulped and washed coffee then exposed to sun rise in appropriate place until the coffee bean become properly dried and those foreign materials in coffee are sorted so that it will be ready grading and sacking. Therefore, most agro processing employees are women (USAID, 2010). Packing dry coffee loading, transporting, and unloading to the warehouses transported to the final market through ECX is also value addition. While some of the large scale, private coffee producers sell their products the exporters either in Addis Ababa or to international importers.
In summary, the main lesson that we learnt from these recorded literature is that, at least in theory, production of coffee may enable farm households to obtain more income that they could obtain by devoting the same household resources to staple crops. In addition these theoretical literatures suggest that coffee is the main source of export revenue for
15 many developing countries. This is also true in Ethiopian case, since Ethiopian export is primarily agricultural commodities. And many reports and facts indicate that, these crops are basically produced by smallholder agriculture sub-sector. Thus, it is important to analyze the status of farmers in production and marketing participation of coffee crop, based on the available theory. Here the main effort is not to analyze the issue by considering these farmers as they are specialized in coffee cropping, rather we focus on analyzing the issue by considering as these farmers can produce the two crops simultaneously by well management of household resources. Of course, production of some cash crops may totally depend on agro-ecological conditions. This also requires special focus and we accounted for the issue in this study. With these theoretical establishments, the researcher turns to focus on factors affecting farmers to participate in production of this crop.
2.2 Empirical Literature
2.2.1. Determinants of Farmers’ Participation in Production and Marketing of coffee.
Cadot(2006) demonstrated that private asset accumulation is a prerequisite for smallholders’ graduation from subsistence production. The author suggests that one possibility for farmers to accumulate private assets is to enter into cash cropping. And investment in public infrastructure such as roads, and information communication facilities are the major determinants of participating in coffee productions.
Jayne (1994) argues that high costs related to purchasing food on the market make cash crop such as coffee production unattractive, despite higher returns of cash crops on the farm. The author suggests so that, it is economically unviable to replace food crop production with cash crop production in this cases. Thus, according to the author food security condition is the one possible factor in limiting smallholder farmers to produce any cash crops, coffee. Similarly, Boughton et al (2007) argued that the main challenge and constraint factor for farmers’ to participate in coffee production is the low productivity in food crop production and its market failure. According to these authors, as farmers have access to secure their food demand they are most likely to participate in production of market-oriented crops.
16
Cotton is one of the known cash crops. The production of it in Zimbabwe observed that the most critical determinants of smallholder decision to produce cotton in Zimbabwe include farmer education levels, distance from the nearest buyer, and the early clearance of the tsetse fly. Their result also revealed that traction equipment and draft power were among the key determinants of households’ ability to diversify into cotton production in the country (Govereh& Jayne, 2003).
It is generally expected that farmer’s decision to cultivate a given coffee can be influenced by factors including household characteristics; economic factors (including the crop profitability and market availability); institutional factors (e.g. availability of extension, inputs and credit services); and environmental factors that involve the crop’s compatibility to existing climate, soil, disease and pest conditions (Lukanu et al., 2004).
Aysheshm (2007) assessed a sesame value chain analysis in MetemaWereda and verified that lack of improved variety seed that properly fits the woreda agro ecology and lack of agro-chemicals supply at the right time and at fair prices constrained sesame production in Metema. In addition, according to Aysheshm, water logging problems has a contributing factor for the reduction of output, yield and thus marketed supply of sesame in the area as well as other cash crops. Furthermore, his findings indicates that sesame marketing has been constrained by diverse factors such as shortage of modern inputs, shortage of capital, lack of timely and accurate market information, and poor quality of packing materials as a few of the inherent problems in the field.
Abdurahman (2005) study the determinants of the elasticity of coffee supply using both cross sectional and time series data from Hararghe high lands. He collected cross sectional data from 60 households residing in two peasant associations of the Hararghe zone. His study found a short run price elasticity coefficient of 0.6, which in line with the argument that individual crop price elasticity is larger because farmers can shift their variable in puts between different crops more easily. On the other hand, he also found that availability of consumer goods has a positive impact on the supply of coffee. In his estimation, he found that the sign of all the parameters to be consistent with his prior expectations except the coefficients for the coffee in the parallel market. The study suggested that increased relative producer price of coffee alone cannot be enough to
17 induce a significant positive response by coffee farmers because they face various non price constraints in coffee production. Thus the increase in price should be accompanied by various structural reforms to remove these constraints and to encourage coffee production and supply.
Teshome (2009) studied the determinants of coffee export supply by taking coffee arrival as dependent variable. The study uses time series data collected from different institutions mainly from national bank of Ethiopia. He employs vector autoregressive and vector error correction model .The study includes world price of coffee, producer price and rain fail, credit access, extension service, Gross Domestic product and real exchange rates as the explanatory variables of the model. The major findings of the study indicates that world price and producer price of coffee affects coffee production negatively their price elasticity was -1.62 and 0.69 respectively. The impact of rain fall is significant in both short run and long run .However, credit access and extension service are insignificant in the long run but significant in the short run .The study also indicates gross domestic product and real exchange rate does not have any impact on the export supply of coffee. Finally he recommends that providing of credit access and extension service at each woreda for coffee farmers are supposed to proved significant effect on export supply of coffee.
The study by Alemu and W.Meijerink (2010) suggests that the presence of high transaction costs, related to the lack of sufficient market coordination between buyers and sellers, the lack of market information, the lack of trust among market actors, the lack of contract enforcement, and the lack of grades and standards narrows market channels in Ethiopia at present. They argued that the persistence of such high transaction costs and contract risk have resulted in limited arbitrage and weak investments by private traders, leading to limited market volumes, weak responsiveness to price signals and high price volatility, all of which have a negative impact on smallholder producer livelihoods. In addition, in his survey study at Humera and east Wellega Zone, Sorsa (2009), argued that despite the potential for increasing the production and productivity of sesame seed in Ethiopia, a number of challenges inhibits its production and productivity in the country. Among the many production constraints the author have listed, the most important includes lack of improved cultivars, poor seed supply system and a lack of adequate
18 knowledge of farming and post-harvest crop management. In addition, the author found that the severe biotic stresses were also the major sesame production related problems in Humera and east Wellega areas. The same author concludes that smallholder farmers’ lack the necessary technical and material input to improve their sesame production and productivity in Ethiopia.
Marcia (2006) studied about four countries production of coffee the data shows that the average coffee yield in Vietnam (2733kgs per ha), Guatemala (970 kgs per ha), Honduras (627 kgs per ha) and Nicaragua (452 kgs per ha). These differences observed in coffee yield can be attributed to three factors (1) the types of coffee that is cultivated by Central American countries (primarily Arabica) and Vietnam (primarily Robusta); (2) organic vs. conventional production; and (3) differences in input use and tree age. Technical efficiency scores indicate that the mean technical efficiency score for all is 0.72, which implies that the production, on average, is about 28% below the frontier. This means that a considerable amount of output, on average, was missed due to technical inefficiency or that inputs were not at their optimal levels. The technical efficiency estimates varied from 8%-92%. Results from the inefficiency model reports that small farm size was a reason for inefficiency in coffee production. In addition, it was found that labor and organic fertilizer were factors for inefficiency, implying that, more used of this input the less technically efficient farmers are. However these variables were not significantly different from zero. All parameter estimates have the expected signs. Labor, tree age, pesticide, chemical and organic fertilizer all are positively correlated with yield.
According to Marcia (2006), the elasticity of yield with respect to labor is 0.33. This means a 1% increase in the level of labor is associated with a 0.33% increase in yield. In addition, the contribution of pesticide to yield is 0.07, indicating that a 1% increase in the amount of pesticide is correlated with a 0.07% increase in yield. Furthermore, the contribution of chemical fertilizer to yield is 0.09. This means that a 1% increase in the amount of chemical fertilizer is correlated with a 0.09% increase in yield. These input elasticity’s show that yield is sensitive to changes in Input levels for labor, pesticide, and chemical fertilizer. This suggests changes in input prices could affect yield by changing the incentives for input levels.
19
In general, the bodies of literature suggest that, increased productions of crops for markets are both an inevitable feature of rural development and essential in the countries where agricultural sector was believed to support the general economic development in these countries. This part of literature evidences the accompanying greater productivity and higher household incomes as a sign of such development benefit from coffee production by farmers. This evidence suggests that in many cases small-scale coffee cropping is both technically and economically efficient. Poulton et al. (2006) argue that, in general, traditional export coffee can make a significant contribution to poverty reduction when there is broad based participation by farmers in an area, labor-intensive production processes, and potential positive linkages to staple crop productivity in cash crop production especially on coffee. Some authors suggested that increased relative price of coffee alone cannot affect coffee farmers production rather world price of coffee ,rain fail, credit access ,extension service, exchange rate, age of coffee tree, family size and fertilizer application. Additionally most researchers focused on the extent of coffee production excluding the farmers those who are not decided to participate in production of coffee. Other researchers assessed the determinants of coffee price by depending on international market using secondary data, which cases biased inconsistence and didn’t show the local market price and related marketing system.
Here, in this research I was more focus on house hold level of farmers to show the determinants which are obstacle to the farmers to decide production participation of coffee, to increase level of coffee production participation and especially the marketing in the study area. The researcher was not depending much on secondary data rather depends on cross sectional primary data. In addition, the researcher used double hurdle econometrics model and including additional explanatory variables to address the determining factors of farmers’ decision to coffee production, extent of coffee production participation and marketing in Oydaworeda.
20
CHAPTER THREE
METHODOLOGY
This chapter describes specific procedures that the researcher anticipates adopting for his thesis. In other words, this section succinctly articulates specific procedures for addressing the research problem. This chapter introduces the study area, research design, and methodology of our research.
3.1. Description of the Study Area
This study was conducted in GamoGofa Zone, in Oydaworeda which is one of the fifteen woredas and two town Administration of GamoGofa zone, in SNNPR. Oyda is bordered on the North by GezeGoffaworeda, on the East by DembaGoffaworeda, on the South by UbaDebretsehayworeda and West Ari woreda (South Ommo zone).It is situated at 267 kilometers far apart from Arbaminch (the Zonal Capital), 297 Kilometers from Hawassa (regional capital) and 525 kilometers from Addis Ababa (national capital).
According to CSA total population of the district is 41,545; males account 19,967 (48.07%) while females covered the rest 51.93 percent (in absolute term 21,578). Oydaworeda has 19 rural and one urban kebeles. There are three agro-ecological zones; temperate, semi – tropical and tropical. Like other parts of the region, agriculture is the main means of livelihood for the population both in terms of crop production and livestock.
3.2. Data Type and Source
To obtain information on the socio economic condition of the households in the woreda data was collected through structured questionnaires. The households were interviewed by using structured questionnaires. Before the main survey, the enumerators were given a training mainly focusing on the technicalities of the questionnaire. A pilot survey was conducted to check its wording, ordering, and timing. The data were both quantitative and qualitative types. The factors that contribute to farmer’s demands to participate in production are analyzed by using qualitative methods. Quantitative research methods are used to measure demographic characteristics (sex and age distribution, family size), social situation, economic situation, education, and credit access.
21
Figure 3.1 Location Map of the Study Area
Source: Study area map generated from EthioGIS (2007)
In addition to statistical investigation conducting discussion with key informant and carrying out observations are crucial to understand the problem in depth. Therefore both quantitative research method and qualitative research methods are jointly used for this
22 research. The secondary data was used to supplement the primary data and obtained from the woreda administration, agricultural office and from selected kebeles.
3.3 Population of the Study
Oydaworeda has currently 20 Kebles. From these kebeles only 12 kebeles are coffee producers (OWAO, 2014). The populations of this research are the number of households in these twelve coffee producer kebeles.
3.4 Sampling Techniques
Because of time and financial constraints reaching all coffee producer kebeles of Oydaworeda is practically impossible. The twelve coffee producing kebeles are located at the same climatic condition (mid land or woynadega) agro ecological zone of the woreda. To select the representative kebeles, we used simple random selection method and four kebeles out of twelve were selected. To identify the representative household heads, list of households from each kebele Agricultural and Development Office is used as sample frame and sampling points are selected by using systematic random sampling method.
3.5 Sample Size
A total of 214 respondents surveyed from the study areas. This 214, sample is determined using the minimum sample size formulae of Fowler (2001) cited by Meneyahel (2015) given by the following formula.
(z) (p)(1 − p) n = − − − − − − − − − − − − − − − − − − − − − (3.1) e
Where, no = sample size, e = the level of risk the researcher is willing to take that true margin of error may exceed the acceptable margin of error = 0.062, z = standard error associated with the chosen level of confidence 94 percent (1.88). And
P=sample proportion in a population (0.52%) house hold participation in coffee production and (0.48%) house hold with non-participating in coffee production
Based on the above formula, the sample size becomes:
1.88 ∗ 0.52 ∗ 0.48 n = = 245 − − − − − − − − − − − − − − − − − − − − − (3.2) 0.06
23
This sample size then can be adjusted to final sample size by considering the total target population of the study area. Therefore, Cochran’s (1977) formula should be used to calculate the final sample size by considering the total target population (Glenn, 2013). These calculations are as follows
= − − − − − − − − − − − − − − − − − − − − − − − − − − − − − −(3.3)
Where, N= total number of the target population of the study area, n0= required return sample size according to Cochran’s (1977) formula=245, and n1= the final sample size.
From the CSA (2007) recent estimation, rural household head of the OubaDamakebel,Shefit 01kebel, kamo kebel and Lame kebel is 591, 349, 323 and 472, respectively. The summation is 1735. The total sample size for the study areas becomes:
245 = = 214 − − − − − − − − − − − − − − − − − − − − − −(3.4) 1 +
Table 3.1.Sample size of each Kebele
No Name of Kebele Total house hold in the Kebele Sampled households 1 Shefit 01 349 43 2 OubaDama 591 73 3 Kamo Kebele 323 40 4 Lame 472 58 Total 1735 214
3.6 Data Collection Techniques and Instruments
Agriculture extension workers collected the primary data; while observation and discussion with woreda as well as kebele governmental officials and expertise conducted by the researcher. The key questions were prepared for discussion with key informants, and governmental officials. The survey was conducted for four weeks in the month of January, 2016. Before the main survey, the enumerators were given a training focusing on the technicalities of the questionnaire. Structured interview questionnaires were designed to collect quantitative data.
24
3.7 Model Specification
3.7.1. Method of Data Analysis and Respective Empirical Models
Different methods can be employed to analyze farm household decision problem. One approach to analyze the issue is to use the well-known Tobit model. However, Tobit model assumes that both the decision to participate in activity and the level of participation are determined by the same variables and with the same sign (Wooldridge, 2002). That is, according to Tobit model, the decision to participate in production of a certain crop and the intensity of production participation are jointly determined and influenced by the same parameters. This is the main limitation of the Tobit model in which it restricts variables and coefficients in the two decisions (production participation and the level of participation decisions) to the same sign and signature (Wooldridge, 2002). That is why recent empirical studies have shown the inadequacy of the Tobit model in cross-sectional analysis, stressing the relevance of alternative approaches.
The appropriate approach is to use the double-hurdle model. This model assumes farmers faced with two hurdles in any agricultural decision making processes (Cragg, 1971; Sanchez, 2005; R.Humphreys, 2010). Accordingly, the decision to participate in an activity is made first and then the decision regarding the level of participation in the activity follows. In this study, thus, double-hurdle model was chosen because it allows for the distinction between the determinants of production participation and the level of participation in coffee production through two separate stages. This model estimation procedure involves running a probit regression to identify factors affecting the decision to participate in the activity using all sample population in the first stage, and a truncated regression model on the participating households to analyze the extent of participation, in the second stage. In our case, we will apply the first stage of double hurdle model to examine the factors determining the decision to participate in coffee production and it is analyzed by a means of the probit regression.
According to Burke (2009), double hurdle model is useful because it allows a subset of the data to pile-up at some value without causing bias in estimating the determinants of the continuous dependent variable in the second stage, hence you can obtain all the data in the remaining sample for the participants (Burke, 2009). Thus, in double hurdle model, there are no restrictions regarding the elements of explanatory variables in each decision stages. That means it is possible to separately analyze the determinants of production 25 participation decision and the level of participation decisions. Due to this separablity, the estimates of production decisions can be obtained by a means of probit regression and that of the level of production participation decision can be analyzed by use of a truncated regression.
The log-likelihood function for the double hurdle model that nests a univariate probit model and a truncated regression model is given following Cragg, (1971) by:
∗ ∗ log = ∑ 1 − Φ( ∗ ) + ∑ Φ( ∗ ) … … .3.5 Where “0” indicates over the zero observations in the sample, while “+” indicates summation over positive observation, Φand∅ refer to the standard normal probability and density function respectively, X*1i and X*2i represents independent variables for the probit model and the truncated model respectively, and are parameters to be estimated for each model’s” is the variance of error terms. .The first portion is the log- likelihood for a probit, while the second portion is the log-likelihood for a truncated regression, with truncation at zero value of the continuous dependent variable in the second stage (the amount of coffee produced in the survey year, in our case). Therefore, the log-likelihood from the Cragg type double hurdle model is the sum of the log- likelihood from a probit and a truncated regression. More useful, is the fact that these two component pieces are entirely separable, such that the probit and truncated regression can be estimated separately (Ground and Koch, 2008; Aristei and Pieroni, 2008; Burke, 2009).
A hypothesis test for the double hurdle model against the Tobit model was examined. The 2 likelihood ratio test statistics ᴦ =-2[ − ( + )]~ X k, where LT is the likelihood for Tobit model; LP is the likelihood for the Probit model; LTRis the likelihood for the truncated regressions model; and k is the number of independent variables in the equations. if the test hypothesis is written as: Ho : λ = and H1:λ ≠ , then Ho is rejected on a pre-specified significance level, provided ᴦ>X2k, confirming the superiority of the double hurdle specification over the Tobit model .In such a cause, the decision to state a positive value for farmers participation in production and the level of participation in production(Greene 2003).
Based on the above backgrounds, the linear probit model can be specified as follows: ∗ = +
26
= 1 ∗ > 0 = 0 ℎ
( = 1) = + + − − − − − − − − − − − − − − − − − − − − − (3. 6) Where Y is the probability of an individual farm household to participate in coffee production, βi is the vector of parameters will be estimated, Xi is the vector of explanatory variables expected to influence the participation decision probability and is the error term.
Probit model specifies the functional relationship between the probability of participating in an activity (coffee production in our case) and the list of various explanatory variables thought to influence the participation decision. These factors can be either continuous or discrete explanatory variables. Therefore, the reduced functional relationship between the binary dependent variable (producing coffee or not) and a list of explanatory variables for the empirical analysis of the current study can be specified as follows using basic probit model specification.
( ( ) = 1) = + ( ) + ( ) + ( ) + ( ) + ( ) + ( ) + ( ) + ( ) + ( ) + ( ) + + − − − − − −3.7
Where Pr - is the probability at which an individual household participate in coffee production represent by (PRODPART=1) and (PRODPART= 0) otherwise.
Βi `s – are the regression parameters, is the error term .The regression parameters estimated by maximum likelihood technique.
Using probit regression method we can compute estimates of the coefficients (β’s) and their corresponding standard errors that are asymptotically efficient. As noted in Wooldridge (2002), the estimated coefficients from probit regression give the signs of the partial effects of each Xi on the response probability (dependent variable). For the continuous explanatory variables, these marginal effects give partial effects of these variables at the sample means. While for the discrete or categorical variables, the marginal effects are used to calculate percentage changes in the dependent variable when the variable shifts from zero to one, ceteris paribus (Newman et al., 2003).
In the second stage of double-hurdle model the researcher examined factors affecting the level of coffee production, conditional on participation decision, which implemented
27 using the truncated regression analysis. Thus, it involves the truncated regression that can be specified as:
∗ = + + = ∗ ∗ > 0 = 1 = 0 , ℎ
From this, we can specify the reduced form of the truncation model as:
= + + − − − − − − − − − − − − − − − − − − − − − − − − − (3.8)
Where Q - the observed quantity of coffee produced, Q* is the latent variable which indicates the level of coffee production is greater than zero, βi is the vector of parameters to be estimate, zi- is the vector of exogenous explanatory variables and is the error term.
The empirical model used in this study assumes that the total quantity of coffee produced in the survey production year (2015/2016) is a linear function of continuous and dummy explanatory variables and is specified as follows:
= + ( ) + ( ) + ( ) + + ( )
+ ( ) + ( ) + − − − − − − − − − − − − − 3.9
Where Q– is the quantity of coffee produced in 2015/2016 production year
(COFFPROD=1)Βi `s – are the regression parameters, is the error term.
Finally, the third objective of the present study can be achieved by defining the amount of income earned from coffee sale as a linear function of continuous and binary explanatory variables. The intention here is to identify important factors explaining marketing of coffee in the study area and determines households’ income which is generated from coffee sale in this area. This can be analyzed by using the truncated regression model, because the dependent variable in this case has many observations at zero. And as noted in Pindyck and Rubenfeld (1991), analyzing such problems using an OLS method would yield biased and inconsistent results (ibid). Due to this we might forced to exclude non- producer farmers from the analysis, because the value of dependent variable (the amount of income earned from coffee sale) is zero for non-producer farmers. Therefore, by using the truncated regression model, we can account for these zero observations; hence this
28 model provides a more accurate estimation (Wooldridge, 2002). Thus, the truncated regression model is chosen and takes the following specification:
∗ = + +
∗ ∗ = > 0 0 , ℎ
+ + − − − − − − − − − − − − − − − − − − − − − − − − − −(3.10)
Where Yi* is the unobserved latent variable; Yi is the actual observed outcome (the level of income generated from coffee sale); βi is the vector of parameters, Wi is the vector of explanatory variables and is the error term.
The empirical model assumes that total farm income earn by a farm household from agricultural product sales is a linear function of continuous and discrete independent variables and is specified as follows:
= + ( ) + ( ) + ( ) +
( ) + ( ) + − − − − − −3.11
Where Yi– is the household income generated from coffee sales (INCOME)
βi’s – are the parameters and Vi-error terms.
3.8. Statistical and Specification Tests
Before executing the final model regressions, all the hypothesized explanatory variables will be checked for the existence of statistical problems such as multicollinearity problems. Basically, multicollinearity may arise due to a linear relationship among explanatory variables and the problem is that, it might cause the estimated regression coefficients to have wrong signs, smaller t-ratios for many of the variables in the regression and high R2 value. Besides, it causes large variance and standard error with a wide confidence interval. Hence, it is quite difficult to estimate accurately the effect of each variable (Gujarati, 2004; Woodridge, 2002).
There are different methods suggested to detect the existence of multicollinearity problem between the model explanatory variables. Among these methods, variance - inflating factor(VIF) technique is commonly used and is also employed in the present study to detect multicollinearity problem among continuous explanatory variables (Gujarati,
29
2004). In Gujarati (2004) it was defined that VIF shows how the variance of an estimator is inflated by the presence of multicollinearity.
According to Gujarati (2004), the larger the value of VIF indicates the more co linearity among one or more model explanatory variables. As a rule of thumb, if the VIF of a variable exceeds 10, which will happen if a multiple R-square exceeds 0.90, that variable is said be highly collinear (Gujarati, 2004).
Alternatively, we can use the inverse of VIF (1/VIF) called Tolerance as a measure of multicollinearity. The closer is tolerance of one explanatory variable (Xi) to zero, the greater the degree of co linearity of that variable with the other regressors. On the other hand, the closer tolerance of Xi is to 1, the greater the evidence that Xi is not collinear with the other regressors (Gujarati, 2004).
Similarly, contingency coefficient (CC) method was used to detect the degree of association among discrete explanatory variables (Healy, 1984). According to Healy (1984), the discrete/dummy variables are said to be collinear if the value of contingency coefficient (CC) is greater than 0.75.Mathematically:
= − − − − − − − − − − − − − − − − − − − −3.12 +
Where CC- is contingency coefficient n- is sample size X2-is chi-square value
Finally, the double hurdle model can be tested against the Tobit model using a standard likelihood ratio test, as the Tobit model is nested in the double hurdle model (Humphreys, 2010). To do so, let LLDH is the log likelihood value from the double hurdle model (which is the sum of log likelihood values from Probit and Truncated regressions) and LLT is the log likelihood value from the Tobit model. Then the likelihood ratio test can be carried out as follows: LR = −2 (LLDH − LLT) and the test statistic has a Х2 distribution with degrees of freedom.
30
Variable Description and their Expected Signs
Dependent variables
Production participation decisions (PRODPART)
This is a binary dependent variable taking value “1” if the farmers participate in coffee production and “0” otherwise.
Amount of coffee produced in cropping season (COFFPROD)
This is a continuous dependent variable and measure in terms of quintal. The researcher used this variable as dependent variable to analyze factors that influence the extent to which farmers decide to produce coffee (the level of production participation, based on the decision to produce the crop) by using truncated regression.
Income generated from sale of coffee (INCOME)
This variable is a continuous dependent variable to analyze factors that determine the income farmers generate from sale of coffee. This allows the researcher to identify those factors that explain the marketing of the coffee in the study area.
Independent Variables:
Total farm size (FARMSZE)
Land is one of the major and the key asset for rural household farmers everywhere. Thus, the decision made by any household is basically and highly influenced by their land holding size. Especially, in my study area the decision to produce cash crop is mainly influenced by farmers land holding size, because cash crop and other staples crops mainly compete for such basic resources. Thus, the researcher expects that a household who holds a greater farm land are more likely to participate in coffee and allocates a significant size for its production. Farm size positively affects farmer’s participation in coffee production.
31
Sex of household head (SEX)
This is a discrete variable that takes a value of “1” if the household head is male and “0”, otherwise. In this study, it is assumed that male household heads have more exposure and access to information and new interventions than female household heads, which might enable them to participate in production of coffee. Thus, male household head expected to participate more than female household heads.
Age of household head (AGEHH)
This is a continuous variable and defined as the number of years of household head age. In this study it is assumed that as age increases farmers would acquire knowledge and experience through continuous learning which help them to actively participating in production of market-oriented cash crops. Thus, in this study this variable is used as a proxy for farmers experience in farming.
Educational level of household head (EDCN)
It is generally recognized that education equips individuals with the necessary knowledge of how to make living. Thus, for the purpose of this study, we believe that those who are literate and have at least some education are better able to make the transition to cash crops. This is so because it is believed that producers with higher levels of education tend to have greater access to production and market information, hence expected to produce market-oriented cash crops.
Number of oxen owned (OXEN)
This is a continuous variable that refers to the number of oxen the respondents own. An ox is the most important means of land cultivation in poor rural areas and is one of the major assets to farm households in Ethiopia. Thus, we expect that the number of oxen available to the household positively enhances the probability of becoming coffee producer and motivates farmer’ significantly.
Number of active family labor (FAMLAB)
Family labor is a continuous variable referring to farmer’s access to family labor. We consider a family labor as active if it can participate in the household agricultural activity. Thus, this variable is expected to positively affect the decision probability to produce coffee and its quantity. This is because coffee is a labour intensive crop, thus requires
32 high labor and in these rural areas there is no employed labor. Thus, family labour is the main source of labor and it has positive effect on both farmer’s participation and level of production participation.
Access to Credit (CREDIT)
Credit access is a dummy variable, which takes value 1 if farmers have access to credit service and 0 otherwise. Since production of any cash crop requires capital which is scarce to most smallholder farmers, it importantly explains farmers’ decision to produce coffee. Coffee especially requires sufficient finance throughout its production processes, farmers who have adequate access to credit service are expected to produce market- oriented cash crops like coffee.
Food sufficiency for the whole year (FOOD)
Smallholder farmers in developing countries are always prone to participate in production of cash crops if they could produce more family foods only. This is because these farmers first want to secure foods for their family. Thus, if farmers have potential and experience in producing sufficient family food for the whole year, such farmers are more likely to participate in production of cash crops such as coffee in the study area.
Household’s access to off-farm activities (NONFRM)
Off-farm activity is a dummy variable indicating farmer’s access to it. If farmers have access to alternative works to farm income sources they are less likely to participate in coffee production. On the other hand, since coffee production requires high working capital it is argued that farmer’s who have access to non-farm activities and generate additional income, are likely to produce high value cash crops such as coffee. Therefore, the impact of this variable on farmers’ decision in coffee production participation is inconclusive.
Distance to extension service centers (DSTEXTN)
This variable is a continuous variable represented by walking distance (in meter) from farmers’ residence/home to the nearest extension service centre. Proximity to such service center is expected to enable regular contacts with agricultural experts, hence motivate to produce coffee.
33
Traveling time to the nearest market place (DSTMRKT)
This is a continuous variable represented by walking time (in minute) from home to the nearest market place. Closeness to market centers may motivate farmers to produce market-oriented crops as it provides easy access to inputs, transport facilities and price externalities. Therefore, closeness to market place is expected to be positively correlated with farmer’s participation in coffee production.
Farmer’s experience on coffee production (EXPER)
This is a discrete variable used to account for farmers’ experience in coffee production. Participation of farmers in coffee production during last two consecutive production years is expected to positively affect farmer`s level of participation in coffee production as it enhances their experience.
Access to market information (MRKTEINFO)
This is a dummy variable taking value 1 if farmers have access to price information by any means, and 0 otherwise. Market information highly influences commodity production, and hence has a significant impact on income earning. Therefore, it is hypothesized that access to price information positively affects the income earn from coffee sale in the study area.
Major buyers of coffee from farmers (COLCTRS, COOP, and MRKTRAD)
This is a dummy variable representing for whom producers sell their produce in the study area. The researcher used three dummy variables in this case. Where dummy variable “COLCTRS” is to mean farmers sold their produce to collectors, variable “COOP” represents farmers sold their produce to cooperatives and “MRKTTRAD” refers to farmers sold their produce to traders at markets. These variables considered since the type of buyer to whom farmers sale their produce may matter for price they receive, hence determine the income earn from coffee sale.
Selling Channels (DIRECT, BROKERS)
This is a dummy variable referring the channel through which farmers sell their produce. This expected to influence the price that farmers can receive which also has an impact on their income. The available options include selling directly to their buyers and selling
34 through brokers. Thus, by using the first dummy variable (DIRECT) as a reference, we can identify whether there is difference in income earn from coffee sale among producer farmers due to using these two channels.
The quantity of coffee marketed (QUANTTY)
This is a continuous variable referring the amount of coffee marketed in the specified year measured in quintal. Quantity marketed is one of the major and key factors in determining the amount of income received by farmers. Even, sometimes this alone determines the amount of income generated from agricultural sale. However, in this study the researcher assume that this variable alone cannot be considered as the determinant of income farmers are receiving from coffee sale, because there are also other factors that determine their income. Whatever the case, the researcher expected positive and significant result for this particular variable.
Market price of coffee (MRKTPRICE)
Own price of coffee is continuous variable and expected to be positively related with income obtained from coffee and farmers production participation as well. It is because farmers’ supply of coffee and their participation on production will depend on its price.
Time of selling (IMMIDIATE, ONEMNTH, TWOMNTH, THREMNTH,)
This is also a categorical variables indicating the time in which farmers sell their produce. These categorical variables allow us to understand the role of time in which farmers sell in explaining the price they charge and hence income they earn. Thus, the researcher expected that these variables explain the income farmers earn from coffee sale
35
Table 3.2 The description and expected sign of farmer`s participation in production and marketing of coffee cash crop.
Variable name Description of Measurement Expec variables ted sign 1 Total farm Household total Continuous variable , measured in + size(FARMSZE) land hold size hectare 2 Sex of house hold Household head Discrete variable and measured as + (SEX) of a farming (1= male headed household, and 0 family otherwise)
3 Age (AGEHH) Age of the Continuous variable, measured in + household years
4 Education(EDC) House hold level Continuous variable, measured in + of education education level 5 Oxen(OXEN) The total oxen the Continuous variable and measured + farmer possesses in number 6 Family Number of active Continuous variable, measured in + labor(FAMLA) family labor number 7 Access to credit Access to credit Dummy variable, measured + (CREDIT) market as(1=access to credit, and 0 otherwise) 8 Food(FOOD) Availability of Dummy variable, measured + family food for as(1=available family food the whole year produce, and 0 otherwise) 9 Off-farm Household`s Dummy variables and measured -/+ activity(NONFRM) access to off-farm as(1= farmers participation in off- activity farm business and 0 otherwise) 10 Distance(DSTEXTN) Distance to Continuous variable measured in _ extension services meter 11 Farmer’s experience experience on Dummy variable and measured + on coffee produ coffee production as(1= produce coffee last two years (EXPER) 0 otherwise) 12 Traveling time to the Traveling time to Continuous variable measured in _ market(TTMARKT) the nearest market minute 13 Access to Access to market Dummy variable measured + information(MARKE information as(1=farmers have market price TINF) information and 0 otherwise) 14 Major buyer of Major buyers of Dummy variable measured + coffee(COLCTRS, coffee from as(1=COLCTRS 2=COOP COOP, and farmers 3=MARKTRAD) MRKTRAD) 36
15 Selling channels Direct sell and by Dummy variable measured + (DIRECT, brokers sell as(1=direct and 2=broker) BROKERS) 16 Quantity of coffee The quantity of Continuous variable measured in + (QUANTTY) coffee marketed quintal 17 Market price Market price of Continuous variable measured in + (MRKTPRICE) coffee ET Birr 18 Time of selling Time of selling Dummy variable measured - (IMMIDIATE, as(1=IMMIDIAT,2=ONEMONTH, ONEMNTH), 3=TWOMONTH,4=THREEMON TWOMNTH, TH and 5=LATTER) THREMNTH)
37
CHAPTER FOUR
RESULTS AND DISCUSSIONS
This chapter presents and discusses main findings of the study. Determinants of smallholders’ decisions to participate in coffee production, the extent of production participation and the marketing issue of the crop in the study area will be presented and discussed.
4.1. Descriptive Results
4.1.1. Socio-demographic characteristics of households
Totally, 214 household heads were considered in this study. Out of which 35 (16.36%) households were female headed and the remaining 179 (83.64%) were male headed. The mean age of the sampled household head is about 40.3 years. The average family size for coffee producers was about 6.4 persons per household, and about 4.8 persons per household for non-producer farmers. Table 4.1.1 presents summary statistics of sampled household’s demographic characteristics in terms of the two sample groups
Table4.1.1: Demographic characteristics of households
Variables Coffee Producers Coffee Non-producers Obs Mean min Max Obs Mean min Max Total HH size 144 6.4 3 10 70 4.8 1 8 Active family labor 144 3.5 1 6 70 1.47 1 4 Dependent family 144 2.9 0 6 70 3.44 0 7
Source: survey result, 2016
The composition of household members in terms of dependent family is nearly similar for coffee producers and non-producers. However, the mean number of active family labour is higher for coffee producers (3.5) and lower for non-producers (1.47).
Educational status of the household head is also an important element in smallholder economic activities. The survey result revealed that 38.31 percent of the sampled farmers never attended schooling, while 61.69 percent were literate at different levels of schooling. Among the literate farmers, majorities (about 33.64%) of them attended schooling below grade five and none of these farmers have attended above grade eight. Table 4.1.2 presents full information on different educational levels of sampled farmers. 38
Table 4.1.2: Educational level of households
School levels Total Sample Frequency Percent 0 82 38.31 1__4 72 33.64 5__8 60 28.05 >9 0 0 Source: survey result, 2016 4.1.2: Land ownership status of farmers Own survey result indicates that about 99.5% of respondents own land. That means, only 0.5% of sampled farmers do not possess their own land. The farm size of farmers varies from 0.1 to 6.5 hectare and the average farm size is found to be 2.64 hectare.
Table 4.1.3: Land ownership of the respondents
Land size (ha) Coffee Non Coffee Total sample Producers (%) Producers (%) (%) < 1 4.15 80 28.97 1.125 < 2.945 29.88 20 26.64 > 3 65.97 0 44.39 Sum 100 100 100 Source: own survey result, 2016
Table 4.1.3 indicates that 65.97% coffee producing farmers possess land size greater than 3 hectares while 80% coffee non-producing farmers possess less than 1 hectare of land. This table depicts that coffee producers own larger land size than non-producers, indicating may be smaller land size inhibits diversification of crops in the area.
From own survey result, 99.7% of farmers who have participated in the production of coffee in 2015/2016 production season cultivated on their own land while the remaining 0.3% of them used rented land in this survey year. The minimum cultivated land under coffee is found to be 0.06 hectare and the maximum is 5.2 hectare. And the average cultivated land under coffee in this survey year is about 0.68 hectare.
39
4.1.3. Livestock ownership of households
Livestock is one of the major assets for farmers. And it is one way of indicating farmer’s level of wealth in some areas of Ethiopia, since the number of livestock owned by each household is considered as the indicator of living standards in rural areas. Especially, in a mixed farming system the contribution of livestock to crop production is great. For example, livestock in the study area (OydaWereda) can be used as an alternative source of income, as a means of transportation and serve as a store of wealth. Oxen and donkeys are among the major livestock resource used in any crop production. These two resources are considered as the main influential variables in decision of farmers to produce and to what extent they can participate in any agricultural production. For example, this survey result revealed that, 96 percent of the sampled households use ox for land preparation purposes, while the remaining portion use traditional hand hoe for the same purpose.
Table 4.1.4: Oxen and donkey ownership of households
Number Coffee Producers Coffee Non producer Total sample Oxen Donkey Oxen Donkey Oxen Donkey 0 8.33 6.25 54.29 18.59 23.36 10.28 1 12.51 32.65 40 64.28 21.49 42.99 2 35.41 59.72 5.71 14.28 25.72 44.85 >3 43.75 1.38 0 2.85 29.43 1.88 Source: survey result, 2016
As we can observe from table 4.1.4, 8.33% of coffee producers; 54.29 percent of non- coffee producers and 23.36% of the whole sampled farmers have no any oxen. Similarly, 6.25% of coffee producers, 18.59% of non-coffee producers, and 10.28% of the overall sampled households own no any donkey. In addition, according to the survey result, about 12.51% and 32.65% of coffee producers owns only one ox and donkey, respectively. Similarly, 40% and 64.28% of non-coffee producers own only one ox and donkey, respectively. This survey result also revealed that, on average, about 21.49% and 42.99% of the total sampled households owns only one ox and donkey respectively. Further, 79.16% of coffee producers own 2 and above oxen. However, only about 5.71% of non- coffee producers own 2and above 3 oxen. Only 29.43% of sampled farmers reported as they own three and above oxen and all these farmers are coffee producers
40
4.1.4. Coffee production and associated problems
Respondents listed a number of problems associated with coffee production in the study area. Among the major problems, lack of improved coffee seed in the study area is the most serious problem reported by sample respondents which is 49.3%. The main problem in this regard is that, farmers usually use the traditional seed that results in low crop yield and is vulnerable to climate change associated problems. This requires immediate intervention through finding improved coffee seeds that properly fit the agro-ecology of the area.
Table 4.1.5: Farmers perception about problems hindering participation on coffee production
Problem type Frequency Percent Disease 68 32.00 Pest infestation 35 16.35 Rain fail problems 5 2.35 Improved coffee 106 49.3 seed Source: Own survey, 2016
Coffee diseases are another serious problem that affects coffee production in the study area. These farmers very much complain about this problem as it destroys the coffee crop at different stages, and discourages farmers participate in coffee production. Respondents also complain on the local governments for not responding to their complaints regarding coffee disease even though reported several times. Pest infestation is another challenge hindering coffee production. About 16.35% of farmers responded pest infestation is one among the major problems that affects coffee production and productivity in Oyda. Rainfall irregularities are mentioned by farmers as one of calamities affecting coffee production and productivity. Thus, if better participation of smallholder farmers in coffee production is to be achieved in the study area, it is necessary to encounter the articulated impediments.
4.1.5. Income sources of households
The survey data revealed that the major source of income for the farmers is on-farm activities (both from crop and livestock production). 35% of the respondents reported involvement in non-farm activities to generate some additional income. The other 41 possible source of cash for rural household is credit. Accordingly we asked the farmers if they have access to credit from any rural institution. From total of the respondents (exactly 35.5%) replied that they have access to credit. The remaining 64.5% answered they have no access to any credit.
For the majority (65%) crop production is the only source of income. Maize, chat and coffee are the main sources. On average, farmers generated income birr 32123.99 from sale of coffee in year 2014/2015. In 2015/2016 production season, the average income from sale of coffee is about 48636.94 birr, ranging birr 1000 – 190000 per farmer see table 4.1.6.
Table 4.1.6: Summary of income from coffee sells
Years Obs Mean Std. Dev. Min Max
2014/15 144 32123.99 31066.32 456 149000
2015/16 144 48636.94 39328.73 1000 190000
Source: Own survey, 2016
4.1.6. Coffee Marketing Practices in Oyda
The majority (97.44%) of coffee producer exchange coffee in market and about 76.39 % sell directly to traders or purchasers at nearby markets. About 23.61% of coffee producers sold their produce through brokers (Figure 4.1).
Figure 4.1 Selling channel
Most farmers (33.35) sold their coffee produce immediately after harvest,(26.38%) sell their produce one month later after harvest, and the remaining(23.6 %) sell two month
42 after harvest. Only 16.66% of coffee farmers’ store their produce for three months after harvest. The reason they present is that coffee seed lost its weight and quality if it is stored at home longer and they have no well managed cemented places. Thus, they prefer to sell immediately after harvest (Figure 4.2). The respondents also complained about existence of serious problems at market place at the time of selling their produce.
Figure 4.2 Time of selling coffee
Source: own survey, 2016
4.1.7. Institutional Issues on Coffee Production in Oyda
Among the institutional issues, membership status of farmers in rural cooperatives, access to different technical advisory services and access to any contractual opportunities with different bodies in production/marketing of agricultural products were assessed (Figure 4.3).
Results of the survey revealed that about 13.08 percent of the respondents were a member of local cooperatives. This percent show us large number of respondent farmers is not member of local cooperatives the reason may be lack of awareness of cooperatives. In addition, different necessary technical advisory services from agricultural extensions are also important and required by rural farmers. In this regard farmers are expecting
43 more services from these institutions. One very important issue they raised is the problem of different diseases affecting different crops such as Coffee, mango, chat and banana. They are looking for immediate solutions for the problems attacking these crops.
80
86.92% 60
Percent
40
20
13.08% 0 Not member of coop .Member of coop
Figure 4.3 Farmers membership states of cooperatives.
Source: own survey, 2016
4.2. Econometric Results
4.2.1. Production Participation (Probit regression)
In this section, we analyze factors affecting farmers’ participation decision in coffee production. To analyze the problem we employed the probit regression and eleven explanatory variables (seven continuous and four discrete), were hypothesized to influence the probability of participation decisions and included in the analysis.
However, prior to running the final regression analysis, both continuous and discrete explanatory variables need to be checked for existence of multicollinearity using Variance Inflating Factor (VIF) and the contingency coefficient (CC) methods, respectively. Accordingly, as can be seen from the results presented in Appendix 10 and 44
11, our test result suggests that, there is no serious multicollinearity problem in our model, since there is no strong association among the hypothesized explanatory variables. Therefore, all of the proposed potential explanatory variables were included in the final probit regression.
The probit model regression was carried out and the result is presented in Table 4.2.1. From the regression result, the joint significance of the explanatory variables were tested by using the Wald test with a null hypothesis of coefficients of all explanatory variables included in the models are equal to zero. The Wald test, which follows χ2 distribution with 11 degrees of freedom (DF), is about 245.46. From χ2 distribution table with 11df the critical value is 19.675 at 5 % level of significance. This implies that the joint null hypothesis of all slope coefficients of explanatory variables are equal to zero is rejected see table 4.2.1. Thus, the overall significance of the model is good (i. e. Explanatory variables have some joint effect on farmers participation in production). The estimated probability greater than chi-square value (Prob> chi-square = 0.0000), suggests that all the model parameters are jointly significant in explaining the dependent variable at less than 1and 5 percent significance level.
Significant explanatory variables from probit regression: Out of the included regressors, the coefficients of six variables were found to have a significant impact on the likelihood of participating in the production of coffee in the study area. According to Wooldridge (2002), the probit regression coefficient gives signs of the partial effects of each explanatory variable on the response probability of the dependent variable.
Household’s landholding size (FARMSZE): The estimated coefficient result for this variable was found to be positive, reflecting positive effect on producing coffee. This result implies that farmers, who have more farm size, are most likely to produce coffee, keeping the effects of other variables constant. In other hand, it indicates as households’ farm size increases, the probability to produce coffee increases, ceteris paribus. This result is expected since land is one of the basic factors of production in any agricultural activities, including cash productions. This is supported by the obtained statistically significant coefficient at less than 1 percent probability level, which confirms the logical association between producing any cash crop and the level of farm size owned by smallholder farmers. The study by Poulton et al
45
(2001) suggests that land is an important factor in influencing farmer’s decision to produce any cash crop, hence support the finding of the current study.
Number of active family labour (FAMLAB): The estimated result also shows that, having more working family member increases the probability of producing coffee. The positive and significant coefficient obtained for this variable confirms that, existence of higher number of working family labour encourages the production of coffee as a cash crop. The result is expected since family labour is the major source of labour force in the area, hence those households who have access to more family labour are likely to produce more quantity of coffee. The reason is that labour markets are lacking in this area but coffee production from land preparation to its harvest requires labour. For example, coffee harvesting is a very critical activity which should be completed at a short period of time; otherwise insects and associated problems can damages and decrease quality of the crop within a short day. This suggests that labour is among the critical variable in influencing decisions of households to produce coffee. The findings by Sorsa, D. (2009) support the finding of the present study.
Access to credit (CREDIT):
The obtained result for this variable confirms that access to credit service significantly influences the likelihood of producing coffee. The estimates show that, farmers who have access to credit are more likely to produce coffee than their counterparts, ceteris paribus. The plausible explanation is that, access to credit enables smallholder farmers to finance purchase of inputs and other production equipment’s, hence encourage farmers to produce a given cash crop like coffee.
Thus, as credit becomes more available for farmers, they are more likely to produce market oriented crops. The findings by Immink and Alarcon (1993); and Lerman (2004) support the finding of the current study by arguing for agricultural credit as it plays a vital role in the process of smallholder commercialization.
46
Table 4.2.1: Determinants of coffee production participation (Probit regression)
Variables dY/dx Std. Err. Z p>| | [95% Conf. Interval ] x-bar
FARMSZE .002136 .0070319 3.09 .002*** 2.64436 -.011645 .015919
AGEHH .045499 .0015933 0.82 .705 42.4432 -.002623 .003622 SEX(male) .027628 .0976564 0.76 0.447 .836449 -.163775 .219031 FAMLAB 001257 .0041486 4.80 0.001*** 2.83645 -.006873 .009389 DSTEXTN -.06890 .090006 -1.19 0.0235** 2.83778 -.030006 .010006 TRVTMRKT -.00025 .0008157 -2.22 0.226 30.5187 -.001853 .001344
CREDIT .004035 .0113968 5.88 0.000*** 1.64486 -.016373 .018302 NONFARM .001243 .0035436 0.07 0.332 .636440 -.004627 .267463 FOOD .001234 .0076211 4.06 0.003*** 2.46753 -.011754 .825362 OXEN .001004 .0032004 6.44 0.000*** 1.78505 -.005268 .007277 EDCN .000854 .0027765 2.34 0.219 2.56075 -.004588 .006296 Number of observations 214 Log pseudo likelihood -13.252236 Wald chi2(11) 245.46 Pseudo R2 0.5025 Prob> chi2 0.0000 ** Significance at 5%, ***significance at 1%, Source: own survey, 2016
Distance to extension service centre (DSTEXTN): Evidence from the probit regression result also indicates that the actual distance of households’ home from extension service centre significantly influences the probability decision to produce coffee in the study area, which is statistically significant at less than 5 percent probability level. The estimated coefficient for this variable shows that there is a negative correlation between distance from agricultural extension service centers to households’ home and the likelihood of producing coffee. This result suggests that farmers require advisory and other services to actively participate in production of market oriented crops, thus those farmers who live near the extension service centre are more likely to participate in production of the considered crop, ceteris paribus.
47
Number of oxen owned (OXEN): The estimated coefficient for this variable suggests that, having more number of oxen increases the probability of producing coffee and their association is statistically significant at less than 1 percent significance level. This result indicates that household’s who have a larger number of oxen are more likely to participate in production of coffee, keeping the effects of all other variables at constant. This is so since ox is used as a major means of land preparation in the study area. The survey finding by Sorsa (2009) supports the finding of the current study. According to Sorsa (2009), in east Wellega zone, more than 92.5 percent of sampled farmers used oxen for land preparation purposes and the remaining use hand hoe. Thus, we obtained evidence that shows the importance of having more oxen is influencing the likelihood of producing coffee in the study area(Cadot et al 2006). Availability of family food (FOOD): Our regression result also reveals that, availability of family food for the whole year has a substantial effect on increasing the probability of producing coffee in the study area, keeping the value of other variables constant. The plausible explanation is that as farmers have good experiences and ability to produce the family food for the whole year, their likelihood to participate in the production of high value cash crops like coffee is higher under ceteris paribus assumption. In other words, this is to mean households who can produce family food for the whole year are more likely to produce coffee than those farmers who cannot produce the family food for the whole year. This is informed by the obtained coefficient result for this variable with positive sign and statistically significant at less than 1 percent significance level. The study by G.Lukanu et al (2004) verified that household food availability is one among the factors that affects farmers’ decision to cultivate a given cash crop, hence supports the current finding.
4.2.2. Factors Determining the Extent of Coffee Production Participation in Oyda
This section focuses on factors determining the extent of farmers’ coffee production participation, conditional on decision to produce coffee. Truncated regression is used in this case, which is the second stage of the double-hurdle model, to analyze the problem.
However, before running the final regression, it is necessary to check for existence of statistical problems such as multicollinearity. In this regard, we employed the Variance
48
Inflating factor (VIF) technique for continuous explanatory variables and the Contingency Coefficient (CC) method for discrete regressors. These test results are presented in Appendix 2 and Appendix 3.
According to Gujarati (2004), VIF value greater than 10 indicates a severe collinearity among regressors. Similarly, Contingency Coefficient (CC) test uses a correlation coefficient of 0.75 as its tolerable critical value in which CC value more than 0.75 indicates collinearity problem. The test estimates show that there is no serious correlation among the proposed explanatory variables.
The presence of heteroskedasticity was checked by Brush pigan test and the p value were 0.000; imply that absence of the problem of heteroskedasticity (appendix4).
A model specification error can occur when one or more relevant variables are omitted from the model or one or more irrelevant variables are included in the model .It can substantially affect the estimate of regression coefficients .more over the model specification errors were checked by linktest ,the test of hat and hatsqwere 0.000 and 0.000 respectively which are significant. This is to say that the linktest has failed to reject the hypothesis that the model is specified correctly. There, fore it seems to us that we do not have a specification error (Appendix 5).
Here, we also use the likelihood ratio (LR) test to check the irrelevance of Tobit model in this case. On the basis of a likelihood ratio (LR) test, the Tobit model was found to be irrelevant (LR= 29.192), with a critical Х2 (7) value of 18.475. The implication of this result is that production decision participation and the level of production participation decisions are not based on the same decision-making process. That means, these two decisions are influenced by different parameters with different signs and signatures. This supports the inadequacy of the Tobit model in our case Appendix 8 and Appendix 9.Informed by these test results; we proceed to present the truncated regression result. Analyzing the estimated parameters, it is possible to highlight that the coefficients of five variables are statistically significant at different significance levels.
One of the significant variables in influencing the level of coffee production participation in the study area is the number of active working family members (FAMLAB). This variable has an important impact on the extent of farmers’ coffee production participation and the result was significant at less than 1 percent probability level. This positive and
49 significant obtained coefficient reveals the importance of family labor in the intensity of coffee production participation as well as the decision to produce crop. The possible explanation is that as we have said in the above section (probit analysis), coffee production is labor intensive and in rural areas where labor markets are non-existed or lacking, family labor is the key and the only source of farming labor. Thus, access to more family labor significantly influences farmers’ participation decision in any agricultural activity and determines the level of participation in those activities.
In line with this, we found similar result from the probit regression, in which this variable significantly and positively influences households’ decision probability in coffee production participation. This shows the importance of working family labor to participate in production of coffee as a cash crop in the study area.
The next significant variable in truncated regression estimation is the dummy variable indicating access to rural credit service (CREDIT). The estimated coefficients for this dummy variable reveal the existence of different level of coffee production participation based on credit access status. The obtained result suggests that, those farmers who have access to credit service are more likely to produce significant amount of coffee than their counter parts, ceteris paribus.
This highlights the importance of access to rural credit service in both decision to produce coffee and the level of production participation in the study area.
Household’s landholding size (FARMSZE) is one of the significant variables which have positive coefficients. Truncated regration result implies that farmers, who have more farm size, are most likely to produce coffee, keeping the effects of other variables constant. In other hand, it indicates as households’ farm size increases, the level of coffee produce increases. This result is expected since land is one of the basic factors of production in any agricultural activities, including coffee productions. This is supported by the obtained statistically significant coefficient at less than 5 percent probability level, which confirms the logical association between producing any cash crop and the level of farm size owned by smallholder farmers.
50
Table 4.2.2: Determinants of the extent of Coffee production participation
Variables Coefficient Std. Err. Z-Value P>|z| [95% Conf. Interval] SEX(male) 0.7140052 1.403466 0.51 0.611 -2.036739 3.464749 EDCN 0.0232598 0.1507177 0.15 0.877 -.2721414 .3186611 FAMLAB 1.423359 0.39982 3.56 0.000*** -.1618832 1.405382 OXEN 1.572168 0.5458919 2 .88 0.001*** -.0434292 2.096428 CREDIT 1.220066 0.7011877 5.74 0.039** -1.893812 .854793 EXPRINCE 1.584189 0.565782 2.80 0.002*** -.2483524 5.889401 FARMSIZ 3.190613 1.393281 2.29 0.023** -1.462894 2.68424 CONS -1.498996 2.867928 -0.52 0.601 -7.120032 4.122039 SIGMA 4.665559 1.014821 4.60 0.000 2.676546 6.654571 Number of obs 144 Wald chi2 (7) 182.48 Prob> chi2 0.0000 Log pseudo likelihood -416.98384
***, ** shows significance of the coefficients at 1% and 5% levels, respectively
Source: own survey, 2016
The estimated coefficient for the dummy variable indicating households’ experience on coffee production (EXPERNCE) reveal the positive and significant impact of this variable on the level of coffee production participation in the study area. The plausible reason is that, farmers who have an experience on the coffee production are most likely to produce the crop in significant amount. This result indicates that household’s, who have produced coffee at least in the last two years produces more amount of coffee than their counterparts keeping the effects of other variables constant.
The positive and significant coefficient obtained from the variable “OXEN” gives evidence that shows the number of oxen owned has a positive and significant impact on the level of coffee produced by sampled farmers. This implies that households who have more oxen are likely to produces more amount of coffee, giving evidence that shows oxen is one of the important factor in determining the extent of coffee production participation in OydaWereda. This further indicates that coffee production in the study area is still dominated by the traditional means. The result from this regression, coupled with result
51 obtained from the previous probit regression, confirms the key role of having more oxen in coffee production participation in the study area.
Household head sex and education level have positive coefficients but insignificant impact on level of coffee production the reason may be the education level of household heads low due to these level of coffee production participation more significant on other variables.
4.2.3. Factors Affecting Coffee Marketing in Oyda
This section focuses on factors explaining marketing of coffee by smallholder farmers in Oyda. The objective is to analyze factors that affect marketing of coffee, by taking the amount of income generated by sampled households as a dependent variable. We can run our model and analyze the problem, given that all the proposed regressors are uncorrelated with the error term, assuming all regressors are exogenous.
However, as we have done in the previous sections, we should carry out statistical tests for the proposed regressors before using these variables in the final estimation. Accordingly, we carried out tests of multicollinearity (both for continuous and discrete variables) by applying VIF and contingency coefficient (CC) techniques. As one can observe, we obtained from both Appendix 6 and Appendix 7 that, there is no serious linear correlation among the proposed explanatory variables, which can cause a multicollinearity problems. Therefore, all the proposed variables were included in the final regression. We use the corrected – robust t-ratio since we suspect heteroskedasticity problem, which is commonly arise in a cross sectional data. As noted in Verbeek (2004) if we use the robust standard error, the resulting test statistics are appropriate, whether or not the errors have a constant variance. For Heteroskedasticity and link testes see Appendix 12. After all, ten variables entered the final regression and the estimated coefficients of these variables are reported in the Table 4.2.3.Out of the included explanatory variables, eight (8) variables were found with statistically significant coefficients. Out of these significant variables, the coefficients off our variables were found with positive signs, implying direct correlation of these variables with the dependent variable. In contrast four significant variables were found to have a negative signs, indicating the inverse relationship between these regressors and the dependent variable.
52
Turning to individual explanatory variables, the quantity of coffee marketed (QUANTTY) has a positive and significant effect on the derived income from coffee sale in the study area. This outcome is expected and logical, since there is positive relationship between quantity supplied and income generated. This result indicates that, the amount of coffee marketed is one among the major factors determining the amount of income earned from coffee sale in the study area.
We obtained result that confirms direct relationship between income generated and the price of the commodity at markets. The positive and significant coefficient obtained for this variable, highlights the evidence that show the coffee price and the received income are positively correlated. The result is also statistically significant at less than 5 percent significance level. The descriptive survey result reveals that, sampled farmers receives different price levels for their coffee produce, hence the market price importantly determines their level of income could be derived from coffee sale. This is supported by (Abdurahman 2005).
The next significant variable in determining the income earned from coffee sale is “TRVMRKT” which represents traveling time to the nearest market place. Negative sign coefficient was obtained for this variable from the regression result, giving evidence that show the income earned from coffee has inversely affected by the longer walking hour from households home to the nearest market place they sale their coffee produce. This implies longer travelling time negatively affects smallholder farmer’s income. The outcome is expected because, long traveling time from market centers affects the price of the crop, hence producer farmers prefer to sale at local area to local traders at lower prices. In addition, long traveling time from market is one of the transaction cost related problems, which is common in rural areas where access to transportations is non- existence, the problem is serious. Thus, the actual time between farmers’ home and the nearest market place is one of the determinant factors in influencing the amount of income earned from coffee sale in the study area, other things being constant.
We also obtained result that confirms direct relationship between income generated and access to market information. The positive and significant coefficient obtained for this variable, highlights the evidence that show access to market information and the received income are positively correlated. The result is also statistically significant at less than 5 percent significance level.
53
Table 4.2.3: Factors affecting income earned from coffee sale in the study area
VARIABLES Coefficient Std. Err. Z-Value P>|z| [95% Conf. Interval] MRKPRIC 10.67007 1.261571 8.46 0.030 ** 8.19744 13.14271 TRVMRK -398.274 158.6752 -2.51 0.031** -550.5304 71.4648 MRKINFO 6609.816 3240.106 2.04 0.011** -4602.984 8097.998 QUNTITYMRKT 3501.309 339.1142 10.32 0.000*** 2836.657 4165.961 SELLING TIME ONE MONTH -9907.447 2899.691 -3.42 0.001*** -15590.74 -4224.156 TWO MONTH -10487.44 4298.134 -2.44 0.001*** -14595.95 2252.423 THREE MONTH -41172.60 15420.45 -2.67 0.005 *** -55935.35 4511.694 To whom Coffee sold (Buyers) COOP 7973.9345 2523.397 3.16 0.004*** -2029.896 7861.639 TRADERS 13647.75 4457.561 3.06 0.242 4911.09 22384.41 BROKERS 5749.708 2275.416 2.53 0.312 1289.975 10209.44 Constant -32845.03 13235.6 -2.48 0.013 -58786.32 -6903.737 Sigma 11419.81 1122.465 10.17 0.000 9219.818 13619.8 Number of obs 144 Wald chi2(10) 1084.63 Prob> chi2 0.0000 Log likelihood -1514.791 ***, ** shows significance of the coefficients at 1% and 5% levels, respectively
Source: own survey result, 2016
Furthermore, the estimated coefficients for dummy variables indicating coffee selling periods(ONEMNTH, TWOMNTH and THREMNTH) shows the inverse correlation between selling coffee in latter months after harvest and the income earned. These are dummy variables we constructed to identify how the difference in coffee selling time can affect the income of farmers which could be derived from the sale. Accordingly, we constructed four dummy variables including: IMMIDIATE (to refer farmers have sold their coffee produce immediately after harvest), ONEMNTH (to refer farmers have sold their produce one month later after harvest), TWOMNTH (which represent farmers waits for two months after harvest and sold their coffee produce), THREEMNTH (to refer they sold their produce three months later after harvest. In this case, the first variable (IMMIDIATE) was considered as the reference dummy variable for comparison purpose 54 and the rest of three dummy variables were included in the model regression. For three variables, we obtained negative sign coefficients which imply selling coffee at one month and later after harvest results in less income from coffee sale, when compared to income earned from selling it immediately after harvest. The coefficients of all these dummy variables were found with significant results.
As one can observe from the regression result (Table 4.2.3) the difference in the amount of income earned from coffee sale increases as farmers sale at latter periods which reduces the income. For example, the estimated coefficient for variable “ONEMNTH” is only negative9907.447, but the estimated coefficient for variable “THREMNTH” is negative 41172.60. This indicates that, selling coffee one month later after harvest results in less income when compared to selling immediately after harvest, keeping the effects of other variables at zero. However, those farmers who sold their coffee produce three months later generate less income than those farmers who have sold their produce immediately after harvest. These all results, gives evidence that show coffee selling time explains and determines the income farmers derive.
Next, we discuss how the type of coffee buyer (to whom sell) affects famers’ income from the sale. These dummy variables are represented by as “COLCTRS” to refer the household sold the produce to traders at their area local collectors, “COOP” to refer the household sold to local cooperatives, and “MRKTRAD” representing the household sold to traders at markets. This is considered since different buyers can provide different prices, hence affects the farmers income. In this case, “COLCTRS” was used as a reference dummy variable and the remaining two variables COOP and MRKTRAD were used as additional regressors. The estimated coefficients for these variables shows that selling to cooperatives and traders at markets results in more income from coffee sale when compared to selling to local traders, ceteris paribus. However, only the result from one variable was found to have a significant impact, giving evidence that shows the difference in income earned from the two types of buyers (local collectors and cooperatives) is significant. This result suggests that farmers secure better income from being selling to local cooperatives which indicate cooperatives are an important institutional innovation in encouraging farmers to produce cash crops, in which it provides better incentive for their participation.
55
CHAPTER FIVE
CONCLUSIONS AND POLICY IMPLICATIONS
5.1. Conclusions
Coffee is the major cash crop for smallholders in OydaWereda. And there is a potential arable land for further production in the Wereda. The production technique was still dominated by traditional means (more than 95 percent of sampled farmers use oxen for land preparation, the remaining use traditional equipments like hand hoe). Lack of improved seed, lack of awareness about the importance of coffee in the area and lack of knowledge and capacity to use fertilizer for coffee production are the other major factors resulting in low productivity of the crop in the study area. This discourages farmers to produce coffee, despite the available potential and opportunities. Fear of crop failure due to rain irregularities and existence of different coffee diseases (which currently become a common problem in the Wereda), were also found as the major determinant factor in limiting smallholder farmers participation in coffee production.
In addition, from the probit model regression, we observed that number of family labour, number of oxen owned, the size of farmland owned, family food availability and access to credit service influences the decision probability of farmers to produce coffee in the study area, positively and significantly. We also obtained that, distance to agricultural extension service centers decreases the likelihood of farmers to produce coffee and its impact was found to be statistically significant. This result suggests that, household specific characteristics and asset endowments are the major determining factors for smallholder farmers to produce coffee in the study area.
Individual household specific factors matters for different level participation status of smallholder farmers in coffee production in the area, which results them in differently responding to the available potential and opportunities. In addition, access to rural credit service was found to be a significant factor, both in participation decision and the level of coffee production participation in the study area. This implies that credit availability is one of the key institutional factors that determine farmer’s decision status in coffee production in OydaWereda. In addition, the number of family labour and the number of oxen owned significantly influences coffee production participating (both decision to produce and how much to produce) in the study area. This is because coffee production is
56 a labour intensive and oxen ownership is the major means of land preparation for the crop production, hence these two variables were found to be the major determinant factors both in decision to produce the crop and the extent of production participation. These two variables are also household specific factors in determining farmers’ participation status in coffee production in the study area.
Furthermore, from the truncated regression result, we highlighted that although the income farmers generate from coffee sale increases with the amount of coffee marketed, the relationship was found to be not a one-to-one. That is, in addition to the quantity of coffee marketed, other factors explain and determine significantly the amount of income earned from coffee sale in the study area. Accordingly variables such as market price of coffee produce, the usual selling time, travelling time to the nearest market and usual buyers of coffee were found to be the major and the significant ones. One possible conclusion from this result is that, time at which farmers will sale their coffee produce matter in generating better income from coffee. Thus, it is better for farmers to sell their coffee produce before two months after harvest, the survey result reveals. The type of coffee collectors or traders who buys coffee from farmers also matters for variations in income earned from coffee sell. Cooperatives are found to be the major channel for farmers to secure better income from coffee produce in the study area. This is because cooperatives are believed to pay better price and provides other market related information; hence those farmers who have sold their produce to local cooperatives were found to generate better income than others. Market price was also found to be an important factor in securing better income from coffee sells for smallholders. This is because coffee is one of the international crops in which its price is linked to international markets; hence market price is necessary and significantly determines the level of income farmers derives.
5.2. Policy Implications
Coffee as usual maintains its role as important agricultural export commodity for Ethiopia. However, its production is significantly dominated by smallholder farmers and is limited to selected areas in the country due to limited availability of agro-ecological zones for its production and productivity. Therefore, to promote and encourage farmers in production of this cash crop, a number of improvements are required. Based on the
57 findings of the study, the following points need to be considered as possible recommendations.
Sampled farmers complained about lack of improved coffee seed varieties in the area. In this regard, farmers require immediate intervention and support. Therefore, providing improved coffee variety that properly fit the agro-ecology of OydaWereda is one possible solution. In addition to this, smallholder farmers have complained about the crop failures at different stages due to coffee diseases, rainfall related problems, and pest infestation problems. This requires research and development works in the area to sustainably solve these problems. Furthermore, sampled farmers have also complained about lack of awareness and capacity to use fertilizer for coffee production. Therefore, building smallholder farmers’ knowledge on fertilizer application and improve fertilizer supply for coffee production is essential.
Furthermore, the findings of this study suggests that institutional services like producer cooperatives and credits are the key factors in influencing both farmers decision to participate in coffee production and the enhancing level of production participation. This is so because coffee production entails high working capital throughout its production processes. Thus adequate availability of credit service can help to facilitate farmers to participate in its production and to produce a significant amount. Broadening and expanding sources of such institutional service is another possible recommendation from the present study, if active participation of smallholder farmers is required in coffee production and marketing in the study area. In this regard, contract farming activities and experiences are importantly needed to facilitate farmer’s participation and to make conducive environment for these marginalized farmers. As different literatures suggests, contract farming practices are important through different ways in smallholder commercialization through cash crops, especially by solving the liquidity constraint for farmers, by increasing the quality of product produced and in reducing transaction costs. Thus, developing the contract farming practices is another important recommendation from the current study. And as coffee is a smallholder crop; which is produced by a number of farmers at very remote and hardly transported places, infrastructure investments are also needed and recommended to encourage farmers in production of such high value-export potentials crops in the country.
58
6. References
Abdurrahman, A. (2005) - Coffee supply response in the HarargheHighlands. Master’s thesis. Addis Ababa University.
Abu,T. (2012).Coffee Annual Report.GAIN report number African Centre for Economic Transformation,http://acetforafrica.org/Agricultural Growth in Sub- Saharan Africa. Development Policy Review, 24(November).
Alemayehu, E. (2010). Analysis of Factors Affecting the Technical Efficiency of Coffee Producers in Jimma Zone: a Stochastic Frontier Analysis .Master’s thesis .Addis Abeba University
Alemayehu,A. (2014).Coffee Production and Marketing in Ethiopia.European Journal of Business and Management Vol.6, No.37, 20142004), 243-277.
Alemseged,A. and Getaneh,A.(2013). Ethiopian Coffee Exporters Association ECEA:
Alemu, D. and Gerdien, W. (2010); Sesame Traders and the ECX: An Overview with Focus on Transaction Costs and Risks, VC4PPD Report #8, Addis Ababa
Anwar, Ab.( 2010). Assessment of coffee quality and its related problems in Jimma zone of oromia regional state. Master’s thesis. Jima university.
Aristei, D. and Pieroni, L.(2008). A double-hurdle approach to modeling tobacco consumption in Italy. Applied Economics, 40(19):2463 – 2476.
Aysheshm, K. (2007). Sesame market chain analysis: the case MetemaWoreda, north Gondar zone, Amhara national Regional State. Master’s thesis. Haramaya University.
Bayissa,G, (2014) A Double-Hurdle Approach to Modeling of Improved Tef Technologies Adoption And Intensity Use in Case of Diga District of East Wollega Zone Global Journal of Environmental Research 8 (3): 41-49, 2014 DOI: 10.5829/idosi.gjer.2014.8.3.1106 (theory concept)
Boughton D., Mather.D, Barrett.C, Benfica R., Abdula D., Tschirley.D and Cunguara. B, (2007); “Market Participation by Rural Households in a Low-Income Country: An Asset- Based Approach Applied to Mozambique;” Faith and Economics Vol 50: 64-101
Burke, William (2009), Triple Hurdle Model of Smallholder Production and Market
59
Participation in Kenya’s Dairy Sector,.Msc. thesis.
Cadot, O., L. Dutoit and M. Olarreaga (2006); How Costly Is It for Poor Farmers to Lift Themselves Out of Subsistence? World Bank Policy Research Working Paper 3881
Central Statistical Agency Of Ethiopia (CSA, 2007) the Population and Housing Census of Ethiopia,Statistical Abstract, February, 2008.
Chauvin, Nic. (2012); more cash from cash Crops: Giving Farmers a Better Deal in Africa, African Centre for Economic Transformation, http://acetforafrica.org/
Cochran, W. G. (1977). Sampling techniques (3rd Ed.). New York: John Wiley & Son.
Cragg, J. (1971). Some statistical models for limited dependent variables with application to the demand for durable goods, Econometrical39, 829-844.(Germ thesis)
CSA(Central Statistical Agency). 2013. Agricultural Sample Survey 2012/2013 (2005E.C.), Report on area and production of major crops. Central Statistical Agency, Addis Ababa
De Janvry, A., M. Fafchamps, and E. Sadoulet. 1991. Peasant Household Behavior with Missing Markets: Some Paradoxes Explained. Economic Journal 101(1991):1400– 17.EPAR Brief No. 94
Ethiopia Commodity Exchange Proclamation (ECEP), 2007, Art 20(1), Proclamation No 550, Federal Neg. Gaz.,13th Year, No 61
FAO,(2011). Development Economics Division the Food and Agriculture Organization of the UnitedFund I. M, 2013. The federal democratic republic of Ethiopia
Federal democratic republic of Ethiopia Ministry of Agriculture and Rural Development (MARD)(2010).Addis Abeba Ethiopia
Federal Democratic Republic of Ethiopia Ministry of Trade Coffee Opportunities in Ethiopia February (2012). Addis Ababa EthiopiaFoodCrop Debate" Crop Post Harvest Program, Issue Paper 3.From Zimbabwe. Economic Development and Cultural Change, 42(2), 387–402.
G.Lukanu; Maryann, Green; Peter, Greenfield & Steve, worth (2004) .Farmers' cash crop
60
Cultivation decisions in Southern Niassa province, Mozambique, Development Southern Africa, 21:3, 531-554
Gebreselassie,S., and Sharp. K,(2008).commercialization of smallholder agriculture in selected Teffee growing areas of Ethiopia, Ethiopia Journal of economics ,volume XVI ,No 1
Getu,B .(2011). National Coffee Research Project Coordinator and Researcher 8th Eastern African Fine Coffees Association Conference and Exhibition19th FebruaryArusha,TanzaniaNgurdotoLodge,VictoriaHallhttp://www.ico.org/sus nitasp#sthash.GjvzSjvy.dpuf
GGZADR,( 2014/15). GamoGoffa Zone Agricultural Department Annual Report For year 2014/15, Arba Minch. Glenn, D & Israel, K. (2013). Sampling the Evidence of Extension Program Impact. Program Evaluation and Organizational Development, IFAS, University of Florida. PEOD-5.
Govereh, J., & Jayne, T (2003). Cash cropping and food crop productivity: synergies or tradeoffs? Agricultural Economics, 28(1), 39-50.
Greene WH (2003). Econometrics analysis. 5th Edition. Macmillan, NewYork, USA. Ground, M. and Koch, S. F. (2008). Hurdle models of alcohol and tobacco expenditure in South households. South African Journal of Economics, 76(1):132–143.
Gujarati, D.(2004). Basic Econometrics (4th ed). The McGraw Hill Companies. New York.(germ thesis)
Healy, F. J. (1984). Statistics: A Tool for Social Research. Wads Worth Publishing Company, California. Henderson, J. Baylis, K., and Barton, J., (2010). Conversion from Staples to Cash Crop Production in Mexico After NAFTA: Effects of PROCAMPO and Credit Constraints. Denver, Colorado, July 25-27, 2010 Immink, MDC and Alarcon, JA (1993). Household income, food availability, and commercial crop production by smallholder’s farmers in the western highlands of Guatemala. Economic Development and Cultural Change 41:319–342.
Israel Glenn D. (2013). Determining Sample Size. Agricultural Education and Communication Department, Unversity of Florida /IFAS Extension.
61
Jayne,T. (1994). Do High Food Marketing Costs Constrain Cash Crop Production? Evidence from Zimbabwe. Economic Development and Cultural Change, 42(2), 387– 402
Jones Govereh, T.S. Jayne, and James,N. (1999). Smallholder commercialization, interlinked markets and food crop productivity: cross-country evidence in eastern and southern Africa
Kendra ,G, (2009). “International Sustainability Standards in Agriculture: The Case of Eco- labeled CoffeeKentucky
Kifle,B. , Demelash ,T. and Gabisa,G. (2015). Screening of somecoffee arabicagenotypes against coffee wilt diseases (gibberellaxylarioidesheim and saccus) at jimma, southwest Ethiopia.
Kydd, J., Dorward, A., Morrison, J. and Cadisch, G. (2004).Agricultural development and pro-poor economic growth in sub-Saharan Africa: potential and policy, Oxford Development Studies, 32: 1,37 — 57
Lerman, Zvi, (2004). Personal report: Policies and institutions for commercialization of subsistence farms in transition countries, Journal of Asian Economics 15: pp 461 479
Lukanu et al.(2004). Personal report: Policies and institutions for commercialization of subsistence farms in transition countries, Journal of Asian Economics 15: pp 461– 479
Marcia ,S. (2006). An Economic Analysis of Smallholder Coffee Production in Guatemala, Honduras, Nicaragua and Vietnam. Master’s thesis. Faculty of Purdue University.
McMillan, M., Assefa,T., Yohannes,A., Kibre ,M. and Amdissa,T. (2003). ETHIOPIA: Trade and Transformation Challenges. Agriculture and Trade Diagnostic Trade Integration Study. Annex 8, Volume 2, Addis Ababa, Ethiopia.)
Md. AtaulGaniOsmani, Md. Khairul Islam, Bikash Chandra Ghosh, Md. Elias Hossain(2014). Commercialization of Smallholder Farmers and Its Welfare Outcomes: Evidence from Durgapur UpazilaofRajshahiDistrict, Bangladesh. Journal of World Economic Research.Vol. 3, No. 6, 2014, pp. 119-126. doi: 62
10.11648/j.jwer.20140306.16Montclair State University Upper Montclair, New Jersey 07043.
Meneyahel, Z. (2015). Farmers Willingness to pay and choice of smallholder water lifting technologies: evidence from Adamitulu, Lemo and Dangelaweredas, Ethiopia.MSc thesis. ArbaMinch University
Ministry of Agriculture and Rural Development (MOARD) (2010). Annual report. Addis Abeba.
Ministry of Finance and Economic Development (MoFED-2010); Growth and
Transformation Plan (GTP), volume I and Volume II, Addis Ababa
MOFED (2006). A Plan for Accelerated and Sustained Development to End Poverty (PASDEP) (2005/06-2009/10). Volume I: Main Text. Addis Ababa. Ministry of Finance and Economic Development.
Newman, C., M. Henchion, and A. Matthews (2003) A double-hurdle model of Irish household expenditure on prepared meals, Applied Economics, 35, 1053-1061. OWAO.( 2015). OydaWoreda Agricultural and Development Office. Annual Report For year 2014/15, Shefit. PAERT (2008). Analysis of coffee supply, production, utilization and marketing issues and challenges in Ethiopia,Addis Ababa Pindyck,R. S.,and D. L. Rubinfeld. 1998. Econometric models and economic forecasts. 4th ed.New York: McGraw-Hill.(germ thesis
Poulton, C., Kydd, J., &Dorward, A. (2006).Overcoming Market Constraints on Pro-Poor Agricultural Growth in Sub-Saharan Africa. Development Policy Review, 24(November2004), 243-277.
Poulton, C., R. Al-Hassan, G. Cadish, C. Reddy and L. Smith (2001) "The Cash Crop versus Food Crop Debate" Crop Post Harvest Program, Issue Paper 3.
R.Humpheys, B. (2010); Dealing With Zeros in Economic Data, December 14, 2010 available athttp://www.ualberta.ca/~bhumphre/class/zeros_v1.pdf(theory concept)
Salami, Adeleke; B. Kamara, Abdul; and Brixiova, Zuzana (2010), “Smallholder Agriculture in East Africa: Trends, Constraints and Opportunities”, African Development Bank Group,
Samia,Z. (2000). Policy Options for Ethiopia’s Coffee Exports. School of Business.
63
Samuel, G. and Eva, L. (2008). Agricultural Commercialization in Coffee Growing Areas of Ethiopia
Sanchez, V. (2005): The Determinants of Rural Non-Farm Employment and Incomes in Bolivia. Master’s thesis. Michigan State University.
Schneider, K. and K.Gugerty, Mary (2010).The Impact of Export-Driven Cash Crops OnSmallholder Households, Evans School of Public Affairs University of Washington;
Sorsa, D. (2009). Sesame trade arrangements, costs and risks in Ethiopia: A baseline survey. VC4PD Research Papers.
Taye ,K.( 2013). Status of Arabica coffee Germplasm in Ethiopia center director & Senior Coffee Researcher, EIAR/Jimma Research Center E-mail: [email protected]
Teshome.K. (2009).Determinants of coffee export equation in Ethiopia. Africa Economic Research consortium, Nairobi, November 2009 Texas A&M University Themselves Out of Subsistence? World Bank Policy Research Working Paper 3881.
Tora,Ba. (2009). Fair-trade coffee and development a field study in Ethiopia national konomiska institutional vidlunds university Department of Economics at the University of Lund Minor Field Study Series No. 188, S-220 07 Sweden.
USAID (2010). Ethiopian Coffee Industry Value Chain Analysis Profiling the Actors, Their Interactions Costs Constraints and Opportunities Chemonics International
Verbeek M., (2004). A Guide to Modern Econometrics, 2nd ed., Erasmus University Rotterdam, John Wily and Sons Ltd., England.
VonBraun, J and Kennedy, E. (1994). Agricultural Commercialization, Economic Development and Nutrition.Johns Hopkins University Press, Baltimore, Maryland.Working Paper Series No. 105, April 2010
Wooldridge, J.W. (2002); Econometric Analysis of Cross Section and Panel Data, Cambridge,Massachusetts: MIT Press.
64
APPENDICES
APPENDIX1. Survey Questionnaire
Prepared by: TenkirTenkaMamo, Arba Minch University – Field of Study Economics (Specialization in Economic Policy Analysis) Purpose: This questionnaire is prepared to collect data pertaining to determinants of farmers’ participation in coffee production and marketing (The case of Oydaworeda in GammoGoffa Zone Southern Nations Nationalities and peoples Regional State). It will provide a major input for my master’s thesis and it is purely conducted for academic purposes.Therefore, the respondent is kindly requested to provide his/her valid responses to the sets of questions included in the questionnaires. All your responses remain confidential. I thank you in advance for your cooperation. Woreda ______Kebele______Date of interview ______A. Household Head Demographic Characteristics
1. Sex: Male Female 2. Age (in years) ______3. Educational level of household head (in years of schooling) ______Better to categorize level of education as: No formal education/no schooling Primary education (1 -8grades) Secondary education (9-12 grades) Tertiary education (colleges diploma & university degree) 4. Number of total family members ______5. Number of active household members aged between 15 and 64 years fulltime on farm activity ______
Year Age 0-15 15-64 65 and above Number of active family members Number of non-active family members 6. Is your family labour adequate for farm activities?
65
Yes No 7. Total amount of hired labor for the production year (2014/15) ______8. Total land holding size (in hectare) ______9. Land size suitable for coffee production ______(in hectare) 10. Did you involve in land renting activity in 2014/2015 production year? Yes No 11. If your answer to question #10 is “Yes”, are you: 1 = Rented out 2 = Rented in Rented out Rented in
B. Source of Household Income 1. From where did you get income you used to cover all family expenditures? Crop sales Livestock sales Remittances Credit Labour sale Others, please specify (if any)______2. Would you rank your income sources from major to minor (use the abovecode):1st=______2nd =______3rd =______4th = ______5th = ______3. Would you list the major 5 crops you grow currently?
Type of crop Area cultivated Quantity Quantity Price per Value sold (ha) produced(qntl) sold(quintal) quintal (in Birr) 1 2 3 4 5
66
4. Livestock ownership
Livestock Cows Oxen Donkey Mules Sheep Goats Poultry Heifers No Owned Have you 1= 1= 1= Yes 1= 1= 1= 1= Yes 1= Yes planned to Yes Yes 2=No Yes Yes Yes 2=No 2=No sell 2=No 2=No 2=No 2=No 2=No in 2007 E.C?
5. If you get income from sale of crop productions, which crop type you used to sell in the market most of the time? 1= food crops 2 = cereals 3 = vegetables 4= cash crops 5=fruits 6. Would you rank these crops according to primary crop income sources from major to minor (use the above code) 1st=______2nd =______3rd =______4th = ______5th = ______7. What are the major crops produced for market (cash crops) you grow in your area? 1 =______2= ______3= ______4=______5= ______8. Would you list these according to your level of production participation? 1st=______2nd =______3rd =______4th = ______5th = ______9. Are you a member of any rural cooperatives? Yes No 10. Do you have access to credit/loan? Yes No 11. Do you participate in non-farm income generating activities? Yes No 12. Do you produce sufficient food for your family for the whole year? Yes No 13. Traveling distance from home settlement to extension services ______(in meter) 14. Traveling time from home to nearby markets ______(in minutes) 15. Did you receive advisory services on coffee production? Yes No 16. Did you participate in production of coffee in any year of the last two crop seasons (2005/2006 or 2006/2007 E.C. Crop seasons): Yes 17. What direction had the farm gate price of coffee shown in these two years? Increased decreased remain the same 18. Was there any coffee crop failure in any of these years? Yes No 19. If yes, what are the sources of such failures? (Multiple answers are possible) Coffee disease pest infestations long/short rain Other (specify______)
67
20. Did you participate in the production of coffee in 2006/2007 (E.C) cropping season? Yes No (If your answer to Q#20is “No”, skips to question #27) 21. Land size allocated for coffee in 2014/2015 cropping season______(in hectare) 22. Which means of land preparation methods you used for coffee production:- 1= own oxen 2 = rented oxen 3 = traditional instruments 23. Type of coffee (bean) seed used: Traditional improved 24. From where did you get the coffee (bean) seed? Own production Market cooperatives agricultural offices Other (specify______) 25. Did you use fertilizer for coffee production? Yes No 26. If your answer to question #25 is” No”, what is the reason? No need Not available No potential to purchase others (specify______) 27. If your answer to question #20 is “No”, what are the main reasons that limit you from production of coffee?
No Possible reasons 1=serious 2=Minor problem problem 1 Decreased productivity of coffee from year to Year
2 Lack of improved coffee seeds 3 Shortage of land 4 Fear of market related problems 5 Lack of awareness about its importance 6 Fear of food shortages
C. Marketing Aspects: 1. Quantity of coffee produced in 2007 E.C ______(in quintal) 2. Quantity of coffee marketed ______(in quintal) 3. Quantity of coffee consumed ______(in quintal) 4. Time of sale: Immediately after harvest after a month after two months After three months/later 5. How did you sale your coffee produce?
68
Directly to the purchaser/traders through brokers others 6. Where did you sell mostly your coffee produce? Local buyers (collectors) Cooperatives traders at primary market 7. From whom you get better price? Local collectors’ cooperatives Traders at primary market others (specify______) 8. Did you face difficulty in finding coffee buyers? Yes No 9. If your answer to question #8 is “Yes”, is it due to: inaccessibility of market low price offer lack of price information other 10. Who set your selling price? Yourself market Buyers negotiations other ______11. Did you know the nearby market price before you transport to your coffee to market? Yes No 12. What is the price of coffee per kilogram in your local? ______13. What is the price of coffee per Kilogram at nearby market? ______14. Do you have a transport access to the nearest market? Yes No 15. How did you transport your coffee produce from home to market places? Head/back loading pack animals Vehicles other ______16. Do you have access to market information? Yes No
17. From where did you get market information? Local traders neighbor cooperatives media other______18. What are the major costs you incur in selling your coffee? 1. Transportation cost______(birr per quintal) 2. Packaging Cost______(birr per quintal) 3. Costs while waiting at the market ______(birr per quintal) 4. Others ______(birr per quintal) 19. What is the amount of total income you earned from coffee produce? 1. 2006 E.C______2. 2007 E.C ______20. What is the farm gate price of coffee per kilogram last year-2007 E.C? ______(in birr)
69
21. Did you considered this price when you decide to produce coffee in 2006/2007 E.C crop season? Yes No 22. What is your prediction about the coming year coffee price? Increase Decrease remain constant no idea 23. If you have any comment please list here: ______
Key Informant Interview with Agriculture development experts Prepared by: TenkirTenka Mamo, Arba Minch University- Post Graduate School Purpose: This questionnaire is prepared to collect data pertaining to production and market participation of farmers in Oydaworeda. It will provide a major input for a master’s thesis research purely conducted for academic purpose. Therefore, the respondent is kindly requested to provide us his/her valid responses to the sets of questions included in the questionnaires. All your responses remain confidential. I thank you in advance for your cooperation. A. Personal background 1. What is your job responsibility? 2. How long have you served in this Woreda/Kebele and in what capacity? B. Production, Marketing, and Farm Characteristics 1. What is the primary means of livelihoods for the people in this Woreda/Kebele? 2. What are the main food and cash crops grown in this Woreda/Kebele and why? 3. What services and assistance do the farmers get from your office? 4. What efforts are done to integrate the farmers with the market? What are the challenges and opportunities at their disposal? 5. What are the major non-farm activities farmers in your Woreda/Kebele mainly engaged in? 6. How many hectare of land is potentially suitable for production of coffee in your Woreda/Kebele? 7. What portion of land is allocated for the production of coffee currently? 8. Who is the primary buyer of the commodity from the farmers? 9. Are there any marketing cooperatives in this Woreda/Kebele? 10. If so, is coffee product traded through these cooperatives?
70
Appendix 2: VIF test result for continuous explanatory variables (Level of participation)
Variables vif 1/vif OXEN 1.23 0.813226 FARMSIZ 1.22 0.818778 FAMLABOR 1.12 0.889855 EDCN 1.03 0.969631 Mean vif 1.15 Source: own survey result, 2016
Appendix 3: Contingency Coefficient test result for discrete regressors
SEX CREDIT EXPER
SEX 1.0000
CREDIT 0.1125 1.0000
EXPER -0.1352 0.0067 1.0000
Source: own survey result, 2016
Appendix 4Heteroskedasticity test
. hettest
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity
Ho: Constant variance
Variables: fitted values of quantityprodu
chi2(1) = 711.83
Prob> chi2 = 0.0000
Source: own survey, 2016
71
Appendix5 Functional Misspecification Test
Linktest
Source | SS df MS Number of obs = 144
------+------F( 2, 141) = 438.20
Model | 6886.74051 2 3443.37026 Prob> F = 0.0000
Residual | 1107.97255 141 7.85796137 R-squared = 0.8614
------+------Adj R-squared = 0.8594
Total | 7994.71307143 55.9070844 Root MSE = 2.8032
------quantitypr~u | Coef. Std. Err. t P>|t| [95% Conf. Interval]
------+------
_hat | 2.233597 .0908483 24.59 0.000 2.053997 2.413198
_hatsq | -.0275849 .0018323 -15.05 0.000 -.0312073 -.0239625
_cons | -10.51425 .9009604 -11.67 0.000 -12.29539 -8.733117
------
Source: own survey, 2016
Appendix 6: VIF test result for continuous regressors (Income Generation)
Variable VIF 1/VIF MKTPRIC 1.29 0.776203 TRVTIME 1.20 0.833434 QUANMRKt 1.09 0.919369 Mean 1.19 Source: own survey result, 2016
72
Appendix 7: Contingency coefficient test for discrete regressor variables (Income Generation)
Markinfo onemonth twomonth threem Broke coop Traders
Markinfo 1.0000
Onemonth -0.0556 1.0000
Twomonth 0.2094 -0.2722 1.0000
Threemonth 0.0173 -0.2148 -0.2534 1.0000
Broke 0.4002 -0.2671 0.2949 0.4534 1.0000
Coop -0.0877 -0.2568 -0.3029 -0.2390 -0.2578 1.0000
Traders -0.1240 0.2075 -0.0520 -0.0756 -0.2377 -0.4518 1.0000
Source: own survey result, 2016
Appendix 8: Test for comparison of Tobit with double hurdle model
Statistics Tobit Double hurdle Probit Truncated Wald X2 (7) 29.95 245.46 182.48 Prob> X2 0.0000 0.0000 0.0000 Log likelihood -415.63979 -13.252236 - 416.98384 Observation 144 214 144 X2 test Double Hurdle and Tobitᴦ =29.192 > X2(7)= 18.475 Source: own survey, 2016
*** Statically significant at 1% probability level
73
Appendix 9: Tobit regression Result
Variables Coefficient Std. Err. T P>|t| [95% Conf. Interval]
SEX(male) .5358594 1.187323 0.45 0.652 -1.81199 2.883709 EDCN -.021126 .1394749 -0.15 0.880 -.2969279 .254676 FAMLAB .5010695 .3405813 1.47 0.144 -.1724065 1.174546 OXEN .9231028 .471056 3.96 0.002*** -.0083781 1.854584 CREDIT -.5685116 .6687761 -0.85 0.397 -1.89097 .7539472 EXPRINCE 2.78782 1.531315 1.82 0.071 * -.2402494 5.815889 FARMSIZ 9.986201 4.391341 2.27 0.025 ** -1.302626 18.66978 CONS -.1808645 2.293922 -0.08 0.937 -4.716937 4.355208 SIGMA 4.612642 .9364128 2.76095 6.464334 Number of obs 144 Wald chi2 (7) 29.95 Prob> chi2 0.000 Log pseudo likelihood -415.63979 Pseudo R2 0.1469 Source: own survey, 2016
Appendix 10: VIF test result for continuous explanatory variables
Variable VIF 1/VIF
AgeHH 1.35 0.742054
Fam labor 1.92 0.521099
Farm size 2.50 0.400512
Oxen 1.98 0.504249
Dstextens 1.32 0.758722
Educ level 1.30 0.771408
Travmrkt 1.45 0.691600
Mean VIF 1.69 Source: own survey, 2016
74
Appendix 11: Contingency Coefficient test (for discrete explanatory variables)
SEX Credit Non farm Food Sex 1.0000 Credit -0.0641 1.0000 Nonfarm 0.2476 -0.4546 1.0000 Food -0.2677 0.4479 -0.6317 1.0000 Source: own survey, 2016
Appendix 12: Heteroskedasticity and link testes hettest
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity
Ho: Constant variance
Variables: fitted values of income2007
chi2(1) = 619.98
Prob> chi2 = 0.0000 linktest
Source SS df MS Number of obs = 144
------F( 2, 141) = 114.34
Model 6.0774e+11 2 3.0387e+11 Prob> F = 0.0000
Residual 3.7472e+11 141 2.6576e+09 R-squared = 0.6186
------Adj R-squared = 0.6132
Total 9.8246e+11 143 6.8704e+09 Root MSE = 51552
------
Income2007 | Coef. Std. Err. t P>|t| [95% Conf. Interval]
------
_hat -.6948225 .2082745 -3.34 0.001 -1.106567 -.2830781
_hatsq .0000106 1.20e-06 8.84 0.000 8.23e-06 .000013
_cons 31414.96 7124.281 4.41 0.000 17330.75 45499.18
------
Source: own survey, 2016 75