i? ? ANALYSIS OF BUTTER SUPPLY CHAIN: THE CASE OF ATSBI- WENBERTA AND ALAMATA WOREDAS, TIGRAY, ETHIOPIA M.Sc. Thesis BY EMBAYE KIDANU January 2010 Haramaya University ANALYSIS OF BUTTER SUPPLY CHAIN: THE CASE OF ATSBI- WENBERTA AND ALAMATA WOREDAS, TIGRAY, ETHIOPIA A Thesis Submitted to the Department of Agricultural Economics, School of Graduate Studies HARAMAYA UNIVERSITY In Partial Fulfillment of the Requirements for the Degree of MASTER OF SCIENCE IN AGRICULTURE (AGRICULTURAL ECONOMICS) BY Embaye Kidanu Kindeya January 2010 Haramaya University i? ? SCHOOL OF GRADUATE STUDIES HARAMAYA UNIVERSITY As Thesis Research advisors, we here by certify that we have read and evaluated this Thesis prepared, under our guidance, by Embaye Kidanu entitled ?Analysis of Butter Supply Chain: The Case of Atsbi-Wonberta and Alamata Woredas, Tigray, Ethiopia. We recommend that it be submitted as fulfilling the thesis requirement. ?Berhanu Gebremedhin (PhD) ______________ ___________ Major Advisor Signature Date Mr. Dirk Hoekstra _________________ ___________ Co-Advisor Signature Date As member of the Board of Examiners of the M.Sc Thesis Open Defense Examination, we certify that we have read, evaluated the Thesis prepared by Embaye Kidanu and examined the candidate. We recommended that the Thesis be accepted as fulfilling the Thesis requirement for the Degree of Master of Science in Agriculture (Agricultural economics). Belayneh Legesse (PhD) _______________ __________ Chairperson Signature Date Jema Hajji (PhD) _________________ _________ Internal Examiner Signature Date Kindie Getnet (PhD) _________________ _________ External Examiner Signature Date ii? ? DEDICATION To the smallholder dairy producer households of the country iii? ? STATEMENT OF THE AUTHOR First, I declare that this thesis is my bonafide work and that all sources of materials used for this thesis have been duly acknowledged. This thesis has been submitted in partial fulfillment of the requirements for an advanced M.Sc. degree at Haramaya University and is deposited at the University Library to be made available to borrowers under rules of the Library. I solemnly declare that this thesis is not submitted to any other institution anywhere for the award of any academic degree, diploma, or certificate. Brief quotations from this thesis are allowable without special permission provided that accurate acknowledgement of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the School of Graduate Studies when in his or her judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author. Name: Embaye Kidanu Kindeya Signature: ____________ Place: Haramaya University, Haramaya. Date of Submission: January 2010 iv? ? BIOGRAPHICAL SKETCH The author was born on march18, 1984 in Wukro town, eastern zone of Tigray region. He attended his elementary school at Dinglet Primary School and Secondary school at Agazi Comprehensive High School during the period between 1991 and 2002/3 G.C at Wukro. Joined the then Debub University in 2003/4, he graduated with B.Sc. degree in Agricultural Resource Economics and Management on 12 th of July 2006. Until he joined Haramaya University for postgraduate studies in Agricultural Economics in October 2007, he was working as socio-economic researcher for Tigray Agricultural Research Institute. ? ? v? ? ACKNOWLEDGEMENTS The only viable way of completing a project of any significant magnitude is to partly relinquish control to others. My thesis was not an exception to this rule. I, therefore, share the ownership of this work with several others and wish to acknowledge their contributions. I, primarily, desire to express my special appreciation and heartfelt gratitude for the guidance, support and inspiration of Dr. Berhanu Gebremedhin (scientist in agricultural economics), my major adviser. His critical comments starting from the research proposal till the final thesis script and his motive to develop my knowledge by providing me interesting materials obliged me to convey warmest pleasure for him. In spite of his tight schedule, he keeps his office open any time I needed his help and spending spare time effectively with me. Without him, this thesis might not have developed the way it did. In general, neither words nor complex mathematical equations can fully articulate his roles in the materialization of this effort. My equal special appreciation, without repeating similar words, also goes to my co-advisor Mr. Dirk Hoekstra, for his helpful and constructive comments on realization of the thesis. I am indebted to Dr. Gebremicheal Nigusse for his assistance in the development of the project proposal and concern throughout the study. I wish also to express my thanks to Dr. Gebremedhin for his helpful guidance during the study. I gratefully acknowledge Ato Gebregzibher Gebreyohans (TARI Director General) for valuable support during the design of the survey questionnaires. I would like to express my deepest gratitude to the International Livestock Research Institute (ILRI)/Improving Productivity and Market Success (IPMS) of Ethiopian farmers? project for granting me research fund and other different services during the research work. I wish to convey my heartfelt thanks to Dr. Gebreyohans Berhane, Dawit Weldemariam, Abrham Gebrehiwot, Kebede Manjur, Weldemariam, Teklay, Degol Fshaye, Efrem Gidey, Kiflom Meshesha, Berhan Abera, Simos Seyum, Zewdu Mesfin, Eyob Shenkut, Muluken Yohans, Wubshet Chala, Fasil Lencho, Amsalu Mitku, Abdulkerim Bedewi, Assefa and others for multidimensional support and continuous encouragement in my endeavor. vi? ? Furthermore, I would like to be grateful to Muluhiwot Getachew and Birke Enyew of IPMS for their responsive financial facilitation of my research work. I would like to extend my profound appreciation to butter producers and traders in Atsbi-Wonberta woreda, Alamata woreda and Mekelle city for their sympathy, hospitality and for they spent many hours to explaining their livelihoods. I also wish to convey my cordial thanks to everybody who has a hand on my accomplishment without mentioning each. Finally, I am grateful to extend my warmest pleasure to my father Ato Kidanu Kindeya, my mother, Weletegebreal Kidanemariam and all my brothers and sisters for their continuous encouragement during the study. Without their support, completion would have not been real. On top of all, thanks God for everything of merits is due to your benevolence. You made it possible for things to happen. ? vii? ? ABBREVIATIONS BMP Butter Market Participation BMS Butter Market Supply BoARD Bureau of Agriculture and Rural Development CSA Central Statistical Authority DA Development Agent DDA Dairy development agency DECSI Dedebit Credit and Saving Institution EARO Ethiopian Agricultural Research Organization ETB Ethiopian Birr FAO Food and Agriculture Organization FTC Farmers? Training Center GDP Gross Domestic Product GCC Global commodity chain Ha Hectare IFPRI International Food Policy Research Institute ILRI International Livestock Research Institute IPMS Improving productivity and Market success kms Kilometers m.a.s.l meters above sea level MEDAC Ministry of Economic Development and Cooperation MIS Market information system MoARD Ministry of Agriculture and Rural Development MOI Ministry of Information MT Metric Tone NGO Non-Governmental Organization NRM Natural resource management OLS Ordinary Least Squares PA Peasant Association viii? ? ABBREVIATIONS (CONTINUED) Qt Quintal REST Relief Society of Tigray RMA Rapid Market Appraisal Rs Indian Rupees S-C-P Structure Conduct Performance SPSS Statistical Package for Social Sciences SSA Sub Saharan Africa TAMPA Tigray Agricultural marketing Promotion Agency TARI Tigray Agricultural Research Institute TGMM Total Gross Marketing Margin TLU Tropical Livestock Unit VIF Variance Inflation Factor WET World economic triangle WoARD Woreda Office of Agriculture and Rural Development ix? ? TABLE OF CONTENTS STATEMENT OF THE AUTHOR iii BIOGRAPHICAL SKETCH iv ACKNOWLEDGEMENTS v ABBREVIATIONS vii LIST OF TABLES xiii LIST OF FIGURES xiv LIST OF APPENDIX TABLES xv ABSTRACT xvi 1. INTRODUCTION 1 1.1.Background 1 1.2. Statement of the Problem 2 1.3. Objectives of the Study 3 1.4. Research Questions 4 1.5. Scope and Limitations of the Study 4 1.6. Significance of the Study 4 1.7. Organization of the Study 5 2. LITERATURE REVIEW 6 2.1. Overview of Dairy Production Systems in Ethiopia 6 2.1.1. Dairy marketing systems in Ethiopia 8 2.1.2. Overview of butter production and marketing in Ethiopia 9 2.1.3. Common challenges of dairy production and marketing in Ethiopia 11 2.2. Methodological and Conceptual Frameworks 11 x? ? TABLE OF CONTENTS (CONTINUED) ? 2.2.1. Conceptual frame work of marketing 11 2.2.2. Methodological framework of marketing 14 2.2.2.1. Approaches to study marketing 14 2.2.2.2. Supply chain analysis and Factors affecting market supply 15 2.2.2.3. Market structure, conduct and performance analysis (S-C-P model) 16 2.2.2.4. Market participation 20 2.2.2.5. Value chain; definition and importance 21 3. METHODOLOGY 23 3.1. Description of the Study Areas 23 3.2. Data Type, Sources, and Methods of Data Collection 27 3.2.1. Data type and sources 27 3.2.2. Methods of data collection 28 3.3. Sample Size and Procedure 28 3.3.1. Farmers sampling 28 3.3.2. Wholesalers and assemblers sampling 29 3.3.3. Retailers sampling 29 3.4. Methods of Data Analysis 29 3.4.1. Structure of production costs and profitability of butter production 30 3.4.2. Analysis of butter market supply 31 3.4.3. Hypothesis and definition of variables 33 3.4.4. Analysis of production and marketing supporting services 37 3.4.4.1. Input supply 37 3.4.4.2. Access to formal credit 42 3.4.5. Analysis of structure -conduct -performance 46 3.4.5.1. Structure of market 46 3.4.5.2. Market conduct 47 xi? ? TABLE OF CONTENTS (CONTINUED) 3.4.5.3. Market performance 48 3.4.6. Major production and marketing problems 49 4. RESULTS AND DISCUSSION 50 4.1. Butter Production Characteristics and Households Socio-Demographic Settings 50 4.1.1. Butter production characteristics 50 4.1.2. Socio-demographic characteristics of butter producer households 52 4.2. Butter Production and Marketing Support Services 56 4.2.1. Input supply 56 4.2.1.1. Supply of crossbreed cows and factors determining households? access to crossbreed cows 57 4.2.1.2. Feed supply and factors affecting households? access to feed 61 4.2.1.3. Equipments supply 65 4.2.1.4. Access to veterinary service 67 4.2.1.5. Access to drug 68 4.2.2. Credit supply and households? credit constraint condition 71 4.2.3. Extension service 75 4.2.4. Access to market information 76 4.3. Structure of Production Costs and Profitability of Butter Production 78 4.3.1. Structure of butter production costs 78 4.3.2. Returns of butter production 80 4.4. Butter Marketing and Butter Traders? Characteristics 82 4.4.1. Butter traders? social and intellectual capital 82 4.5. Structure, Conduct and Performance of Butter Markets 86 4.5.1 Structure of butter markets 87 4.5.1.1. Market concentration 87 4.5.1.2. Barriers to entry 88 xii? ? TABLE OF CONTENTS (CONTINUED) 4.5.1.2.1. Licensing 88 4.5.1.2.2. Lack of capital 88 4.5.1.2.3. Experience in butter trading 89 4.5.1.2.4. Risk barriers 89 4.5.2 Market conduct 90 4.5.2.1 Existence of formal and informal producer and marketing groups 90 4.5.2.2. Market price information behavior 90 4.5.2.3. Buying strategy 91 4.5.2.3. Selling strategy 92 4.5.3. Performance of butter market 92 4.5.3.1. Marketing margin 95 4.6. Analysis of Factors Determining Butter Market Supply 97 4.6.1. Determinants of butter market participation decision 98 4.6.2. Level of butter market participation 100 4.7. Production and Marketing Opportunities and Constraints 102 4.7.1. Butter production and marketing constraints 103 4.7.2 Marketing constraints of butter traders 106 4.7.3. Butter production and marketing opportunities 107 4.8. Identification of Important Constraints and their Possible Intervention Points throughout the Entire Butter Supply Chain 108 5. SUMMARY, CONCLUSION AND POLICY IMPLICATIONS 112 5.1. Summary 112 5.2. Conclusion and Policy Implications 116 REFERENCES 121 APPENDICES 129 xiii? ? LIST OF TABLES Table Page 1. Butter production characteristics .............................................................................. 51 2. Socio-economic and demographic characteristics of butter producer households ...... 53 3. Socio-demographic characteristics of market participants and non-participants ........ 55 4. Probit estimates of households? access to crossbreed cows ...................................... 59 5. Source of cross breed cows ...................................................................................... 60 6. Evaluation of crossbreed cows .................................................................................. 61 7. Probit estimates of factors determining households? access to feed .......................... 63 8. Feed source and evaluation of feed supply ................................................................ 64 9. Source and evaluation of butter equipments by household characteristics................. 66 10. Access to and evaluation of veterinary service ........................................................ 67 11. Sources, access and evaluation of drug supply ........................................................ 68 12. Results of probit model for access to drug .............................................................. 70 13. Results of probit model for farmers credit constraint condition ............................... 72 14. Access to extension service and frequency of da contact per month ........................ 76 15. Butter market information sources ........................................................................... 77 16. Structure of butter production costs and returns ....................................................... 81 17. Socio-demographic characteristics of butter traders by market center ...................... 83 18. General characteristics and intellectual, social and financial endowment of butter traders ..................................................................................................................... 85 19. Concentration ratio of four firms for the three markets ............................................ 88 20. Years of experience in butter trading ....................................................................... 89 21. Traders market information source .......................................................................... 91 22. Butter market channels and marketing margin analysis. ........................................... 96 23. Probit results of butter market participation decision (first stage) ............................ 98 24. Ols estimates of level of butter market participation .............................................. 102 25. Production and marketing constraints of butter producer households ..................... 104 26. Marketing constraints of butter traders .................................................................. 106 xiv? ? LIST OF FIGURES Figure Page 1. Location of the study area, Alamata woreda ............................................................. 25 2: Location of study Area, Atsbi-wenberta woreda ........................................................ 26 3. Butter market channels .............................................................................................. 94 4. Diagrammatic representation of production and marketing constraints and corresponding strategic intervention points in the entire butter supply chain ........ 111? ? xv? ? LIST OF APPENDIX TABLES Appendix table Page 1. Cow population in atsbi- wenberta and alamata woredas ........................................ 130 2. Conversion factors used to compute tropical livestock units. .................................. 130 3. Conversion factors used to compute man-days equivalent ...................................... 131 4. Test for multicollinearity ........................................................................................ 131 5. Contingency coefficient .......................................................................................... 132 6. Commitment level of export subsidies on milk products (Rs. /kg) .......................... 132 xvi? ? ANALYSIS OF BUTTER SUPPLY CHAIN: THE CASE OF ATSBI-WEMBERTA AND ALAMATA WOREDAS, TIGRAY, ETHIOPIA ABSTRACT Butter is an important source of food, cosmetics and common marketable form of dairy product in the study areas. The total butter production in the survey year was 18880 kg in both the woredas, out of which 87.6% was supplied to market. The objective of this study is to investigate the butter supply chain in Atsbi-wonberta and Alamata woredas of Tigray region. The specific objectives were to analyze production and marketing supporting services, to identify structure of production costs and determine profitability, to analyze determinants of butter supply, to analyze structure-conduct and performance of butter market and to identify major production and marketing constraints and opportunities. To accomplish these tasks? formal and informal data collection tools of both primary and secondary data were used. a survey was conducted in the districts in which data was collected from randomly selected 200 butter producer households of Atsbi and Alamata woredas using a structured questionnaire and from 56 butter traders at different level of the chain from Alamata, Atsbi and the terminal market, Mekelle. The econometric result of market participation decision indicated quantity produced, number of extension visit, market information access, family size, distance to nearest market and distance to development center are the significant factors. Similarly, quantity produced, distance to nearest market and distance to development center are significant factors affecting level of supply. As to the probit model of determinants of access to crossbreed cows; number of extension visit, participation in extension demonstration, access to veterinary service and distance to the woreda town are significant factors. Likewise, dairying experience, labor availability, distance to feed market and access to formal credit are significant factors determining farmers? access to feed in the probit model. Among the hypothesized determinants of access to cattle drug; number of extension visit, number of cows and distance to the nearest market found to be significant factors. The probit model of households? credit constraint condition reveals herd size, distance to development center, off farm income and frequency of extension contact are the significant factors associated with credit constraint condition. The net returns, generated after deducting all economic cost of resources used for butter production, found to be 1623ETB per cow/year xvii? ? from crossbreed cows and 213ETB per cow/year from local breed cows with overall average net return of 918.3ETB per cow annually. Following the market structure criteria of concentration ratio, butter market shows competitive nature in Atsbi market with C 4 of 31%, and weak oligopolistic nature in Alamata and Mekelle markets as their concentration ratios are 39% and 44% respectively. Entry barriers were not observed in relation to licensing and working capital. However, considerable burrier was observed with respect to years of experience at the wholesale level and risk of seasonal variation in demand and price of butter. Based on the channel comparison and marketing margin analysis of butter market performance, the producer?s share of the consumer?s price was found to be the highest along producer-consumer channel followed by producer-woreda retailer-consumer and producer- rural assembler-wholesaler-consumer. The major constraints of butter production and marketing in the studied areas were inadequate availability and supply of feed, the low productivity of the endogenous cattle breeds, low supply of crossbreed cows, lower demand of butter during fasting periods, adulteration and seasonal fluctuation of butter price. Therefore, taking into account these factors in designing butter production and marketing improvement programme may help policy makers come up with policies aimed at ameliorating the butter supply chain in the districts. 1. INTRODUCTION 1.1. Background In the realm of economic growth, markets may provide the incentives to profit maximizing participants to develop new technologies, products, resources of supply, new markets and methods of exploiting them. Agricultural marketing acts as an agent of rural development. Moreover, agricultural marketing will play a coordinating rule, steering supply and demand with respect to place, time and form utilities. If the production system works efficiently, it produces suitable incentives to meet consumer?s needs more accurately in terms of type, quality and quantity of supply. Production is thus adapted to the need of consumers in response to price signals transmitted by the marketing system (Vincent, 1967). Agricculture in Ethiopia, depending on agro-climatic conditions in the respective areas, is characterized by smallholder and traditional cultivation. Agricultural production is based on obsolete practices and low utilization of productivity-enhancing inputs. Consequently, the quantity of surplus supplied to the market is very limited. Some studies indicate that even in a bumper harvest year, the marketable surplus is no more than 28% on the average. The situation is equally bleak in the livestock sector. Despite the huge livestock population, the low off-take rate and the poor quality of the livestock itself implies that the marketed surplus of livestock products (meat, milk, butter, eggs, etc) is very low (Dawit, 2005). Similarly, Ethiopia ranked first in cattle population in Africa but the dairy industry is not developed even as compared to east African countries like Kenya, Uganda, and Tanzania. Regarding dairy production, the national milk production remains among the lowest in the world, even by African standards (Zegeye, 2003). As the current development in the country is characterized by rapid population growth in general, the demand for dairy products is increasing as ever. However, the current levels of contributions of the livestock subsector in Ethiopia, at either the macro or micro level is below potential. The levels of foreign exchange earnings from livestock and livestock products are much lower than would be expected, given the size of the livestock population (Berhanu et al., 2006). 2? ? In Ethiopia, dairy policy does not favor the producer farmer, no subsidy and no support of price floor is given. Therefore, dairy prices reflect the cost of producing and distributing them. However, many developed countries have established the intervention price of butter. When price intervention was common for butter produced in the European Union. The wholesale price of packed butter was close to the intervention floor and the retail price a few percentage points higher: as low as 5 percent according to the Milk Marketing Board (1987), which argued that butter was keenly priced by supermarkets in order to attract customers into stores. Rural producers around Adisababa sells about 36% of total sales to restaurants, while those to itinerant traders accounted for 33% and sales to individuals and those to wholesalers in Addis Ababa accounted for 31% of sales (Debrah, 1990). Similarly, butter production is the long-life activity in Atsbi-Wonberta and Alamata woredas and due to some cultural barriers of milk transaction; it is also the most common marketable dairy product. Different marketing agents like rural assemblers, retailers, wholesalers and hotels and restaurants are participating along the chain of butter supply. 1.2. Statement of the Problem Dairy production, among the sectors of livestock production system, is a crucial issue in Ethiopia where livestock and its products are important source of food and income, and dairying has not been fully exploited and promoted in the country. Despite its huge numbers, the livestock subsector in Ethiopia is low in production in general, and compared to its potential, the direct contribution it makes to the national economy is limited. A number of fundamental constraints underlie these outcomes, including traditional technologies, limited supply of inputs (feed, breeding stock, artificial insemination and water), poor or non-existent extension service, high disease prevalence, poor marketing infrastructure, lack of marketing support services and market information, limited credit services, absence of effective producers? organizations at the grass roots levels, and natural resources degradation (Berhanu et al., 2006). In addition, policy decision on milk and milk product marketing are taken in the absence of vital information on how they affect dairy producers, traders, exporters, and consumers. Similarly, current knowledge on dairy product market structure, performance and prices is poor for designing policies and institutions to overcome the perceived problems in the marketing system (Ayele et al., 2003). 3? ? However, the fundamental development objective of the country is to establish a strong market-oriented agricultural production system. Likewise, the Poverty Reduction Strategy outlines, rapid economic development that benefits all citizens can only be achieved when the market-led transformation of the agricultural sector of the economy could enact. Thus, it is necessary to survey the present problems of production and marketing, formulate recommendations, devise, and implement a strategy, which enables the establishment of an efficient production system and transparent and efficient marketing system of these products. Moreover, as milk transaction is culturally prohibited and considered as a taboo in the rural areas of the districts, butter is found to be a sole marketable commodity of dairy products. Furthermore, butter is an important cash source for household consumption expenditure in the woredas. Based on this ground, for progressive development of the dairy sector, then households? income generation and transformation the small-scale and subsistence producers to commercial operators, investigation of butter supply chain needs to be carried out, as there was not done such research in this area. Therefore, in line with the market-oriented production strategy of the country?s policy, the study is intended at bridging the information gap with regard to butter production and marketing in Atsbi-Wonberta and Alamata districts of Tigray region. 1.3. Objectives of the Study The overall objective of this study is to examine butter supply chain in Atsbi and Alamata woredas. The specific objectives of the study are: 1. to analyze the production and marketing support services such as extension, input supply, credit and marketing 2. to analyze the structure of production costs and determine profitability of butter production 3. to analyze the determinants of butter market supply 4. to analyze the market structure, conduct and performance of butter marketing 5. to identify major constraints and opportunities of production and supply 4? ? 1.4. Research Questions ? The thesis attempted to answer the following research questions: 1. Which factors determine butter market supply in Atsbi and Alamata districts? 2. How is butter marketing system organized, functioning and performing? 3. What is the structure of butter production costs and how profitable is butter production for producers? 4. What are the key constraints and opportunities in butter production and supply? 5. What are the production and marketing support services and to what extent they serve to producers and different marketing actors? 1.5. Scope and Limitations of the Study As the study is being the first in the region, it lacks many detailed investigations, which could have reinforced understanding of the whole system especially in relation to demand side and consumption preference studies. Hence, due to time and financial constraints, the study narrowed down to concentrate on butter-supply chain in Atsbi-Wonberta and Alamata woredas as well as final market of the product, Mekelle. In addition to this, butter production accounts for the major proportion of dairy production in the areas and passed through a number of marketing stages and agents. Furthermore, other dairy products are not included because transaction of these products is culturally prohibited in the rural areas of the studied woredas. As a result, special attention was given to butter of the dairy products. To this effect, butter market channels, chain actors and their roles, institutions involved directly or indirectly in butter production and marketing as well as factors affecting marketable supply of butter has been discussed and identified. 1.6. Significance of the Study Various efforts to promote small-scale farming have been noted in the past decades. Farmers in these areas are not really part of commercial agriculture. This is one of the reasons why contribution of smallholder agriculture to the gross national product is still not realized as what would be expected. This kind of subsistence farming is characterized by low production (and productivity), poor access to markets, and poor access to inputs and credit. 5? ? It is, however, possible for smallholder farming to survive economically when given a set of opportunities. After all, subsistence farmers are used to take rational decisions in order to adapt to conditions they find themselves in. Since such a study was not done before in the region, many stakeholders could use the output of this finding for development of the subsistent farmers. Hence, the critical analysis of butter supply chain is very important before suggesting for production and marketing development issues. Therefore, the study gives detail information on how butter production and marketing is functioning particularly in the specified districts, which are the potential butter producing areas in the region. Mainly, the results of the study favor small-scale dairy farmers and the actors of the supply chain. Analysis of the whole system and identifying clearly the challenges will benefit policy makers and implementers in indicating the area of advantage for what should be done to improve butter production and marketing. Besides, it would be a useful reference for researchers and other interested persons in the area of study. Therefore, it was hoped that, results from this study would have practical use mainly to this area and can serve as a base for any further studies to be conducted in other similar areas within this line of study. 1.7. Organization of the Study Firstly an over view of the unique characteristics of Ethiopian dairy production and marketing was discussed. A review of the selected conceptual and methodological framework is dealt in chapter two. Chapter 3 introduces background information about the study area and verifies the methods of data collection and data analyses, followed by chapter four that presents the results of the study. Finally, chapter five offers a brief summary, conclusion and policy implication of the empirical findings. 6? ? 2. LITERATURE REVIEW ?2.1. Overview of Dairy Production Systems in Ethiopia In the late 1980s, agriculture in Ethiopia contributed about 45% of national gross domestic product. The livestock sector contributed about 40% of agricultural GDP or 18% national GDP, and 30% of agricultural employment. Dairy output accounted for about half of the livestock output (Feleke and Geda, 2001). During 1993-2001, per capita income remained at about US$100. Livestock production increased by much less than the production increase for the agriculture sector as a whole, so relative share of livestock to AGDP declined over time. During this period, per capita livestock, output fell by 5% while crop, food and agriculture grew at 14, 7 and 6% respectively (Halderman, 2004). Four main dairy production systems may be identified in the country: a small commercial sector consisting of large private farms and state farms. Small urban/peri-urban systems raising crossbred or both crossbred and local cattle having access to milk collection centers or cooperatives. Smallholder mixed farming systems in the highlands using indigenous breeds, and pastoral/agro- pastoral system in the low lands. Reliable figures on the relative importance of these systems in terms of number of farms/herds, dairy population, share of milk produced are not available. However, a rough estimate indicate that currently, out of about 1430 million liters of milk produced annually, 900 million liters (63.3%) is produced by rural small-scale mixed farms in the highlands. 205 million liters (14.3%) by small urban/peri-urban farms in the highlands, 320 million liters (22.4% ) by pastoral/agro- pastoral producers in the lowlands, and 5 million liters (less than 0.03%) by large private and state farms (Ahmed et al, 2003; Feleke and Geda, 2001). Household consumption and expenditure surveys indicate that livestock products comprise only 8% of total food expenditure, with 4 percent of expenditure allocated to dairy products. About 56% of milk in the country is processed into butter, cheese and yoghurt and 44% is consumed fresh. Although levels of consumption vary according to income levels, relative shares of fluid milk and other products mainly butter remain about the same across income groups 7? ? The rural system is non-market oriented and most of the milk produced in this system is retained for home consumption. The level of milk surplus is determined by the demand for milk by the household and its neighbors?, the potential to produce milk in terms of herd size and production season, and access to a nearby market. The surplus is mainly processed using traditional technologies and the processed milk products such as butter, ghee, ayib and sour milk are usually marketed through the informal market after the households satisfy their needs (Tsehay, 2001). Pastoralists raise about 30% of the indigenous livestock population, which serve as the major milk production system for an estimated 10% of the country?s human population living in the lowland areas. Milk production in this system is characterized by low yield and seasonal availability (Zegeye, 2003). The highland smallholder milk production is found in the central part of Ethiopia where dairying is nearly always part of the subsistence, smallholder mixed crop and livestock farming. Local animals raised in this system generally have low performance with average age at first calving of 53 months, average calving intervals of 25 months and average lactation yield of 524 liters (Zegeye, 2003). Peri-urban milk production is developed in areas where the population density is highland agricultural land is shrinking due to urbanization around big cities like Addis Ababa. It possesses animal types ranging from 50% crosses to high grade Friesian in small to medium- sized farms. The peri-urban milk system includes smallholder and commercial dairy farmers in the proximity of Addis Ababa and other regional towns. This sector owns most of the country?s improved dairy stock (Tsehay, 2001). The main source of feed is both home produced or purchased hay; and the primary objective is to get additional cash income from milk sale. This production system is now expanding in the highlands among mixed crop? livestock farmers, such as those found in Selale and Holetta, and serves as the major milk supplier to the urban market (Gebrewold et al., 2000). Urban dairy farming is a system involving highly specialized, state or businessmen owned farms, which are mainly concentrated in major cities of the country. They have no access to grazing land. Currently, a number of smallholder and commercial dairy farms are emerging mainly in the urban and peri-urban areas of the capital and most regional towns and districts (Nigussie, 2006). Smallholder rural dairy farms are also increasing in number in areas where there is market access. According to Azage and Alemu (1998), the urban milk system in 8? ? Addis Ababa consists of 5167 small, medium and large dairy farms producing 34.65 million litres of milk annually. Of the total urban milk production, 73% is sold, 10% is left for household consumption, 9.4% goes to calves and 7.6% is processed into butter and ayib (cheese). In terms of marketing, 71% of the producers sell milk directly to consumers (Tsehay, 2001). 2.1.1. Dairy marketing systems in Ethiopia In the African context, markets for agricultural products would normally refer to market- places (open spaces where commodities are traded). Conceptually, however, a market can be visualized as a process in which ownership of goods is transferred from sellers to buyers who may be final consumers or intermediaries. Therefore, markets involve sales, locations, sellers, buyers and transactions (Debrah and Berhanu, 1991). There are formal and informal marketing systems in the transaction of dairy products. The term ?informal? is often used to describe marketing systems in which governments do not intervene substantially in marketing. Such marketing systems are also referred to as parallel markets. The term ?formal? is thus used to describe government (official) marketing systems (Debrah, 1990). Dependable system has not been developed to market milk and milk products in Ethiopia (Zegeye, 2003). Fresh milk is distributed through the informal and formal marketing systems. In both rural and urban parts of the country, milk is distributed from producers through the informal (traditional) means. This informal market involves direct delivery of fresh milk by producers to consumers in the immediate neighborhoods or to any interested individuals in nearby towns (Debrah and Berhanu, 1991). Initial intervention to promote formal dairy marketing started with the establishment of a 300 dairy farm and a small milk processing plant under the UN Relief and Rehabilitation Program in 1947 in the premises of the now Dairy Development Enterprise (DDE) (Sintayehu, 2003). The same report stated that in 1959 UNICEF helped establish a processing plant with a processing capacity of 10 thousand litres per day with milk collection and purchasing centres around Addis Ababa. The radius of milk collection was later expanded to 70 km around the capital. Capacity of the processing plant was increased to 30 thousand litres in 1969. In 1979 the DDA (Dairy Development Agency) was transformed to the DDE when processing 9? ? capacity was increased to 60 thousand litres/day and the radius of collection expanded to 150 km with donor assistance. The only organized and formal milk marketing and distribution system comes from the two milk-processing plants which are both located in the capital Addis Ababa (Zegeye, 2003). As reported by many authors, farmers? milk marketing groups and dairy cooperatives play a key role for milk marketing outlets, which as a result encourages farmers to produce more (Zegeye, 2003). A study of the milk marketing system in Kenya has shown that there are at least eight different marketing channels, with the number of intermediaries ranging from 1 to 4 (FAO 1996). A study in Addis Ababa milk shed revealed that dairy producers sold milk through different principal market channels (Debrah 1990, Mbogoh, 1990), which included: ? Producer?consumer (P?C) channel: direct sales to individual consumers, which accounted for 71% of the total channels. ? Producer?catering institution?consumer (P?CI?C) channel: catering institutions includes/ itinerant traders, small private shops and kiosks, coffee and tea sales, hotels, and supermarkets/, and ? Producer?government institution?consumer (P?GI?C) channel: sales to government institutions such as the armed forces, schools and hospitals 2.1.2. Overview of butter production and marketing in Ethiopia Butter was known in the classical Mediterranean civilizations. The ancient Greeks and Romans seemed to have considered butter a food fit more for the northern barbarians. The potential butter producer nations in the world are, India (620,000 MT), United States (522,000MT), France (466,000MT), Germany (422,000MT), and New Zealand (307,000MT) respectively (Dalby, 2003). Butter produced from whole milk is estimated to have 65% fat and is the most widely consumed milk product in Ethiopia, of the total milk produced, around 40 percent is allocated for butter while only 9 % is for cheese (Mohamed etal, 2003). In a study conducted in Borena region of Ethiopia, butter was found to be an important source of energy 10? ? as food for humans, and is used for cooking and as a cosmetic. The storage stability of butter still makes preferable to other dairy products. This gives butter a distinct advantage over fresh milk in terms of more temporal flexibility for household use and marketing (Layne, 1990). Efficiency of traditional butter production was measured for 28 instances in which soured milk was churned by women in 20 households of Borena region. Prior to churning, the milk had a temperature of 20.0 ? 0.420 C and an acidity of 1.06 ? 0.03%. The milk was churned for 40.0 ? 2.5 minutes and afterwards the temperature of the buttermilk was 23.7 ? 0.320 C . The sour milk contained about 46.8 g of fat, compared with 7 g of fat in the buttermilk after churning. Thus some 85% of the butterfat was extracted by churning. Butter yield was 66.9 ? 5.6 g but moisture content of the butter was not determined (Layne et al., 1990). Most of the milk produced in the country is processed by the producers themselves on-farm into butter and soft cheese (ayib) for home consumption and sale. Rural producers far away from urban markets usually process surplus milk into butter because of difficulties in selling fresh milk locally and the main butter markets are in the towns and cities (CSA, 2003). Apart from income, consumer preferences and dietary customs also partly explain relatively low demand for dairy products. Orthodox Christians comprising about 40% of the Ethiopian population abstain from consuming dairy and other animal products for about 200 days in a year. Thus, low demand for dairy products in Ethiopia compared to demand in other low- income countries in Sub-Saharan Africa appears to be a major reason for the slow growth of the dairy sector. The main outlets for cooking butter for rural producers near Addis Ababa were: (i) Restaurants in Addis Ababa and surrounding areas those serve local foods, (ii) Itinerant traders, and (iii) Individual consumers or butter wholesalers in Addis Ababa. Butter and some dairy products are called yellow fats, which contains a number of products for spreading onto bread or for indirect consumption as ingredients in other foods. There is some debate over product definition, and different systems of classification have distinguished products according to a variety of characteristics: the source of their raw 11? ? material (dairy fat, animal fat, and vegetable fat); their total fat content; their polyunsaturated fat content; and whether they are hard or soft (Traill et al., 1994). For example, a market research agency used the following definitions: butter (80 percent and over dairy fat); margarine (80 percent and over non-dairy fat); dairy spreads (usually a 75 percent fat blend of dairy and non-dairy fats); low-fat spreads (25 to 40 percent fat); and reduced?fat spreads (60 to 80 percent) (Mintel, 1990). 2.1.3. Common challenges of dairy production and marketing in Ethiopia Challenges and problems for dairying vary from one production system to another and/or from one location to another. The structure and performance of livestock and its products marketing both for domestic consumption and for export is generally perceived poor in Ethiopia. Underdevelopment and lack of market-oriented production, lack of adequate information on livestock resources, inadequate permanent trade routes and other facilities like feeds, water, holding grounds, lack or non-provision of transport, ineffectiveness and inadequate infrastructural and institutional set-ups, prevalence of diseases, illegal trade and inadequate market information (internal and external) are generally mentioned as some of the major reasons for the poor performance of this sector (Belachew 1998; Belachew and Jemberu 2003; Yacob as cited in Ayele et al., 2003). In the debate of poverty reduction or small-scale vs. industrial production and in spite of a general consensus on the appropriateness of general recommendations, there seem to be a lack of vision regarding the future structure and roles of the present small-scale producers. Many donors seem ready to protect and preserve the smallholders, but few have a vision of the process requiring ?transforming small-scale subsistence producers into commercial producers supplying a modern, demanding food market? (Kristensen et al., 2004). 2.2. Methodological and Conceptual Frameworks 2.2.1. Conceptual frame work of marketing Market: The choice of market definition depends on the problem to be analyzed. It may be defined by: i) a location (like Alamata market), ii) a product (butter market), iii) a time 12? ? (September butter market) and iv) a level (butter retail market). Originally, the term market stood for the place where buyers and sellers gathered to exchange their goods, such as village square. A market is a point, or place or sphere within which price-making force operates and in which exchanges of title tend to be accompanied by the actual movement of the good affected. A market is the set of the actual and potential buyers of a product (Kotler and Armstrong, 2003). These buyers share a particular need or want that can be satisfied through exchange relationships. The concept market is linked to the degree of communication among buyers and sellers, and the degree of substitutability among goods. A market is thought of as a meeting of buyers and sellers: a place where sellers and buyers meet and exchange takes place, on area where price determining forces (supply and demand) operate, an area where there is a demand for good. However, the term market means not a particular place in which goods are bought and sold. Market means a social institution that performs activities and provides facilities for exchanging commodities between buyers and sellers. Economically interpreted the term market refers not to a place but to a commodity or commodities and buyers and sellers are in free interaction with one another. Marketing: There is no universally accepted definition of marketing, the usefulness and validity of a definition is associated with its application. Specifically for this study the following definitions used. Kohl (1968), defined marketing in a way that is most applicable to agriculture. Accordingly: ?Marketing is the performance of all business activities involved in the flow of goods and services from the point of initial agricultural production until they are in the hands of ultimate consumers. Generally marketing is all those business activities associated with the flow of goods and services from production to consumption?. The marketing of agricultural products begins at the farm when the farmer plans his production to meet specific demand and market prospects (Abbott and Makeham, 1981). Marketing is usually seen as a ??system?? because it comprises several, usually stable, interrelated structures that, along with production, distribution, and consumption, underpin the economic process (Mendoza, 1995). In popular usage, the term "marketing" refers to the promotion of products, especially advertising and branding. However, in professional usage the term has a wider meaning. It can be divided into four sections, often called the "four Ps," 13? ? only one of which is promotion. These four elements are often referred to as the marketing mix. A marketer will use these variables to craft a marketing plan. For a marketing plan to be successful, the mix of the four "P's" must reflect the wants and desires of the consumers in the target market. They are: Product: The product management aspect of marketing deals with the specifications of the actual good or service, and how it relates to the end user has needs and wants. Pricing: This refers to the process of setting a price for a product, including discounts. Promotion: This includes advertising, sales promotion, publicity, and personal selling, and refers to the various methods of promoting the product, brand, or company. Place or distribution: This refers to how the product gets to the customer; for example, point of sale placement or retailing. Marketing Channel: The analysis of marketing channel is intended to provide a systematic knowledge of the flow of goods and services from their origin (producer) to their final destination (consumer). This knowledge is acquired by studying the? participants? in the processes, i.e. those who perform physical marketing functions in order to obtain economic benefits. In carrying these functions, marketing agents achieve both personal and social goals. They add value to production and by so doing help satisfy consumer needs. The price pays for the goods) the physical commodities and services (i.e. transportation, bulk breaking, grading) for the services and renders compensated the marketing agents for this effort. This price also serves as a signal to all actors in the marketing channel, i.e. producers, rural assemblers, and transporters, whole sellers and retailers (Mendoza, 1995). Producers: The first link in the butter supply chain. The producer harvests the products and supplies to the second agent. From the moment he/she decides what to produce, how much to grow and when to grow and sale. Consumer: The last link in the supply chain. The participants and their respective functions often overlap. The most widespread combinations are traders-whole sellers that collect the commodity and supply it to retailers to consumer. Rural assembler: Sometimes also known as transporter or trader, he/she is the first link between producer and other middle men 14? ? Whole seller: Concentrates the various, intermediate sized loads and puts the product into large, uniform units. These activities all contribute to price formation. Retailers: Middlemen, which include supper, market another large-scale retailer who divides large shipments of produce and sell it to consumers in small units. The basic function of that they provide is bulk breaking. Marketable and Marketed Surplus: Marketable surplus is the quantity of the product left out after meeting the farmers? consumption and utilization requirements for kind payment and other obligations such as gifts, donation and charity etc. Thus, marketable surplus shows the quantity left out for sale in the market. The marketed surplus shows the quantity actually sold after accounting for losses and retention by the farmers, if any adding the previous stock left out for sale. Thus, marketed surplus may be equal to marketable surplus, it may be less if the entire marketable surplus is not sold out and the farmers retain some stock and if losses are incurred at the farm or during transit (Malik et al., 199l). . 2.2.2. Methodological framework of marketing 2.2.2.1. Approaches to study marketing Economists take three major approaches to analyzing the marketing sector of the national economy. These include; the functional approach, the system or institutional approach and the individual or commodity approach (Mendoza, 1995; Branson and Norvell, 1983). The functional approach: In this approach, we took all the basic marketing activities (functions) that have to be performed in the agricultural commodities and at the marketing of inputs in to agricultural production. Physical distribution (i.e. functions) and economic activity (i.e. buying, selling) are two dimensions of marketing carried out by institutions or people. An analysis of these two dimensions of agricultural marketing is intimately linked to the institutions created by law or by corporate standards or simply by established procedure, that have emerged as a result of the social and economic relation between the participants in the marketing process (middlemen, consumers, and producers). 15? ? The system (institutional) approach: It is concerned with the number and kind of business firms that perform the marketing task. That means, it covers all market participants (producer, assembler, transporter, wholesaler, retailer and consumer). This approach includes market stabilization agencies boards of foreign trade, supermarket chains, wholesaler or retailer networks, a town?s central market, or agreements between producers and millers. The effectiveness of marketing institutions depends on the involvement of the relevant people. The commodity (individual) approach: This entails an analysis of marketing functions, system, and structure from the viewpoint of an individual product. This approach combines the above two approaches. We can study a list of products. In a commodity subsystem approach, the institutional analysis is based on the identification of the major marketing channels. This approach includes the analysis of marketing costs and margins. 2.2.2.2. Supply chain analysis and Factors affecting market supply ? A supply chain is ?a network of connected and interdependent organizations mutually and co- operatively working together to control, manage, and improve the flow of materials and information from suppliers to end users? (Christopher, 1998). Kohls and Uhl (1985), define supply as a schedule of different quantities that will be offered for sale at different prices at a given time and place. Many factors can combine to increase or decrease supply. Generally, they may be classified as follows: 1) in the short run, there may be a change in the various factors that would induce sellers to offer available stock of goods at different schedule of prices. 2) In the intermediate and long run periods, there may be a change in the costs of production of the commodity. This may be caused by changes in costs of needed inputs or in the technology of the production of the commodity itself. It may also be caused by changes in the costs of producing other commodities that compete for the same resources (Kohls and Uhl, 1985). The main factors, which determine market supply, could be divided into economic factors, which include product price, provision of consumer goods, production cost and market supply costs. Political factors, which include the level of government intervention. One of the expected important variables, which influence the behavior of the market supply of producers, is price. If price increases, producers will gain high revenue and would be 16? ? motivated to increase the market supply. For high-risk crops, high prices are necessary to call forth a given level of production. Prices also show increased variability because production plans are not always achieved ((Maro, 1986; cited in Wolday, 1994)). Agricultural commodities are produced by large number of farmers and consumed by large number of households. With the exception of foodstuffs consumed on farm or sold locally, they are bought and sold a number of times between the farm gate and the final consumer or industrial user. Commodities that are exported often change hands many times between the farm gate and points of final sale. While moving between these two points, the commodity is loaded, off loaded transported, stored, cleaned, graded and processed. Typically, each of these activities takes place on several occasions (Westlake, 2005). In the Ethiopian context, the presence of prohibitively high transaction costs, evidenced by 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, implies that buyers and sellers operate within narrow market channels, that is, only those channels for which they can obtain information and in which they have a few trusted trading partners. Extensive empirical analyses of Ethiopian market behavior thus reveals that market actors conduct business across short distances, with few partners, in few markets, and with limited storage, implying that opportunities for expanding market activity, otherwise known as arbitrage across space (transporting significant distances to market goods) and across time (storing for significant periods), are limited (Eleni, 2001). 2.2.2.3. Market structure, conduct and performance analysis (S-C-P model) Market structure: Market structure is defined as characteristics of the organization of a market, which seems to influence strategically the nature of competition by pricing behavior within the market. Structural characteristics may be used as a basis for classifying markets. Markets may be perfectly competitive, monopolistic, or oligopolistic. The four salient aspects of market structures include the degree of seller concentration, the degree of buyer concentration, the degree of product differentiation, and the condition of entry (Scott, 1995). 17? ? Market structure indicates all the firms engaged in a particular marketing channel. There are two strategic features. The first is the number and relative size of the firms involved. Do one or two so large firms dominate the others? The business relationship between them, are they interdependent or interlinked in ownership and management? Do formal contracts or informal understanding connect them? How easy is it for new firms to come into the system (Abbott, 1981)? So these and other similar questions will be prepared to know the structure of butter market in the study areas. Market concentration measures: The common measures of market concentration are: Concentration Ratio(C): ? r 1=i iS=C (1) i =1, 2, 3...r Where S i =the percentage market share of i th firm and r =the number of largest firms for which the ratio is going to be calculated. Very recently, the concentration ratio was the numerical index most widely used by industrial organizations for measuring the size of distribution of firms in market. While it is possible, use any economic variable such as employment, total assets or value added, for calculating C, sales or purchase figures have been the most popular basis for the index (Shughart, 1990; cited in Admasu, 1998). Kohls and Uhl (1985), suggested that as a rule of thumb, four largest enterprises concentration ratio of 50 percent or more is an indication of a strongly oligopolistic industry, 33 to 50 percent a weak oligopoly, and less than that, not concentrated industry. The problem associated with this index is the arbitrary selection of r (the number of firms that are taken to calculate the ratio).The ratio does not indicate the size distribution of the r firms. Hirschman Herfindahl Index (HHI): ? n 1=i 2 iS=HHI (2) i =1, 2, 3, ----, n Where: S i - is the percentage market share of i th firm, and n, total number of firms. This index takes into account all points on the concentration curve. It also considers the number and size distribution of all firms. In addition, squaring the individual market shares gives more weight 18? ? to the shares of the largest firms, which is an advantage over concentration ratio. Avery small index indicates the percentage of many firms of comparable size, whilst an index of one or near one suggest that the number of firms in small and /or that they have very unequal shares in the market. This method is limited in its application for it imposes additional burden in so far as more data must be collected (Scarborough and Kydd, 1992; cited in Admasu, 1998). Gini-Coefficient: Gini coefficient is an alternative concentration measure that has some similarities to the concentration ratio. It is based on Lorenz curve. To use the Lorenz curve, the firms in an industry are ranked from smallest to largest in terms of their market shares. Then the cumulative percentage of firms is related to their market shares. Gini-coefficient compares the area between the diagonal and Lorenz curve with the area of triangle under the diagonal (Bronfenbrenner, 1971; cited in Admasu, 1998). A simple method to calculate the coefficient is to estimate the area of the trapezoids underneath the Lorenz curve at each quartile, subtracting the total sum from 10,000 and dividing the difference by 10,000. The problem associated with Gini coefficient is that it favors equality of market shares without regard to the number of equalized firms. In other words, the coefficient equals zero. or two firms with 50 percent market share, for three firms with 33.33 percent market share each, and so on. Moreover, the coefficient is sensitive to market errors. The measured degree of inequality in an industry will tend to become larger as relatively small or relatively larger boarder line firms are included (Shughart, 1990; cited in Admassu, 1998). Market conduct: This refers to the market behavior of firms. In what way do they compute? Are they looking for new technique and do they apply them as practicable? Are they looking for new investment-opportunities or are they disinvesting and transforming funds elsewhere (Abbott and Makeham, 1981). Market conduct refers to the pattern of behavior that is followed in adapting or adjusting to the market in which they sell or buy (Scott, 1995). Such a definition implies that the analysis of human behavioral pattern that are not readily identifiable, obtainable or quantifiable. Thus in the absence of theoretical framework for market analysis, there is a tendency to treat conduct variables in a descriptive manner or as Ishak (1988), points it out ,as a spill-over in the assessment of market performance. 19? ? Bain (1968) names two closely interrelated aspects of market conduct, the manner in which, and the devices and mechanisms by which, the different sellers coordinate. Their intrinsically rivalries decisions and actions, adapted to each other, or succeed in making them mutually consistent as they react to demands for their products in a common market. In addition, ?the character of pricing policies and related market policies that the sellers in the industry adopt Assessed in terms of the individual or collective aims or goals that they pursue as they determine their selling prices, their sales promotion outlays, the designs and quantities of their products and so forth.? By examining the relationship between the factors of market structure and their price setting practices, it may be possible to make some predictions about the consequence of these behavioral patterns for performance (Scott, 1995). There are no agreed upon procedures for analyzing the elements of market conduct. Rather, previous researchers? point to some guide lines in the form of questions. These questions provide a systematic way to detect indications of unfair price setting practice and the condition under which such practice are likely to prevail. More specifically, they cover the following topics: the existence of formal and informal marketing groups that perpetuate such practice; formal and informal producer groups that affect bargaining power; the availability of price information and its impact on prevailing price; the distance from the major market and its impact on price; and the feasibility of utilizing alternative market outlets. The questions also provide an indication of the type of data needed and data collection procedures (Scott, 1995). Market performance: This is an assessment of how well the process of marketing is carried out and how successfully it aims are accomplished. Is produce assembled and delivered on time and without wastage? Is it well packed and present attractively? Is its quality reliable and are contracts kept? Is the consumption of products increasing and are sales in competitive market expanding? There are many such practical indications of how well a certain marketing system is operating (Abbott and Makeham, 1981). Market performance refers to the impact of structure and conduct as a measured in terms of variables such as prices, costs, and volume of output (Bressler and king, 1970; cited in Scott, 1995). By analyzing the level of marketing margin and their cost components, it is possible to evaluate the impact of the structure and conduct characteristic on market performance (Bain, 20? ? 1968). For most countries, it is generally acknowledged that a distribution system displaying acceptable performance is one that allows technological progress, has the ability to adapt, innovate and utilize resources efficiently and to transmit prices that reflect costs (OECD, 1982; cited in Scott, 1995). Prices are thus viewed as a stimulus for an efficient allocation of resources. Hence, desirable market performance is directly related to the competitiveness of an industry because distortions thereof tend to impend price efficiency (Scott, 1995). There are two major indicators of market performance: 1) net return and 2) marketing margin. Estimates of net return and marketing margin provide indication of an exploitative nature when net returns of buyers are much higher than the fair amount, that is including all marketing costs and returns to management and risk, and when marketing margins increase not because of higher real marketing costs but because price paid to producers are lower. The analysis of performance using the industrial organization framework is as follows. Collusive pricing (market conduct) become possible if ? Market concentration is high (market structure) ? Entry barriers are high (market structure) ? Market information is not available to all participants (market conduct) This results in net returns and marketing margins that are much higher than the?fair?amount (Scott, 1995). Hence, in this study, structure, conduct and performance (S-C-P) analysis is used to analyze butter market performance at Atsbi, Alamata and Mekelle markets. 2.2.2.4. Market participation The first and historically richest strand of literature concerns the determinants of small farmer participation in markets in semi-subsistence agrarian economies. This strand has focused primarily on 1) understanding the role of transactions costs and market failure in smallholder decision making and 2) resolving the econometric challenges to testing hypotheses concerning smallholder market participation in the presence of possible selection bias. Landmark theoretical contributions by de Janvry et al. (1991) develop formal household models to explain low smallholder supply response in the presence of food or labor market failure. A key conclusion of their work, empirically confirmed in a range of situations by von Braun et al. (1989), is that low productivity in food crop production, in the presence of food market failure, is a constraint to participation in cash crop markets. A corollary of this is that 21? ? a wide spread between food purchase and sale prices makes cash crops relatively more profitable for net food sellers than net food buyers. Thus if net marketable food surplus is causally related to household endowments of productive assets such as land and labor. Better- endowed households should be more likely than poorer households to participate in cash crop markets because the returns to cash crops are then directly related to household endowments (e.g., land, livestock) that typically have a strong, positive effect on the likelihood of being a net seller of basic staples (Barrett and Dorosh, 1996). Empirical analysis of the determinants of smallholder market participation has to deal with the econometric hazard of selection bias (Heckman, 1979). The problem arises because households (or individuals) face different types of decisions in relation to market participation ? a discrete decision over whether or not to participate in a given market as either a buyer or a seller, and a continuous decision as to how much to buy or sell conditional on market participation. Variables affecting the latter, continuous decision may affect the discrete participation decision while some factors ? e.g., fixed costs of market participation due to transport costs or vendor license fees ? that affect the discrete participation decision will not; in theory at least, affect the continuous decision. If unobserved preferences (e.g., risk aversion) or characteristics (e.g., liquidity constraints) affect both decisions, then regression estimates of the continuous choice will yield biased estimates absent correction for the first-stage participation choice. 2.2.2.5. Value chain; definition and importance The value chain describes the full range of activities which are required to bring a product or service from conception, through the different phases of production (involving a combination of physical transformation and the input of various producer services), delivery to final consumer, and final disposal after use (Kaplinsky, 2001). For at least twenty years now, there have been systematic attempts in English-, German-, and French-speaking schools of thoughts, to describe and analyze the vertical integration and disintegration of production and distribution processes. A great many terms were used in this connection, in part with identical and in part with varying meanings. 22? ? Value chain analysis addresses the nature and determinants of competitiveness, and makes a particular contribution in raising the sights from the individual enterprise to the group of interconnected enterprises. By focusing on all links (actors, enterprises, processes) in the chain and on all activities in each link, value chain analysis helps to identify which activities are subject to increasing returns, and which are subject to declining returns. Global Commodity Chains (GCC): the term "Global Commodity Chain (GCC)" was introduced in the mid 1990s (Gereffi, 2001). Gereffi is focusing on the power relations in the coordination of globally dispersed, but linked, production systems. He has shown that generally commodity chains are characterized by a leading party or parties that are determining the overall character of the chain. Gereffi differentiates between "producer-driven" and "buyer-driven" global commodity chains. Capital and technology- intensive industries such as automobiles, aircrafts or computers are typical examples for "producer-driven" global commodity chains, while labor-intensive industries such as consumer electronics or food production are examples for "buyer- driven"-chains. For the latter the specifications are supplied by the large retailers or marketers that order the goods. The four core elements of the GCC approach are the international dimension, power or governance, coordination and organizational learning. The main hypothesis of the GCC is linking up with the most significant "lead-firms" in an industry. Lead firms are distinguished from subordinated companies in terms of access to major resources (e.g. product design, brand names or consumer demand). World Economic Triangle: A concept pointing out that the combination of strong local linkages within global commodity chains might bring upgrading prospects for regions in developing countries; and thus is an approach for showing the importance of linking vertical (chains) and horizontal (clusters) integrations. Other authors (Humphrey et al., 2001) are pointing out that the combination of strong local linkages within global commodity chains might bring upgrading prospects for regions in developing countries. The concept of the "world economic triangle", where actors, governance and regulation systems are determining the scopes of action open to regions in the global commodity chains. He determined six critical aspects in any economic triangle; these are Actor constellations, Interests, Power structures, Situational mindsets, Action orientation and Trust. 23? ? 3. METHODOLOGY 3.1. Description of the Study Areas Alamata woreda Alamata is located in southern zone of Tigray 180km away from, Mekelle, capital of Tigray region on the main road to Addis Ababa. There are 10 peasant association and two-town dwellers association in the District. Agriculture is the main source of income of the population in the district. The total population living in the district is estimated to be 118,557 of which 58,591 were male (CSA, 2007). The total cultivated land is estimated to be 34,503 ha out of which 33,778.8 ha is cultivated through rain-fed while 724.2 ha is through irrigation. From the irrigated land, around 493.6 ha are irrigated through surface irrigation system using perennial rivers and 175 ha using privately owned ponds (shallow wells). There are also two pilot sprinklers and/or drip irrigation system being implemented in the area with total of 55.6 ha with regard to sprinkler and/or drip irrigation system it is believed to have in the near future 99 deep wells with potential of irrigating 3997 ha of land (REST, 1998). At the moment, 30 deep wells dug out in the district with an estimated potential of irrigating 900 ha of land. Altitude in the area ranges from 1178 to 3148 m and 75% of the district is low land (1500 masl or below and only 25% is found in intermediate high lands (1500 and 3148 masl). The small undulating mountains surrounding the district are very steep and with low vegetation cover. a large area and have a series of dissected gullies which serve as a source of runoff water and alluvial soil to the Alamata valley. The district is characterized with bimodal rainfall with average annual rainfall of 663 mm. Flood diversion is the most commonly used traditional system of supplementing the erratic rainfall pattern of the area. In eight of the seasonal rivers that pass through the district , it is estimated that around 6621 ha of land can be irrigated using flood coming from high land areas of the district during summer season (REST, 1998). The average annual temperature is 29.7 c o with the maximum and 14.6 c o the minimum averaging 22.2 c o . 24? ? The dominant crops produced in the district are mostly cereals, pulses and oil seeds of the cereals sorghum, teff, and maize takes the largest portion of production. It is estimated the district has livestock population 106,461 of which cattle population 74,853 comprises the major share followed by small ruminants with a population of 24,971. Having this potential the district is suffering from well organized systematized market oriented production system that discourages production and productivity of livestock as required. Atsbi-Wonberta woreda Location: Atsbi Wonberta district is found in eastern Zone of Tigray region at about 65 Km from Mekelle regional city. The district is bordered in north by Saese Tsaedaemba district, in the south by Enderta district, in the east by Afar regional state and in the west by Kilteawlaelo district. It has an altitude at Dega (high land ),which ranges from 2400 m to 3000 m and at Weynadega (mid land) ranging from 1800 m to 2400 m above sea level. The district has a total area of about 1223 sq Km. Generally the district has 70% and 30% Dega and Weynadega, respectively. The major types of land use are forest 89,185 ha, grazing land 8,742 ha, potential cultivated 35,305 ha, cultivated 13,050.23 ha (ARDO, 2006). There is also bee forage planting practices in the study area such as supplementary feeding which includes sugar, barley flour, peas and beans flour. In both the traditional and modern beehives, supplementary feed is provided. In the study area there is also an extension activity that encourages beekeepers to grow indigenous bee forage such as ?gribiya? (Hypostus ariculata) and ?tebeb? (Basium clandiforbium). These plants are herbaceous and contribute high in honey production of the area. Climate: The climate of Atsbi Wonberta ranges from cool to warm. The average temperature of the area is 18 degree centigrade. Generally, the climate of the area characterized as high land and middle land. Rainfall is usually intense and short duration, with an annual average of about 667.8 mm. Demographic features: According to the information from district Agricultural and Rural Development Office (2006), Atsbi Wonberta District has a total population of 112,639, of which 55,359 (49.15%) are males and 57,280 (50.85%) are females. Urban and rural population is 9,609 and 103,030 respectively. 25? ? ? Figure 1. Location of the study area, Alamata woreda M e c hare H a w e lti H a d e A l g a Korme Ebo Gen e te K a r a A dish a?h?o? W e re A bay e Kuk u fto B?ala U la ga A di?s? ? K i?gni S e len Wu?ha? M? a r? u? T u m u ga Lima a t T s e ts e ra M?e?r iw a S/B e k als i Gerjale T a o D a yu K / G/ Lemlem Me k h o n i Al am at a T own We yr a Wuha T/ W a r gi b a 2?0?0?0?0?0?0? 020 0 0 0 0 0 4 0 0 0 0 0 0Miles N Adis K igni Alam ata Tow n Bala U laga Ebo Genete Ger jale Hade A lga Hawelti Kar a A d isha ho Kor me Kukufto Kulu Gize Lemlem Laelay D ayu Limaat Maru Mecha re Mekho n i Meriw a Selam B ekals i Selen W uha Tao Tsetser a Tsigea Wargiba Tum uga Waja Wer e A baye Wey raWuha Legend ???? ? Alam? ata 26? ? ATSBI-WENBERTA Figure 2: Location of study Area, Atsbi-Wonberta woreda 27? ? Crop and livestock Production: According to the information from Atsbi-Wonberta district ARD planning Office (2006), the dominant cereal crops of the area are barley, wheat, teff, maize and sorghum. The important marketable crop commodities in this area are pulses (beans, field pea and lentils, in the order). Despite the large population of livestock, especially sheep, livestock productivity is low as in many other parts of the region. As most of the district is in the highlands, it is suitable for sheep production. Dairy and horticultural crops are also other important marketable commodities in the area. According to the information from Atsbi-Wonberta district ARD planning Office (2006), livestock sources are cattle-52,496, of which 30588 are cows, goat-15431, sheep-82950, poultry-47265, donkey- 9416, mule-1333, horse-79, camel-54 found in the District. Apiculture is another important source of household income in the study area. Studies indicate that, there are farmers who own up to 100 local beehives in the area. There are 16,915 honeybee colonies in the district, out of which about 2000 are modern hives. Currently, in the district, the maximum honey yield obtained per improved box hive and traditional hive is 40 kg and 12 kg respectively, where as the minimum honey production from both beehives is 10 kg and 2 kg, respectively (IPMS, 2004). 3.2. Data Type, Sources, and Methods of Data Collection 3.2.1. Data type and sources Data on production and marketing of butter, exchange arrangements, system of storage, transport facilities and supporting institutions were collected from sample respondents using questionnaire. Data on quantity of butter marketed, price of butter, price of other dairy products produced in the farm, total volume of butter produced, expenditure on inputs of production, and households socio-economic, demographic and spatial settings was collected from sample informants using questionnaire. In addition to primary data on the above issues, secondary data on number of licensed and unlicensed traders, marketing agents and their role, conversion factors of livestock units, adult equivalent, etc were collected from secondary sources. Sources of primary data are smallholder farmers and marketing intermediaries. Secondary data sources include woreda Office of Agriculture and Rural development, 28? ? cooperatives at different levels, Trade and Industry office, Tigray Regional marketing agency, and different publications etc., which have data relevant to the study. 3.2.2. Methods of data collection The data used for this study was collected both from primary and secondary sources. Secondary data was collected from different institutions, organizations and offices as well as through reviewing documents and publications. Primary data on the production and marketing system was collected from the producing farmer up to the end consumer through structured questionnaires. 3.3. Sample Size and Procedure The decision involved are partly a function of the information currently known, time and resource available, accessibility to and openness of the marketing participants themselves, as well as the estimated size of trading population. There is no ironclad rule to help one determine the number of interviews required for each stage or segment of the supply chain. The establishment of fixed procedure could prove excessive for same segments of the study and insufficient for others. Sampling by segments without size limits established initially can simplify things as a result prior determination of the number of respondents is set to each category of respondents. By adopting the proportional random selection method, 200 butter producer households from the two woredas and total of 56 traders at different levels were selected from the three markets. 3.3.1. Farmers sampling As already noted, this study was designed to assess butter supply chain in the specified districts. In addition to this, an attempt was made to assess the production characteristics, profitability and market supply of butter by the households in the study sites. In order to achieve these goals, a two stage sampling procedure was adopted. The first stage involved the random selection of rural kebelles in the study sites. Based on the distribution of population, a total of 14 Kebelles from Atsbi-wonberta woreda and seven kebelles from Alamata woreda were selected. These criteria were adopted because they determine largely the behavior of 29? ? butter production and marketing characteristics of the districts. The second stage of sampling involved the selection of respondents. Once the list of butter producing household in each selected kebelles was obtained, household heads were selected proportionally using the random selection method. As the number of butter producer households in these woredas is nearly equal, 100 sample producers were selected from each woreda making the total number of sample producers to be 200 from the two woredas. 3.3.2. Wholesalers and assemblers sampling Similarly, butter traders were selected from three markets (Atsbi-wonberta, Alamata and Mekelle), in which these markets are recipient of the produce from the selected producing areas. Sampling here is the very difficult task due to the opportunistic behavior of the traders. However, to have the possible level of representative prior to formal traders? survey, a rapid market appraisal (RMA) was conducted in order to get the overall picture of butter supply chain in the Districts. The sample size of butter whole sellers and assemblers was determined on the number of traders that were known after the informal market survey. As the number of the whole sellers and assemblers at each stage was very low, consequently, all of them were interviewed. A total of 8 wholesalers from the region capital city, Mekelle and 16 assemblers from the three markets were sampled in this stage. 3.3.3. Retailers sampling Selection of retailers in Atsbi-Wonberta, Alamata and Mekelle towns was made based on the size of the markets. A50 percent share for Atsbi-Wonberta and Alamata since they are relatively small markets while Mekelle town had taken a 50 percent share of the samples, total of 32 sample retailers were selected randomly. Data to be collected from retailers was mainly focus on the characteristics of market structure, conduct, and performance of the market. 3.4. Methods of Data Analysis Both descriptive statistics and econometric methods of data analysis are used. 30? ? ?3.4.1. Structure of production costs and profitability of butter production To meet the objective all production costs (expenses), farmer?s sale (revenue) was identified and collected, based on the economic value of resource used in butter production, the structure of production costs and returns was formulated. By deducting economic costs from total return of butter, net return of butter production was calculated, and this result in turn was used to determine the profitability of butter production in the selected woredas using the following formula. n n 1=t t n 1=t _t )r+1( C? =NPV ?? (3) Where NPV = is net present value B t = benefit from one year C t = cost incurred in a year r = interest rate per year n= number of years Since the main target of this section is to identify production cost of butter and determine its profitability, production costs were disintegrated to milk, butter, appreciation of cattle (i.e., calves, heifer and young bull) and cow dang and/or manure. This was done by distributing the production costs according to their share of return to the households? income from the dairy enterprise. Therefore, production costs belong to the other dairy out puts based on their income share were identified and deducted out from the profitability analysis. The procedures employed were: ? The possible outputs from the dairy farm were identified by the farmers and listed as milk, butter, calves, dung, and manure. ? Secondly, the amount of income the farmers earned from these dairy partitions was recorded and computed their proportions from the total dairy income. ? Thirdly, the total production costs were distributed proportional to their income share to the total dairy income. 31? ? ? Then, total production costs of whole milk were identified neatly off the calves, manure and dung costs. ? Finally, based on the proportion of milk converted to butter, the costs also shared accordingly. In addition to these costs, there are also production costs belong to butter only like labor cost of churning, deprecation of churning and other butter equipments are considered. 3.4.2. Analysis of butter market supply If two decisions are involved, such as participation and level of supply, Heckman?s two-stage estimation is the recommended econometric model. This model allows the producer to choose whether to participate in a particular market, and if so, to choose the level of supply. Thus, a Heckman (1979) two-stage procedure is used in which the inverse Mill?s Ratio is calculated from probit estimation of decision to sell and introduced into the supply equation. Procedures for estimating butter market participation decision and level of supply Ideally, the OLS is applicable to determining factors that affect the level of participation. However, some households may prefer not to participate in a particular market in favor of others, where as others may be excluded because of market conditions or households resource constraints. If OLS regression is estimated while excluding the non-participating from analysis, a sample selectivity bias is introduced into the model. Such a problem can be overcome by following two-step procedure, as suggested by Heckman (1979). In this study, therefore, the Heckman?s two-stage selectivity model is used to investigate the factors that influence the probability of being participated in butter marketing. While secondly estimating the factors affecting the level of supply using OLS. The first step of Heckman procedure establishes the probability of participation in the output market. For the individual producer, the decision to participate or not to participate in Butter marketing can be formulated as binary choice model that can be analyzed using the probit equation below. The empirical specification of the probit model to be estimated by maximum likelihood estimation is defined as: ? 32? ? ii * i ?+?X=BMP (4) 1=BMPi If?? 0>BMP * i ? 0=BMPi If?? 050 years 0.55 0.50 Source: storck et al., 1991 Appendix table 4. Test for Multicollinearity Continuous independent variables Collinearity Statistics Tolerance VIF Total annual butter production .697 1.435 Total family size .317 3.150 Labor supply .258 3.881 Years of experience in butter production .653 1.531 Age .697 1.435 Off farm income .928 1.078 Frequency of extension contact .647 2.981 Non dairy financial income .770 1.298 Distance to the nearest market center .624 1.602 Distance to the development center .609 1.643 Distance to the woreda town .688 1.454 Distance to weathered road .728 3.510 livestock owned in TLUequivalent .436 2.292 Number of local breed cows .287 2.392 Number of crossbreed cows .510 1.457 Total cultivated land in hectare .483 2.398 Number of oxen .612 1.673 Source: own computation Note: In all cases, VIF is less than 10 hence, no high degree of multicollinearity 132? ? Appendix table 5. Contingency coefficient SEX EDU VETACS DEMOPAR MINFO SEX 1 EDU .179 1 VETACS .021 .170 1 DEMOPAR .113 .212 .014 1 MINFO .067 .177 .018 .251 1 Source: own computation Note: In all cases contingency coefficient is less than one hence, no high degree of association is observed. Appendix table 6. Commitment Level of Export Subsidies on Milk Products (Rs. /Kg) 1995 1996 1997 1998 Canada Butter 146.23 59.14 62.73 51.20 Skim Milk Powder 24.57 33.70 34.87 33.29 Cheese - - - - European Union Butter & Butter Oil 315.58 333.11 232.57 29087 Skim Milk Powder 40.50 74.02 49.41 65.12 Cheese 45.87 46.84 41.69 48.17 U.S.A. Butter & Butter Oil 22.68 78.17 17.82 186.81 Skim Milk Powder 23.94 125.51 32.79 38.86 Cheese 5.41 4.94 4.21 1.48 Source: Chand, Ramesh and Linu Mathew (2001)