iANALYSIS OF COMMON BEAN CROSS BORDER TRADE IN EAST AFRICA: THE CASE OF KENYA ODERA EMMA VIVIAN KM17/3003/11 A Thesis Submitted to the Graduate School in Partial Fulfillment of the requirements for the Award of the Master of Science Degree in Agricultural and Applied Economics of Egerton University. EGERTON UNIVERSITY January 2015 ii DECLARATION AND APPROVAL DECLARATION I declare that this thesis is my original work and has not been submitted in this or any other university for the award of any degree. Odera Emma Vivian Reg. No. KM17/3003/11 Signature: ………………….………….. Date: ……………………………. APPROVAL This thesis has been submitted to the graduate school with our approval as university supervisors. Prof. Patience Mshenga (PhD) Department of Agricultural Economics and Agribusiness Management, Egerton University Signature: …………………………… Date: …………………………. Dr. Eliud Birachi (PhD) Market value chain specialist with the International Center for Tropical Agriculture (CIAT), Kenya Signature: Date: 30th Jan 2015 iii COPYRIGHT No part or whole of this thesis may be reproduced, stored in a retrieval system or transmitted in any form or means such as electronic, mechanical or photocopying without the prior written permission of Egerton University on behalf of the author. © 2015 Emma Odera All rights reserved iv DEDICATION I dedicate this work to Dennis Ochieng for his unconditional love, support and for being an inspiring husband. My beloved son Isaiah Paul Otieno and my mum and siblings, thanks for your support. vACKNOWLEDGEMENT I would like to thank God for giving me good health during my entire study. I give my sincere gratitude to the entire staff of AGEC/AGBM department under the leadership of Prof. Mutai for their steadfast support. I gratefully acknowledge the support and guidance from my supervisors Prof. Mshenga and Dr. Birachi. Sincere gratitude to the CMAAE Program which sponsored my studies in Pretoria and partially my research work. Material support from CIAT team during data collection along the Kenya border points is dully acknowledged. The contribution and positive critiques from other individuals and friends will not also go unmentioned. To God be the glory! vi ABSTRACT East Africa region has imbalances in the supply and demand of common bean. This can be offset by improving marketing infrastructure. The objectives of the study were to determine the characteristics of common bean traders, to determine the constraints to the observed trade in common bean varieties and finally to assess the extent to which markets have integrated at key selected markets. Multi-stage sampling technique was used to obtain a sample of 240 respondents (120 traders from the border points and 120 traders from key selected markets). The four border points (Busia, Malaba, Isebania, and Namanga) were purposively selected due to the extent of activities, nature of trade and the volumes of common beans that they handled. The three key markets (Nairobi, Mombasa, and Nakuru) were also purposively selected because of high potential demand and supply for common beans. Snow balling method was used to select the traders. Descriptive statistics were used to address objective 1 and objective 2 and co- integration analysis was used to address objective 3. The findings revealed that a greater proportion of the traders were women, majority being retailers. The women traders also had more years of experience on average in the retail business compared to the men. Results also indicated that the major constraints to bean trade were high transportation costs, heavy rains, and irregularities in bean supply. Nyayo and Wairimu bean varieties were the most traded in the markets whereas Saitoti variety was the least traded. The co-integration tests established that Nairobi- Mombasa and Nakuru- Mombasa Rosecoco markets and Nairobi- Nakuru and Nakuru- Mombasa Mwitemania markets were co-integrated. The study therefore recommends that the government should improve marketing infrastructure especially the roads to enable easy flow of the product between the markets. vii TABLE OF CONTENTS DECLARATION AND APPROVAL.......................................................................................... ii DEDICATION.............................................................................................................................. iv ACKNOWLEDGEMENT............................................................................................................ v ABSTRACT.................................................................................................................................. vi LIST OF FIGURES ...................................................................................................................... x ACRONYMS AND ABBREVIATIONS.................................................................................... xi CHAPTER ONE ........................................................................................................................... 1 INTRODUCTION......................................................................................................................... 1 1.1 Background information ....................................................................................................... 1 1.1.1 Common bean production in the world .......................................................................... 2 1.1.2 Common bean consumption in East Africa.................................................................... 3 1.2 Statement of the problem ...................................................................................................... 4 1.3 Objectives of the study.......................................................................................................... 4 1.3.1 General objective............................................................................................................ 4 1.3.2 Specific objectives.......................................................................................................... 4 1.4 Research questions ................................................................................................................ 5 1.5 Justification of the study ....................................................................................................... 5 1.6 Scope and limitation of the study.......................................................................................... 5 1.7 Definition of terms ................................................................................................................ 6 CHAPTER TWO .......................................................................................................................... 7 LITERATURE REVIEW ............................................................................................................ 7 2.1 Common bean varieties produced in East Africa.................................................................. 7 2.2 Common bean trade in East Africa ....................................................................................... 8 2.3 Market integration analysis ............................................................................................. 11 2.4 Price analysis ................................................................................................................... 13 2.6 Conceptual framework ........................................................................................................ 16 CHAPTER THREE .................................................................................................................... 17 METHODOLOGY ..................................................................................................................... 18 3.1 Study Areas ......................................................................................................................... 18 3.2 Respondents ........................................................................................................................ 19 viii 3.3 Sampling procedure and sample size .................................................................................. 19 3.4 Data Types and Sources ...................................................................................................... 20 3.5 Data analysis ....................................................................................................................... 20 CHAPTER FOUR....................................................................................................................... 24 RESULTS AND DISCUSSIONS ............................................................................................... 24 4.1 Descriptive Analysis ........................................................................................................... 24 4.1.1 Characteristics of common bean traders .......................................................................... 24 4.1.2 Common bean varieties traded......................................................................................... 28 4.1.3 Constraints to bean trade .................................................................................................. 29 4.2 Results of Integration Analysis ........................................................................................... 31 4.2.1 Price trends ................................................................................................................... 32 4.2.2 Unit root test results...................................................................................................... 34 4.2.3 Co-integration test results............................................................................................. 36 CHAPTER FIVE ........................................................................................................................ 40 CONCLUSION AND RECOMMENDATIONS ...................................................................... 40 5.1 Conclusions ......................................................................................................................... 40 5.2 Recommendations ............................................................................................................... 40 5.3 Suggestions for further research.......................................................................................... 41 REFERENCES............................................................................................................................ 42 APPENDICES ............................................................................................................................. 46 APPENDIX ONE: INTERVIEW SCHEDULE FOR TRADERS AND TRANSPORTERS ................................................................................................................................................... 46 APPENDIX TWO: INTERVIEW SCHEDULE FOR CUSTOMS OFFICIALS ............. 55 ix LIST OF TABLES Table 1: Eastern Africa common bean production in tons ............................................................. 2 Table 2: Cross border yearly volumes for common bean for the Year 2012 in metric tons......... 10 Table 3: Border points studied...................................................................................................... 19 Table 4: Age and Experience of traders........................................................................................ 24 Table 5: Types of traders .............................................................................................................. 27 Table 6: Category of trade by gender ........................................................................................... 27 Table 7: Common bean varieties traded at the border points and key markets ............................ 29 Table 8: Constraints faced by traders............................................................................................ 30 Table 9: Unit Roots Test for Rosecoco Bean................................................................................ 35 Table 10: Unit Roots Test for Mwezi moja Bean ......................................................................... 35 Table 11: Unit Roots Test for Mwitemania Bean......................................................................... 36 Table 12: Cointegration tests between Nairobi and Mombasa and Nakuru for Rosecoco ........... 37 Table 13: Cointegration test between Nairobi and Nakuru markets for Mwezi Moja.................. 38 Table 14: Cointegration tests between Nairobi, Mombasa and Nakuru markets for Mwitemania. ....................................................................................................................................................... 39 xLIST OF FIGURES Figure 1: Conceptual framework for factors influencing efficient market. .................................. 17 Figure 2: Overall level education level......................................................................................... 25 Figure 3: Level of education by gender ........................................................................................ 26 Figure 4: Seasonal variation of Rosecoco bean prices in Nairobi, Mombasa and Nakuru........... 32 Figure 5: Seasonal variation of Mwezi moja bean prices in Nairobi and Nakuru........................ 33 Figure 6: Seasonal variation of Mwitemania bean prices in Nairobi, Mombasa and Nakuru ...... 34 xi ACRONYMS AND ABBREVIATIONS CIAT International Center for Tropical Agriculture CGIAR Consultative Group on International Agricultural Research COMESA Common Market for Eastern and Southern Africa EAC East African Community EABC East Africa Business Council FAO Food and Agriculture Organization FAOSTAT Food and Agriculture Organization Statistics FEWSNET Famine Early Warning Systems Network KEPHIS Kenya Plant Health Inspectorate Service Kg Kilogram Kshs Kenya shillings RATIN Regional Agricultural Trade Intelligence Network USD United States dollars 1CHAPTER ONE INTRODUCTION 1.1 Background information The global economy is integrating rapidly through trade such that exports from developing countries are becoming increasingly diversified. In turn, these countries have become less dependent on agricultural exports than they were in the past. Currently, developing countries are becoming their own best markets for agricultural products. This is as a result of countries trading with each other. In order to facilitate trade in Africa, it’s important to remove bottlenecks that hinder cross border trade such as; bribery, long custom procedures, and complex import/export requirements. The Common Market for Southern and Eastern Africa (COMESA) and the East African Community (EAC) are some of the regional bodies that facilitate trade. These trading blocs harmonize and standardize trade procedures as well as the administration of border controls. For example, the major role of the EAC Customs Union is to deepen the integration process through liberalization and promotion of regional commercial integration through intra-regional trade (EABC, 2008). Trade in East Africa involves all types of goods such as agricultural commodities and manufactured goods. Agricultural commodities flowing across the borders include staple food commodities such as maize, beans, rice, fish, groundnuts and banana. Manufactured consumer goods traded are shoes, textile, medicine, vehicles and bicycle parts (Uganda Bureau of Statistics, 2007). It is estimated that 26% of Kenya’s exports to the EAC are evenly distributed in Uganda, Tanzania and Rwanda. On the other hand, Kenya is a significant destination for Tanzania’s exports estimated at 44% while Uganda’s exports to Tanzania are approximately 25% (EABC, 2008). Of the agricultural goods traded in East Africa, common bean is the major staple food that is traded across the borders (RATIN, 2011). In recent years, trade opportunities in common bean for both exports within the Eastern and Southern Africa region have increased. Kenya and Malawi show a huge potential for import market of common beans which can be tapped by their neighbors, particularly Tanzania, Uganda and the great lakes region (Katungi et al., 2009). Some of the bean corridors according to Pan-Africa Bean Research Alliance (PABRA) are around North West Tanzania, Burundi, Rwanda and Eastern DRC into Kenya ; 2South west Uganda, Northern Rwanda and Eastern DRC into Kenya; Northern Tanzania into Kenya and Ethiopia including the rift valley destined to Kenya. 1.1.1 Common bean production in the world Common bean is globally grown in nearly 28 million hectares with a production of 20 million tons. The average yield of common bean has been increasing over the past years with a range of 493kg/ha in 1961 to 729kg/ha in 2008 (FAOSTAT, 2008). Production of common bean is highest in Latin America with about 5.5 million metric tons being produced with the major producers being Brazil and Mexico. Africa is the second largest producer of common bean with a production of 2.5 million metric tons. In Africa, the leading producers are Uganda, Kenya, Tanzania, Rwanda, Burundi, Ethiopia, Malawi and Congo (CGIAR, 2012). Most of the common bean varieties grown for the market are the sugar type, red mottled, large red kidneys, small and medium reds, yellows, tan/khaki (pinto), cream, white, purples and blacks. Common bean is grown twice a year in Eastern Africa. The sowing seasons run from March to April and from September to October with the exception of Ethiopia in which the main growing season is June to August (Rukandema et al., 1981; Wortmann et al., 1998; Ferris and Kaganzi, 2008). Table 1: Eastern Africa common bean production in tons Country 2010 2011 2012 Burundi 201,551 200,673 205,944 Ethiopia 362,890 340,280 463,009 Kenya 390,598 577,674 613,902 Malawi 153,815 288,414 185,578 Rwanda 327,497 331,166 432,857 Uganda 455,000 464,105 425,400 Tanzania 867,530 675,948 1,199,267 (Source: FAOSTAT, 2012) Common bean production varies from country to country in East Africa due to varied soil type, climatic conditions and adoption of high yielding bean varieties by farmers. As shown in Table 1, the production of common bean in Kenya has been increasing over the past years. Uganda's 3common bean production has been steady with a slight drop in the year 2012 while in Tanzania there has been a fluctuation in the three years with the highest production being in 2012. This is due to increased acreage in Kenya, Uganda and Tanzania thus increase in production. 1.1.2 Common bean consumption in East Africa Common bean provides dietary protein for over 100 million people in both the rural and poor urban areas. Studies have shown that there is a high per capita consumption of common bean in developing countries (13-40 kg per year) especially with low income families in urban and rural areas (Singh, 1999). The per capita consumption of common bean is high in poor countries such as Nicaragua (22.5 kg per capita per year) and in poorer regions of higher income countries such as Northeast Brazil (18.5 kilogram per capita per year) (CIAT, 2012). Eastern Africa has the highest per capita consumption in the world that ranges from 50 to 60kg (ISAR, 2011). The per capita consumption of common bean in Kenya is estimated at 14 kg per year (Spilsbury et al., 2004; Buruchara, 2007). Consumption of common bean in Karagwe district in Tanzania is higher than the national average and is estimated at 13 Kg per year (Xavery et al., 2005). In Uganda the per capita consumption of common bean is over 58 kg (Soniia, 1999). Apart from the pulse being an important food source, common bean has the potential to generate incomes if key markets are harnessed through contracts from other countries for the overall development of the economies. In addition, production of common bean contributes to the inputs, transport, processing, retailing, packaging and the formal and informal trade sector. However, the importance of regional trade of bean appears underestimated with focus on consumption and export markets. The improved bean varieties are not well known among most smallholder farmers in Kenya thus the need to promote these varieties. This is because the varieties developed in the 1980s that are low yielding and susceptible to diseases have been passed from farmer to farmer and saved from season to season. According to the Ministry of Agriculture (2009), the consumption of beans between 2004 and 2008 was approximately 464, 000 metric tons while the production was 357, 00 metric tons which resulted in a deficit of 107, 000 metric tons. Over the last few years in Kenya, the price per ton of common bean has considerably increased, from USD 760 in 2010, to USD 837.5 in 2011 then USD 866.1 in 2012 (FAOSTAT, 2013). This price change significantly affects the purchasing power of poor people in both rural 4and urban areas and also points to possible inefficiencies in grain distribution system between surplus and deficit areas. According to Kibiego et al. (2003), common bean deficit in Kenya suggests an apparent market failure to stimulate production. This is caused by seasonal price fluctuation and lack of statistical data on bean marketing. To ensure production of different common bean varieties in the market at affordable prices, quantity and quality, an efficient marketing system is required. 1.2 Statement of the problem East African region produces most of the common bean in Africa and Kenya is a major consumer of it. Despite increased regional trade and increased common bean production in Kenya over the past years, its demand still outweighs supply and consequently prices have increased significantly. The constraints to common bean trade and effectiveness of the marketing systems are not yet clearly known. Past studies have also looked into the issue of market information systems in the region with a view to strengthening it so as to correct supply and demand imbalances. However, integration of bean markets has received less attention therefore it is not known to what extent the border and major bean markets are integrated in the region to facilitate steady product flows across markets. This would help stabilize bean prices as well as increase its availability and consumption thereby contributing to food security and nutrition at household level. 1.3 Objectives of the study 1.3.1 General objective The general objective of the study was to examine cross border trade for different common bean varieties in Kenya. 1.3.2 Specific objectives The specific objectives of the study were: 1. To determine the characteristics of common bean traders in key selected border points and markets. 2. To determine the constraints to the observed trade in the selected border points and markets. 53. To assess the extent to which markets of common bean are integrated in key selected markets. 1.4 Research questions 1. What are the characteristics of common bean traders at key border points and markets? 2. What are the constraints to the observed trade in common bean? 3. What is the extent of market integration in key selected bean markets? 1.5 Justification of the study One of the major policy needs in East Africa is to curb food insecurity and promote regional trade. To attain this, the East African community should aim at facilitating agricultural trade among its members by eliminating so of the bottlenecks to cross border trade. This will become increasingly important in linking food surplus areas to food deficit areas especially as development is driven by increased population and activities in towns and cities. Thus, this study sought to generate valuable information on most traded bean varieties and market integration which can be used to develop the efficiency of trade in East Africa. Information on market flow of common bean enables improvement of policies to increase marketing efficiencies and also to maintain quality of varieties being developed by monitoring their movement across the borders. Farmers will be able to produce bean varieties that are highly demanded in the market. Traders will also benefit from the knowledge of the most traded varieties to trade in and which are the best markets to trade in to increase their profit margin. The government will be able formulate policies to assist in promoting new drought tolerant and disease resistant bean varieties. 1.6 Scope and limitation of the study This study focused on traders, transporters, crop inspectorate officials and customs officials at the border points of which the sampling unit was from the key selected border points and markets in Kenya. Reliance on memory recall of those traders and transporters who did not keep records affected precision of data collected but the study supplemented the information through records kept by the customs officials at the border points. The study also utilized secondary data on the monthly average prices for three bean varieties (Mwitemania, Rosecoco and Mwezi moja) collected from the Ministry of Agriculture covering a period of three years (2011 to 2013) because prices for different varieties were incomplete in earlier years. 61.7 Definition of terms Common bean: It is used in this study interchangeably with bean and scientifically known as Phaseolus vulgaris L. Food security: The state in which the food demands of people are met at all times. Cross border trade: The buying and selling of commodities with the seller being in one country and the buyer being in another country. Market performance: This study adopted the definition of market performance from Harris (1993) as representation of economic results of the structure and conduct, in particular the relationship between distributive margins and the costs of production and marketing services. Market efficiency: This study adopted the definition of market efficiency from Barrett and Li (2002) as the transfer of excess demand from one market to another, manifested in the physical flow of the commodity, the transmission of price shocks from one market to another or both. Market: Refers to a place where willing buyers and sellers exchange money for commodities. Market integration: This is the flow of commodities from surplus to deficit markets and transmission of price shocks from one market to another. 7CHAPTER TWO LITERATURE REVIEW 2.1 Common bean varieties produced in East Africa Preferences for common bean varieties vary among farmers, traders and consumers. Furthermore, common bean varieties vary in their adaptation to diverse environment including the biotic and abiotic stress factors (Chirwa et al., 2007). Bean varieties can be classified into nine major classes according to color and size as follows: pure large reds, medium and small reds and red mottled, purple, yellow and tans, cream, navy/white and black. The red and the red mottled are the most common types due to market preference (Wortmann et al., 1998). Further, literature notes that, a wide range of seed colors and sizes is acceptable in many common bean production areas in Africa (Van Rheenen, 1979; Grisley and Mwesigwa, 1991a; Voysest and Dessert, 1991; Grisley and Munene, 1992). The large and the medium sized seeds are mostly preferred but also small seeds are acceptable. Tanzania's production of common bean includes both local and improved varieties which differ by color, shape, size and properties such as cooking time and digestibility (Wortmann et al., 1998 and Fulgence et al. 2009). The most common varieties grown in Tanzania are the Lyamungu 85 and the large red/brown Calima, or Rosecoco. This is because of their relatively high and stable productivity under moderately good growing conditions and the high market preference in Kenya (Wortmann et al., 1998). Other studies done by Kweka et al. (1998) and; Nkonya et al. (1998) found the Lyamungu 85 was rapidly adopted than Canadian wonder in Northern Tanzania because it cooked fast and its palatability is better than Canadian wonder. Other varieties include Soya 4 and 5, Canadian wonder, Tikyakuponza and Lyamungu 90. Soya is preferred by urban consumers in towns of Northern zone and some of the coastal towns like Tanga, Dar es Salaam and Zanzibar and it also fetches high prices in the market (Katungi et al., 2009). The purples locally known as Mwezi moja in Tanzania are preferred in Dar es Salaam because it cooks quickly, is tasty and produces a reddish broth (Wortmann et al., 1998). According to Korir et al. (2005), majority of retailers in northern Tanzania preferred Soya (a local, medium sized, purple bean) to other varieties because it sells faster, has low flatulence, cooks faster and is very sweet. 8In Kenya the common bean varieties grown are the large red kidney beans, medium and small reds, the creams, pinto sugars, browns and the purples. Some of the multiple bean varieties grown are AFR708, G2333, KATX56, KATB1, and RWR719 which are of good color, liked by many sellers and are of good taste (Buruchara, 2007). A different study found out that the most commonly grown and marketed varieties in Kenya are GLP-2 (Rosecoco), GLP-24 (Canadian wonder), GLP-x-92 (Mwitemania), GLP-585 (red haricot) and GLP-1004 (Mwezi moja) (Munene and Grisley, 1992; Mbugua and Munene, 1997). KATX56 is the most preferred variety for production because it has a stable yield even in stress conditions and also high yielding in favorable situations. In contrast to Tanzania retailers, the majority of Kenya retailers preferred Nyayo (Korir et al., 2005). According to Grisley and Munene (1992), the Calima (or Rosecoco) has a high market preference in Kenya. According to Korir et al. (2005), Kenya residents in Githuria have a preference for the Red Harricot beans because of the strong red color that blended with maize when cooked together. On the other hand, Mwitemania was preferred by residents in Taita Taveta. The major bean varieties in Uganda include: large seeded red-mottled, the purple, red type and the pale and white colors. The large red-mottled varieties comprise traditional types such as K20, a variety developed by the national research programme in the 1960's (Rubaihayo et al., 1981) and the semi climbers NABE 4 (locally known as Nambale). Other local grain types available in the country are medium size types such as red-medium type (Kayinja) and the brown-red oval (Kanyebwa). In Northern Uganda, there is the small-seeded (locally known as Lango) and are usually black or creamed colored bush bean varieties. In addition, there are some new improved varieties by the national research organization (NARO) and other partners and these varieties have received high market reception especially K131, K132 and NABE 2 (Kalyebara, 2008). K132 and NABE 4 are preferred for their large seed size, desired market qualities, short cooking time and shorter maturity period. 2.2 Common bean trade in East Africa Most recent studies have shown that there exist two types of trade in the East African region that is, the formal and informal trade. Informal cross border trade is more dominant in East Africa due to a number of factors including: traders’ tendency to evade the previously high 9export taxes and import duties, bureaucratic licensing, registration requirements and market failures as a result of poor policies adopted by countries and poor infrastructure (Tchale, 2002). According to Mauyo et al. (2011), the participants involved in the common beans marketing business in Kenya and Uganda are farmers, middlemen, upcountry assemblers, long distance wholesalers, agents, wholesalers based in each country, exporters, retailers and consumers. In Tanzania, those involved in cross border common beans trade are traders, agents, transporters and consumers. The traders consist of wholesalers, retailers and informal hawkers/dealers (Ogutu and Echessah, 1998). The agents buy common beans at the farm gate where they can either sell in small quantities directly to consumers from their stores or sell them to large traders during peak periods. Travelling traders collect sufficient volumes from the agents and are always individual operators whereas the transporters simply transport common beans from large traders into the major urban centers. Wholesalers role in the commons beans trade is to provide large storage facilities to bulk commodities and send them to urban retailers, export houses or cross border traders. On the other hand, retailers directly sell to consumers in small amounts and they are the participants who add the highest margin to common beans. The exporters in common beans are specialized operators involved in large scale procurement, cleaning, grading and rebagging. In Uganda most traders sell beans by weighing in kilograms with a few in 2kg tins while in Kenya most traders sell in 2kg tins. In addition, traders arrive at uniform prices of beans by agreeing among themselves on the prices to sell depending on the demand and the supply of beans. In farmsteads located in areas with limited access to commercial motor vehicles, people use bicycles to transport common beans from farms to rural markets or to the stores of the commission or market agents. When moderate quantities are involved, pick-ups are used with respect to transporting from rural to urban areas while lorries are used to transport large quantities of common beans to final destination especially outside the country. Traders of common beans experience problems of high cost of transportation during rainy seasons due to poor road condition as compared to dry seasons (Mauyo et al., 2010). Tchale (2002) analyzed the informal cross border trade of beans in Eastern and Southern Africa. He concluded that cross border markets provide a niche market that can ensure income and food security to millions of small scale bean producers in the region. In addition, informal 10 movement of bean varieties has implication on the quality of the varieties released by the bean breeding and dissemination programs. Tchale (ibid) also suggests that, informal markets should be encouraged in order to improve the benefits to small scale farmers resulting from increased bean trade in the region. Furthermore, opportunities offered through cross border markets should be explored as these markets provide an important outlet for producers who may not have access to domestic markets due to poor infrastructure and internal marketing inefficiencies. Mauyo et al. (2011) analyzed cross border trade between Kenya and Uganda using both primary and secondary data. Using the concentration model, they indicated that, at the wholesale level, the bean structure was moderately concentrated with competitive fringe in both Uganda and Kenya. At the retail level, the markets in both Kenya and Uganda were well dispersed with a number of traders controlling only small amount of beans in each market. They concluded that, high quality beans fetched higher prices in both Uganda and Kenyan markets. Furthermore, they found out that majority of traders in Uganda sorted their beans and on the contrary, majority of these traders in Kenya did not sort/grade their beans. In addition the study found that, pricing of common beans was determined by the market forces and that uniform prices were arrived at by traders who agreed among themselves on the prices that they were to sell depending on supply and demand of each day. Kenya has been importing beans from Ethiopia, Uganda and Tanzania despite being the seventh largest producer in the world. For the period 2004-2008, the country’s total value of bean imports amounted to about US$ 5.3 million while total value of exports estimated at US$ 624, 000 over the same period. This shows that Kenya imported of beans 10 times what it exported (USAID, 2010). The studies above clearly show that Kenya is the major importer of common beans in East Africa. Katungi et al. (2009), states that some of the factors that contribute to high demand of common beans in Kenya are high population growth rate, weather turbulence and declining to stagnant yields. Therefore to offset this demand, Kenya has to import common beans from neighboring countries. As shown in Table 2, a lot of common beans were destined to Kenya from Uganda and Tanzania through various border points. Table 2: Cross border yearly volumes for common bean for the Year 2012 in metric tons 11 Border Vol (MT) Source Destination Busia 119,332.00 Uganda Kenya Gatuna 2,420.40 Rwanda Uganda Lwakhakha 1,311.92 Uganda Kenya Malaba 2,561.05 Uganda Kenya Mutukula 2,388.00 Tanzania Uganda Namanga 519.35 Tanzania Kenya Oloitoktok 226.00 Tanzania Kenya Source: Regional Agricultural Trade Intelligence Network (RATIN 2011) The findings of EASSI (2012), on the paradox of women cross border traders indicated that, majority of cross border trade; especially the informal is conducted by women who run small businesses. The challenges the women faced included travel security risk, poor infrastructure that increases the cost of transporting goods across the borders, discrimination and harassment by border officials, complexities of cross border trade regulations and lack of market information. 2.3 Market integration analysis Market integration is an important determinant of food flow, availability, accessibility and price stability. The extent to which markets make food available and accessible, and keeps prices stable depends on the degree of market integration across the region (Nyange, 1999). Barrett and Li (2002) consider market integration to be most usefully defined as the tradability or contestability between markets. This implies the transfer of excess demand from one market to another, manifest in the physical flow of the commodity, the transmission of price shocks from one market to another or both. A different study states that market integration is associated with price transmission, which measures the speed of traders’ response in moving food to deficit zones (WFP, 2007). According to Mushtaq, Gafoor and Dad (2008), markets that are not integrated could result in the inefficient allocation of resources. Kose, Prasad and Terrones (2003), state that, markets that are integrated perform better in improving per capita income and standards of living. Market integration is a tool that ensures that regional balance is maintained between food deficit and food surplus regions and that food will always move to where there is high demand 12 and where the prices are also high. Markets that are integrated need to share the same traded good and the same long run information. The information does not have to be processed simultaneously but the markets need to be connected by trade and long run information either directly or indirectly (Gonzalez-Rivera and Helfand, 2001). Over the last few decades, market integration studies have had a wide application. In agriculture for instance, it has been used to determine the overall performance of the market. Various authors have applied different methods and techniques in studying market integration. Correlation and co-integration analysis are the most used but have failed to account for transaction costs and non-linearity (Moser et al., 2009). The correlation analysis of market integration has been used widely. Price series correlation is regarded as convenient indicator of market integration (Basu, 2006). Two variables are correlated if a change in one variable brings about a change in another. They will be perfectly negatively correlated if they have a co-efficient of negative one. They will not be correlated if the correlation is close to zero (Boisseleau and Hewicker, 2002). This method has methodological flaws such as: failure to recognize common exogenous trend, seasonality or autocorrelated and heteroskedastic residuals in the regression with non-stationary price data (Basu, 2006). In addition, correlation cannot account for many real world complexities (Vollrath, 2003). Meyers (2008) studied long run and short run integration of maize in six markets in Malawi using the co-integration approach within the vector autoregressive modeling framework. He used monthly maize retail prices for period January 2000 to May 2008 and found out that nine out of the fifteen market pairs were integrated in the long run and short run market integration was low implying slow price transmission. Mayaka (2013) assessed market integration of dry beans in four markets in Kenya using price data from the periods 1994 to 2011. The study found out that all the four markets were co integrated and the Granger causality test confirmed independent causality with only one market link showing bidirectional causality leading to symmetric price adjustment between Kitale and Nairobi market. However this study only generalized dry beans and did not take into consideration the different varieties. Barrett and Li (2002) incorporated the effect of transaction costs in the co-integration analysis which is considered as a major problem that affects continuous trade and the direction of trade flow. They introduced a mixture distribution model which takes into account transfer costs 13 and trade flows. The model considers the correlation between trade flows; price spread and cost transfers to explain the four potential market conditions that exist: perfect integration, segmented equilibrium, imperfect integration and segmented disequilibrium. Wim et al. (2010) used co- integration analysis and vector error correction model to analyze market integration and utilized wholesale weekly rice prices over a period of between January 2004 to November 2006. They found out that there existed at least three co-integrating vectors implying that, rice markets in Bangladesh during the study period were moderately linked together and that the long run equilibrium was stable. They also found out that the speed of price transmission between the divisional markets were weak. Mumbeya (2011) analyzed value chain and market integration of cassava in the Democratic Republic of Congo. He used co-integration techniques, an error correction mechanism and an index of market connection. The study established that among the eleven pairs of markets, six of them were segmented implying that price changes in the reference markets were fully transmitted to the regional market. Du Preez (2011) analyzed market integration within the potato industry in eight selected markets in South Africa and used weekly data ranging from January 1999 to June 2009. The study determined market integration by applying the threshold vector error correction model and based on the results, there existed long run relation between all the markets and the markets were not integrated in the short run. Studies in East Africa and especially Kenya have concentrated on dry beans in general and have not taken into account the different varieties. This study therefore analyzed market integration of key major bean varieties. 2.4 Price analysis To help judge the extent of efficiency of the marketing system, price movement analysis in varietal commodities in corresponding and linked markets is used. The observed trends in price changes over years in the long run are associated with developments in technology of production input, supply and infrastructure. Time series of prices and quantities can be analyzed using various methodologies ranging from simple graphs, regression or autoregressive integrated moving average (ARIMA) model. Time series has been decomposed by the classical model into different components such as trend (T), cyclic (C), seasonal (S) and random (E) indices (Goetz and Weber, 1986). 14 The variations in market prices can be classified as temporal variation and spatial variation. Temporal variation is as a result of mixtures associated with cyclical, seasonal and irregular components. The seasonal component is the most important. Spatial price variation is the variation in the prices observed in different markets and they occur due to the differences in the location of production and consumption centers. Inter relationship between prices movements in different markets mostly depends on nature and extent of competition, dissemination of market information and the attitude of market functionaries. The degree to which wholesale prices of a commodity in different markets are related to one another determines the efficiency of any marketing system. Kohls and Uhls (1998) are of the opinion that, pricing signals guide and regulate production, consumption and marketing decisions over time, form and place. Price relationships between spatially separated competitive markets depend on the size of transaction cost. As price differences between different markets exceed transaction costs, arbitrage opportunities are created which make profit seeking participants to exploit this opportunity by purchasing commodities from low price surplus market and transferring them to higher price deficit areas. According to Tomek and Robinson (1990), arbitrage opportunities occur only when the deviation in price is substantial enough for potential profit to exceed the cost of trading. This then raises prices in the surplus region and reduces them in the deficit region. Tomek and Robinson (1990) further state that, the principle underlying the differences between regions in a competitive market structure with homogenous commodities is that price differences between any two regional markets that trade with each other should equal transaction cost. In a situation of autarky price differences will be less than or equal to transaction costs. According to Negassa, Myers and Gabre-Madhin (2003), price relationships between spatially separated markets are generally analyzed within the framework of spatial price equilibrium theory developed by Enke (1951), Samuelson (1964) and Takayama and Judge (1964). The key assumption underpinning spatial price equilibrium theory is that price relationships between spatially separated competitive markets depend on the size of transaction costs. Prices are an important tool in the economic analyses of markets (Oladapo and Momoh, 2008). The overall functioning of the market can be better understood by analyzing the vertical or spatial price transmission between markets. The level of market competitiveness can be determined by 15 studying the extent and speed with which shocks in prices are transmitted within the marketing chain (Serra and Goodwin, 2002a). Vollrath (2003) extends the law of one product to international markets stating that, prices will equalize across freely trading areas. In addition homogenous goods will sell for the same prices in different countries taking the exchange rate into account. If two markets are integrated, a shock in any of the markets in either demand or supply and ultimately price should be transmitted to the other market (Barrett, 1996; McNew and Fackler, 1997; Boisseleau and Hewicker, 2002; Negessa et al., 2003; Mushtaq et al., 2008). Markets that are normally integrated exhibit long run relationship between their prices (Balke and Fomby, 1997; Vollrath, 2003; Negassa et al., 2003). In the short run spatial prices can deviate from each other but still be integrated (Vollrath, 2003). 2.5 Theoretical framework Market integration can be vertical, spatial or inter-temporal. Vertical integration involves different stages in marketing and processing channel. In spatial integration, spatially distinct markets have prices that move together and price signals and information are transmitted smoothly. Finally, inter-temporal market integration refers to arbitrage across a period of time. Spatial market integration is the long run relationship of prices. It is the smooth transmission of price signals and information across spatially separated markets (Golletti, Ahmed and Farid, 1995; Ghosh, 2000). The idea of spatial market integration is always expressed as the law of one price. Market integration depends on trade action and its operational environment, which is determined by transportation and communication infrastructure availability and the policies that affect price transmission (Gilletti et al., 1995). Fackler and Godwin (2001), point out that, although majority of authors have focused on whether or not markets exhibit spatial integration, only a number of them have explicitly evaluated the determinants of market integration. . If geographically separated markets are integrated, then there exists an equilibrium relationship (Goodwin and Schroeder 1991 Sexton et al., 1991). Co-integration model in market integration is usually performed to determine whether price of a commodity in a local market is 16 related to change in the central market (Ghoshray, 2009; Ravaillion, 1986). The long run equilibrium relationship for analyzing market integration is as follows: Yt = α + βXt + Ut…………………………….………………………….. ……………… (1) Where; Yt and Xt = commodity prices of a homogenous good (common beans in this case), in two different markets at time t, and α and β are parameters to be estimated. If two markets are perfectly spatially integrated, then β =1. If this holds, then price changes in one market are fully reflected in alternative market. When β ≠ 1 (i.e. β< 1 or β> 1), then the degree of integration may be evaluated by investigating how far the deviation of α1 is from unity. 2.6 Conceptual framework The conceptual framework below shows that an efficient market or integrated market depends on balance in trade which is brought about by improved market infrastructure, government policies and fair prices in the market. If the marketing costs are high then the volume, price, variety and quality of bean that is supplied in the market will be low to meet the rising demand in the market. Government policies which bring about the trading regimes and constraints or challenges such as bribery, long custom procedures and high fee charges at the market that are faced by both the transporters and the traders play a major role in influencing the adequacy of the volumes and quality of bean that is traded between the surplus and deficit market. Rigid trading regimes discourage traders and transporters from trading in large volumes of beans which also impacts on the prices received by consumers and the quality and variety of beans in the market. The characteristics of the bean trade in terms of traders characteristics such as age, gender, education level, distance to the market or border point, years in education and the nature of business all have an impact on the volumes, varieties and prices of common bean that is traded. 17 NB: Arrows show the flow of beans Figure 1: Conceptual framework for factors influencing efficient market. Source: Own conceptualization, 2014 Surplus market 1 Deficit market 1 Efficient market Balance in bean supply and demand Improved market infrastructure (roads and market information demand) Government policy Fair prices 18 CHAPTER THREE METHODOLOGY 3.1 Study Areas This study involved selection of key border points within East Africa. There are around ten border points in Kenya. The study covered four major border points namely Busia, Malaba, Isebania and Namanga: These border points were purposively selected due to the extent of activities, nature of trade and the volumes of common beans that they handled. Busia border is mainly an exit point for goods destined for Uganda from Kenya and vice versa. There is substantial trade that goes on between Kenya and Uganda however; the trade in agricultural commodities is in favor of Uganda. Imports from Uganda are mainly maize, beans, bananas, tomatoes, dried cassava chips, water melon and pineapples. The exports to Uganda are mainly Irish potatoes and cabbages but in small quantities. Trade in common beans mainly takes place between the months of May to August. Malaba border point is mainly for goods on transit, with small imports that pass for local consumption. The main imports through this border are cereals, beans, water melon, bananas and mangoes. Exports are mainly passion fruits and onions which are in small quantities. Isebania border point is located in Kuria district bordering Tanzania. This point has a district agricultural officer stationed permanently here in collaboration with an officer from KEPHIS. Imports from Tanzania mainly include beans, cotton seed cake, rice bran, rice, oranges, tomatoes, green grams, water melons, sweet potato and maize. It handles relatively high amounts of common bean. Namanga is one of the busiest borders between Kenya and Tanzania. It serves as an entry point for commodities destined to the Nairobi market and the transit goods to other countries through the Jomo Kenyatta International Airport (JKIA). It is also the entry point for commodities destined to other parts of the country. It handles commodities such as dry maize, oranges, watermelon, livestock onions and beans. Common beans are traded throughout the year. 19 Table 3: Border points studied Province/border name Neighboring countries Isebania Tanzania Busia Uganda Malaba Uganda Namanga Tanzania The study also involved three selected key markets in Kenya. The criteria for selecting the markets was generally based on their position from the selected border points i.e. whether they are surplus or deficit region and the potential demand for common bean. These markets were purposively selected due to the following reasons: Nakuru is a major cosmopolitan town and a production zone with relatively high demand for common bean. Nairobi is consumption zone with a high real and potential demand for bean and most of the common bean from the border points is destined to Nairobi. Mombasa also a consumption zone and the largest town in the coastal region for both retail and wholesale market for common bean which is served by Namanga border point. 3.2 Respondents Respondents interviewed included traders and transporters of common beans to get information on their characteristics and the constraints they face in the common bean trade. Customs officials at the border points or crop inspectorate officers were also interviewed on the quantity of both formal and informal trade of common bean and the varieties of common bean crossing the borders. 3.3 Sampling procedure and sample size A multistage sampling technique was used in this study to arrive at a sample of 240 traders. The four border points and three key markets were purposively selected. The second step involved snowballing to determine the traders. At least two transporters were interviewed and one Kephis/customs official was interviewed from each border point. 20 In determining the sample size for traders, the formula as developed by Groebner and Shanon (2005) was used as follows:=݊ (^ݖ2 ݌ݍ)/݁^2 ………………………………………………………………… (2) Where: n = sample size p = proportion of the population of interest q = 1-P (the weighting variable) z = standard variate at confidence level ( = 0.05) e = margin of error The area under the normal curve corresponding to 95% confidence interval is 1.96 (z value in the statistical tables). Using a p value of 0.5 and acceptable error of 8.95%, a sample size of 120 was obtained. This was replicated for both the border points and key markets. 3.4 Data Types and Sources This study utilized both primary and secondary data. The secondary data consisted of common bean monthly average prices for the years 2011 to 2013 collected from the Ministry of Agriculture Kilimo House in the Agribusiness and Marketing department. Primary data was collected from the common bean traders and transporters using a structured interview schedule. Data collected included characteristics of traders, different types of common beans traded, prices of common beans, place where beans are bought and sold, frequency of trade per month, mode of transport, challenges traders and transporters face and market information. 3.5 Data analysis Objective one: Characteristics of common bean traders at key border points and markets. Descriptive statistics (mean, frequencies, standard deviation and graphic representation of the results in charts) was used to establish the following parameters; gender, age, level of education, number of years in business and category of trade (transportation, wholesale and retail).  21 Objective two: Determination of constraints to the observed trade. Descriptive statistics were used in the comparisons of the quantity of different common bean varieties traded and the constraints traders faced. This included the use of frequencies, mean, cross tabulation and percentages. Objective three: Assessing extent of market integration for common bean. This analysis provided important information on the product movement mechanisms and technical information on spatial prices behavior. Relationship among prices in key markets was checked using co-integration analysis. A long run linear relationship exists if different price series are co-integrated. If geographically separated markets are integrated, then there exists an equilibrium relationship (Goodwin and Schroeder 1991 Sexton et al., 1991). Co-integration model in market integration is usually performed to determine whether the price of a commodity in a local market is related to change in a central market (Ghoshray, 2009; Ravaillion, 1986). The long run equilibrium relationship for analyzing market integration is as follows: Yt = α+ βXt + Ut…………………………….………………………….. ……………… (3) Where; Yt and Xt = commodity prices of a homogenous good (common beans in this case), in two different markets at time t, and α and β are parameters to be estimated. If two markets are perfectly spatially integrated, then β =1. If this holds, then price changes in one market are fully reflected in alternative market. When β ≠ 1 (i.e. β< 1 or β> 1), then the degree of integration may be evaluated by investigating how far the deviation of α1 is from unity. A two-step model by Engel and Granger (1987) was used since price time series are usually non-stationary and because standard statistical models do not allow explicit determination of α and β. The first step was to determine the order of integration of each price series by checking for stationarity. A time series (say Yt) is stationary if the joint distribution of Yt and Yt + t is independent of time (t). Augmented Dickey-Fuller test was used to determine the order of integration. This was achieved by regressing ∆Yt on Yt-1 and several lags of ∆Yt (enough to avoid auto correlated disturbances). 22 The model was specified as: ∆Yt = α0+ α1 Yt-1 + Σ αk+t ∆Yt+k + ε t……………………………………………………(4) Where: ∆Yt is the first difference of prices in market Y, Yt-1 is the lagged price of common beans in market Y, α0 and α1 are parameters to be estimated, ε t is the error term. The t-statistic on the estimated coefficient of Yt-1 was used to test the hypothesis that: Ho: Yt ~ I(1) Vs H1: Yt ~ I(0) If the null (Ho) above cannot be rejected then Yt cannot be stationary, it can be integrated of order one or even higher. To find out the order of integration the test was repeated with ∆Yt in place of Yt thus regressing ∆∆Yt on a constant ∆Yt-1 and several lags of ∆∆Yt. ADF test will then be used to test the hypothesis that: Ho: ∆Yt ~ I(1) Vs; H1: ∆Yt ~ I(0) i.e Ho: Yt ~ I(2) VS; H1: Yt ~ I(1) This process was continued until the order of integration was established. The second step involved testing for co-integration based on the idea that if two time series (eg. Yt and Xt) are each ~ I (1), then their residual (say Ut) was integrated of order zero (stationary). Where Ut = Yt – α – βXt. The residual (Ut) was then tested for stationarity. The ADF tests applied to these residuals should yield statistics which are large and negative so as to reject the null hypothesis of I (1) in favor of stationarity. If the first step shows that each time series is integrated of order one, and if the second step results to a stationary residual, then the two time series are said to be co-integrated. This implies that long run equilibrium relationship exists between the two sets of prices. To have a distinction between short-run and long-run integration, an Error Correction Model (ECM) was used. This allowed for derivation of the speed of price transmission from one border point/market to another. The error term in the co-integration was treated as the equilibrium error. To tie the short run behavior of Yt to its long run value, the ECM will be specified as: 23 ∆Yt = α0+ α1 ∆Xt+ α2Ut-1 + ε t ………………………………………………………… (5) Where; ∆ = first difference operator, ε t = random error term and Ut-1 = (Yt-1 – α – βXt-1) ECM states that ∆Yt depends on ∆Xt and also on equilibrium error term, while absolute values of α2 decide how quickly equilibrium will be restored (speed of adjustment). 24 CHAPTER FOUR RESULTS AND DISCUSSIONS This chapter presents the findings from the study of common bean cross border trade. The chapter is divided into two sections. The first section presents the descriptive results comprising types of traders, age, gender, experience, level of education of traders, varieties traded and constraints to bean trade. The second section of the chapter discusses empirical results of market integration of the three selected markets in Kenya. 4.1 Descriptive Analysis 4.1.1 Characteristics of common bean traders The results in Table 4 show that the mean age of all sampled traders was 38.97 with the mean age for women and men traders being 40.79 and 36.34 respectively. This shows that female traders were older than the male traders. This could be due to the fact that women began trading in beans at an older age due to family obligations. Some of the women respondents mentioned that they had to take care of the young children until they reached a certain age before they could venture into trade. It was also found that, the mean trading experience for all sampled traders was 7.64 years. Women traders averaged more years in trading than their men counterparts (7.9 years versus 7.2 years for men). As shown in Table 4, majority of women traders had more experience in the bean trade with the maximum experience being 47 years in bean trade and minimum experience being 6 years. The maximum number of years that men stayed in the trade was 25 years and the minimum being a year. Table 4: Age and Experience of traders Aggregate Women Men Age Mean 38.97 40.79 36.34 Std. deviation 0.71 1.01 0.89 Experience Mean 7.64 7.86 7.21 Std. deviation 6.69 7.62 5.16 Min 6.00 1.00 Max 47.00 25.00 Women traders= 141; men= 99 25 Education level The findings showed that more than half (57.58%) of traders had secondary school education. Majority of them felt that it was valuable to be educated to secondary school level as it gave them an advantage on record keeping, credit access, getting information of sources of bean supply and knowing their customer needs over those who did not reach secondary school level (see figure 2 below). Masinjila (2009) also found similar results that majority of traders felt it was advantageous to be educated to secondary school level to engage in meaningful cross border trade. From the results, 28.13% of the traders had attained primary education while those without formal education constituted 7.36%. Those traders who attained post-secondary school education and secondary long cycle accounted for 3.03% each and they outnumbered those traders who had reached adult alphabetization level (0.87%). Figure 2: Overall level education level Figures 3 (a) and (b) show the level of education by gender. These findings reveal that male traders tend to be more educated than women traders. The results show that most traders had attained secondary education in which 64.21% were men while 52.94% were women traders. 30.15% of women traders and 25.26% of men traders achieved primary level education. About 11.03% of the women traders had no formal education. In contrast, 2.11% of men traders had no 0.87% 3.03% 3.03% 7.36% 28.13%57.58% Adult alphabetization Post secondary Secondary school long cycle No formal education Primary school Secondary school -4 years 26 formal education. The relative high level of illiterate women traders is an indication of poor human capital of women which increases their probability to participate in the informal sector as observed by Njikam and Tchouassi (2011). With regard to those traders who had attained post- secondary education, 2.21% were women while 4.21% were men. In addition, the results also show that, 2.21% of women traders and 4.21% of men traders completed six year secondary schooling. . Figure 3: Level of education by gender The selected sample consisted of 41.8% retailers, 39.2% wholesalers, 9.7% both wholesalers and retailers and only 9.3% transporters as shown in Table 5. The reason for most traders being in retail bean trade could be attributed to the fact that other categories of trade require more capital to start and maintain. Odhiambo et al. (2006) noted that lack of capital is a serious constraint for entry into bean trade in Nairobi. 52.94%30.15% 11.03% 2.21% 2.21% 1.47% Sec sch-4yrs Primary sch No formal educ Sec sch-long cycle Post sec sch Adult alphabet 64.21% 25.26% 2.11% 4.21% 4.21% 0 Sec sch-4yrs Primary sch No formal educ Sec sch-long cycle Post sec sch (a)Women (b) Men 27 Table 5: Types of traders Type Frequency Percentage Transporters 23 9.3 Wholesaler 94 39.2 Retailer 99 41.8 Both wholesale and retail 24 9.7 Total 240 100.00 The results in Table 6 show that 59.1% of sampled traders were women while 40.9% were men. This indicates that women are more active in bean trade than men. The findings are consistent with Kibiego et al. (2003) who observed that, majority of common bean traders in Kenyan markets were women who constituted long distance wholesalers, wholesalers/retailers, agents and retailers operating in shops, market stalls and open air. This difference in gender can be attributed to the fact that women traders have defied the belief that trade is a men dominated economic venture. This also goes against the belief that women in most African societies have been perceived as delicate and their duty is to stay at home and take care of the family since men are considered to be bread winners in the family. Currently, more women are participating in trade so as to minimize increasing family costs. Table 6: Category of trade by gender Trade involved in Women Men Total Chi-square Transporters Count 6 16 22 21.8560*** % 27.3 72.7 100 Wholesaler Count 48 45 93 % 51.6 48.4 100 Retailer Count 74 25 99 % 74.8 25.3 100 Both wholesale and retail Count 12 11 23 % 52.2 47.8 100 Total 141 99 240 59.1 40.9 100 28 *** indicates significance at 99 percent confidence level. Of the sampled common bean transporters, 72.7% were men while 27.3% were women. This could be because truck driving or transportation is considered a masculinity venture. From the study, it was also observed that women wholesalers constituted 51.6% while men wholesalers constituted 48.4%. In addition, there were 74.8% women retailers and 25.3% men retailers. Among traders who participated both in retail and wholesale bean trade, women traders comprised 52.2% while men traders were 47.8%. The results in Table 6 also reveal that most women traders dominated in three of the trade categories. 4.1.2 Common bean varieties traded Regarding the common bean varieties among the traders, Nyayo bean was popular and was sold by 35.0% of the traders while 23.5% of the traders sold Wairimu bean variety as shown in Table 7. This could be because these two varieties sell faster in the market due to consumers’ preference and low prices. These findings are similar to that of Korir et al. (2005), who found that farmers preferred to grow the Nyayo variety since it was very marketable at retail level in Kenya. About 15.5% of traders sold Mwitemania variety with 10.2% of traders stocking mixed bean while 8.8% of the traders sold Rosecoco. The least traded bean varieties were Yellow bean and Saitoti which were sold by 4.8% and 2.0% of traders respectively. Nyayo variety was highly traded in the key selected markets and sold by 40.8% of the traders followed by 35.9% of the traders in Busia. In Namanga, 15.5% of traders sold Nyayo bean variety while in Isebania 7.8% of the traders sold the same variety and no trader sold Nyayo in Malaba border point (see Table 7). This could be because Malaba border point is a transit point and minimal trade is carried out at this point. Other than Malaba, 35.7% traders in Namanga, 28.6% traders in Isebania, 21.4% traders in Busia and 14.3% traders in key markets sold yellow bean. Among the sampled traders, 66.7% in Busia sold Saitoti variety while another 33.3% in the key selected markets traded in Saitoti bean. This could probably be because Saitoti bean shares similar features with Nyayo. Saitoti is rather smaller in size than Nyayo making Nyayo the most preferred by traders to sell among the two varieties. Mwitemania variety was sold in the key selected markets, Busia and Namanga by 56.5%, 39.1% and 4.3% of the traders respectively. Of all the bean varieties traded, it was only Wairimu that was sold in all the markets. It was highly traded in the key selected markets (39.1%) and least traded in Malaba 29 (7.3%). This could be attributed to the fact that Wairimu variety is relatively cheaper and preferred by most consumers. There are more than 10 varieties traded in the Kenyan market and the most popular among traders and with a big market share are Nyayo, Wairimu and Rosecoco (Spursby et al., 2004; Katungi et al., 2010). Table 7: Common bean varieties traded at the border points and key markets Common bean varieties Busia Malaba Namanga Isebania Key selected mkts Total Nyayo 37 (35.9) 16 (15.5) 8 (7.8) 42 (40.8) 103 (35.0) Wairimu 13 (18.8) 5 (7.3) 17 (24.6) 7 (10.1) 27 (39.1) 69 (23.5) Mwitemania 18 (39.1) 2 (4.3) 26 (56.5) 46 (15.7) Mixed 10 (33.3) 2 (6.7) 1 (3.3) 17 (56.7) 30 (10.2) Rosecoco 13 (50) 2 (7.7) 3 (11.5) 8 (30.8) 26 (8.8) Yellow bean 3 (21.4) 5 (35.7) 4 (28.6) 2 (14.3) 14 (4.8) Saitoti 4 (66.67) 2 (33.3) 6 (2.04) Numbers in Parenthesis are percentages 4.1.3 Constraints to bean trade From the interviews on constraints faced by the traders, 33.3% of the bean traders indicated that poor roads was a constraint, 32.5% gave high transport cost as a constraint while 21.9% gave the constraint of seasons (heavy rains). The high transport costs were common during the rainy season due to poor road conditions compared to the dry seasons (Mauyo et al., 2010). Traders expressed that high transport costs which is majorly caused by poor road infrastructure constrained them from purchasing beans from source markets. Heavy rains also affected the supply of beans which later translated to low margins for the traders. 30 Table 8: Constraints faced by traders Constraints Transporters Retailers Wholesalers Wholesalers/ retailers Overall percentage Poor roads 25 50 25 0 33.3% High transport cost 9.30 34.9 32.6 23.3 32.5% Seasons (heavy rains) 16.7 47.6 35.7 0.00 21.9% Low profit margins 0.00 70.9 5.9 23.5 19.8% Irregular quantity of supply 0.00 42.9 28.6 28.6 19.0% Lack of access to credit 8.3 58.3 20.8 12.5 16.9% Lack of storage 0.0 20.0 60.0 20.0 7.2% Not convinced that the demand will last for too long 14.3 57.1 28.6 0.0 6.3% Heavy formal taxes (tariffs) 0.0 75.0 25.0 0.0 5.5% Insecurity 0.0 100.0 0.0 0.0 1.7% . Results in Table 8 also show that 19.8% of the traders expressed low profit margins while 19.0% of the traders gave irregular quantity of supply as constraints to bean trade. Some 31 traders expressed that they would sell off their stock of beans at low prices to avoid it from being attacked by pests because at times the demand for beans was low. The issue of irregular quantities of beans forced traders to source beans from alternate sources at very high prices compared to the original source. Some traders mentioned that their source suppliers would hoard beans so as to sell them later at a high price during peak season; this forces most of them to source beans from neighbouring countries like Uganda and Tanzania. Korir et al. (2003) mentioned that long distance wholesalers especially women travelled to neighbouring Tanzania markets to purchase beans. 16.9% of bean traders expressed lack of access to credit as a constraint, 7.2% gave lack of storage, 6.3% gave uncertainty about demand as a constraint while 5.5% gave the constraint of tariffs. Only 1.7% of the traders gave the issue of insecurity as a constraint to bean trade. This finding is similar to Mauyo et al. (2010) who found out that, traders mentioned insecurity as one of their least worries in bean trade. Issues of theft were very rare in bean trade and were experienced by very few traders. This is because the traders had adequate and safe storage facilities for the beans. Other than transporters, all the other traders gave low profit margin, irregular quantities of supply and lack of storage as a constraint. Only retailers were observed to have expressed insecurity as a constraint to bean trade. A greater number of retailers (75%) gave tariffs as a constraint while 25% of the wholesalers gave the same constraint. Other than those traders who were both wholesalers and retailers, all the other traders expressed poor roads, seasons and uncertainty about demand as constraints to bean trade. 4.2 Results of Integration Analysis Secondary data from the years 2011-2013 for Rosecoco, Mwezi moja and Mwitemania were used. The three bean varieties were used because they were common in Nairobi, Nakuru and Mombasa markets that were to be tested for integration. Mbugua and Munene (1997) noted that, Mwitemania, Mwezi moja and Rosecoco are the most commonly grown and marketed varieties in Kenya. 32 4.2.1 Price trends Figure 4 shows wholesale variation in market prices for Rosecoco in Nairobi, Nakuru and Mombasa between the years 2011 to 2013. It can be seen that there has been continuous price fluctuation in the three markets since beans are normally harvested in August and December. Bean prices generally decline immediately after harvest and are at their lowest around August to September and December to February due to bounty supply. The prices then gradually increase and reach their peak around April to July (Katungi et al., 2010). This is because during the months of April to July, beans are still in the field and the demand increases due to minimal supply hence increased prices in the market. Figure 4: Seasonal variation of Rosecoco bean prices in Nairobi, Mombasa and Nakuru It is evident that the wholesale prices have been fluctuating over the three years with the highest average price being in Nairobi especially in the months of June and September followed by Mombasa and then Nakuru with the lowest. This could be because both Nairobi and 4500 5000 5500 6000 6500 7000 Price in KES of Rosecoco per 90kg bag June 2011 Sept 2011 Dec 2011 Mar 2012 June 2012 Sept 2012 Dec 2012 Mar 2013 Date (Monthly) Nairobi Mombasa Nakuru 33 Mombasa are consumption zones for Rosecoco. Nakuru is a production zone for Rosecoco making the prices relatively cheaper as compared to the other two markets. Figure 5 shows wholesale market price trends for Mwezi Moja bean in Nairobi and Nakuru between the years 2011 to 2013. The wholesale prices reached peak in April in both markets. This is because the planting season for beans is in April and so there is less beans in the market which causes supply shortage in the market thus increase in prices. In the month of October, the wholesale prices were low in Nakuru but very high in Nairobi. The difference of prices in these two markets may be explained by the fact that in October beans are probably still in the field and are yet to be harvested. Nairobi is not a production area for beans therefore has to be supplied by other markets thus high prices in the month of October. Figure 5: Seasonal variation of Mwezi moja bean prices in Nairobi and Nakuru Figure 6 shows wholesale market price trends for Mwitemania in Nairobi, Nakuru and Mombasa between the years 2011 to 2013. The trends in the three markets have varied and 4000 4500 5000 5500 6000 Price of Mwezimoja Bean on KES per 90kg bag Jul 2011 Oct 2011 Jan 2012 Apr 2012 Jul 2012 Oct 2012 Jan 2013 Time (monthly) Nairobi Nakuru 34 fluctuated between the three years and this could be due to either bumper harvest or poor harvest caused by variation in weather conditions or resistance and susceptibility to pests and diseases. Mwitemania prices are highest in Mombasa followed by Nairobi and then Nakuru. The monthly average prices reach their peak in the month of April and are slightly low in the month of January possibly because April is a planting season while January is a harvesting season. ` Figure 6: Seasonal variation of Mwitemania bean prices in Nairobi, Mombasa and Nakuru 4.2.2 Unit root test results Table 9 shows unit root test for Rosecoco bean and the first difference of the selected markets. Prices were converted to logarithm and it was found that price series for Nairobi were not stationary as the test statistic (-2.792) is less than the 5 percent critical value. Differenced price series for Nairobi (D1. Nairobi) was stationary as the test statistic (5.317) is greater than the 1 percent critical value (3.75) in absolute terms meaning we reject the null hypothesis that the series are not stationary. Differenced price series for Nakuru and Mombasa were stationary as the test statistics were greater than the 1 percent critical value. This implies that prices of the 3000 4000 5000 6000 7000Price of Mwitemania beans in KES per 90kg bag Jul 2011 Oct 2011 Jan 2012 Apr 2012 Jul 2012 Oct 2012 Jan 2013 Time (monthly) Nairobi Mombasa Nakuru 35 previous period for Rosecoco in the three markets influenced the current prices. This means that the price series for Nairobi, Nakuru and Mombasa Rosecoco bean are integrated of order one I (1). When price series are integrated once, this makes it stationary and integrated of order one process, denoted as I (1) as observed by Odipo et al. (2014). Table 9: Unit Roots Test for Rosecoco Bean Market Test Statistic 1% critical value 5% critical value 10% critical value MacKinnon approximate pvalue for Z(t) Nairobi -2.792 -3.750 -3.000 -2.630 0.0595 D1. Nairobi -5.317 -3.750 -3.000 -2.630 0.0000 Nakuru -2.950 -3.750 -3.000 -2.630 0.0399 D1. Nakuru -4.326 -3.750 -3.000 -2.630 0.0004 Mombasa -2.232 -3.750 -3.000 -2.630 0.1950 D1.Mombasa -5.423 -3.750 -3.000 -2.630 0.0000 Price series for both Nairobi and Nakuru were not stationary as the test statistics of -2.929 and -2.614 respectively were less than the 5 percent critical value meaning previous period prices influenced current prices. The differenced prices for both Nairobi and Nakuru were stationary as depicted in Table 10 meaning that the price series for Mwezi moja for the two markets were integrated of order one. However, contrary to a study by Mayaka (2013), price series for dry beans in Nakuru and Nairobi were integrated of order zero at 1 and 5 percent significance. Table 10: Unit Roots Test for Mwezi moja Bean Market Test Statistic 1% critical value 5% critical value 10% critical value MacKinnon approximate pvalue for Z(t) Nairobi -2.929 -3.750 -3.000 -2.630 0.0421 D1. Nairobi -4.816 -3.750 -3.000 -2.630 0.0001 Nakuru -2.614 -3.750 -3.000 -2.630 0.0902 D1. Nakuru -3.901 -3.750 -3.000 -2.630 0.0020 Table 11 shows that Nairobi and Nakuru Mwitemania bean markets were stationary series. There was no need therefore to test the first differences of the price series to determine the order of integration. This means that previous prices for Mwitemania in both Nairobi and Nakuru did not influence current prices. Price series for Nairobi and Nakuru were integrated of order 36 zero I (0) Mombasa Mwitemania bean market was not stationary as shown in Table 11, the test statistic of -2.015 was less than the 5 percent critical value. The differenced prices for Mombasa were stationary and therefore integrated of order 1 meaning previous prices influenced current prices. Table 11: Unit Roots Test for Mwitemania Bean Market Test Statistic 1% critical value 5% critical value 10% critical value MacKinnon approximate pvalue for Z(t) Nairobi -3.517 -3.750 -3.000 -2.630 0.0076 Nakuru -3.218 -3.750 -3.000 -2.630 0.0190 Mombasa -2.015 -3.750 -3.000 -2.630 0.2799 D1.Mombasa -3.215 -3.750 -3.000 -2.630 0.0191 4.2.3 Co-integration test results After testing for stationarity, the data was then tested for Co-integration. If two markets in spatially separated markets p1t and p2t contain stochastic trends and are integrated of the same order, then the prices are said to be integrated. p1t - b p2t = ut is I(0) ……………………………………. (6) b is referred to as the co-integrating vector. To confirm whether the selected markets are co- integrated, the above relationship was estimated using Ordinary Least Squares OLS (Engle and Granger, 1987). The null hypothesis of no co-integration was tested by applying unit root tests on the residuals (ut) for each of the price series. The hypotheses H0: Ut ~ I (1) was tested against H1: Ut ~ I (0). Results in Table 12 indicate that Nairobi- Mombasa and Nakuru- Mombasa markets for Rosecoco were co-integrated since the test statistics were greater than the critical value at 5 percent. The reason for Nairobi-Mombasa market for Rosecoco being co-integrated could be because of good infrastructure between the two markets. Nakuru-Mombasa markets being co- integrated could be partly because of good market information flow among the two markets. This means we reject the null hypothesis of markets not being co-integrated at α= 0.05. Nakuru- Nairobi market for Rosecoco was not co-integrated and this can be attributed to the fact that Nairobi market could be supplied with Rosecoco from a different production zone other than 37 Nakuru. Beans are established to move from Arusha to Nairobi through Namanga border point and there appears to be a co-movement or trend of average wholesale prices of Rosecoco between Nairobi and Arusha as observed by Korir et al. (2003). Table 12: Cointegration tests between Nairobi and Mombasa and Nakuru for Rosecoco Market Test Statistic 1% critical value 5% critical value 10% critical value MacKinnon approximate pvalue for Z(t) Nairobi vs Mombasa -3.490 -3.750 -3.000 -2.630 0.0083 Nairobi vs Nakuru -2.627 -3.750 -3.000 -2.630 0.0876 Nakuru vs Mombasa -4.735 -3.750 -3.000 -2.630 0.0001 An Error Correction Model (ECM) was used to distinguish between short-run and long- run price relationships amongst Rosecoco markets in Nairobi, Mombasa and Nakuru. ECM was only carried out on the market pairs that were co-integrated. ECM accounts for both short and long run adjustment to disequilibrium in the markets and the time it takes to eliminate disequilibrium as noted by Mushtaq et al. (2008). Once a number of variables were found to be co-integrated, then there existed a corresponding error correction representation which indicated that changes in the dependent variable are a function of the level of disequilibrium in the co- integrating relationship as well as changes in other variables as noted by Engle and Granger (1987). ߂ ௧ܰ= −0.087 + 0.659∆ܯ௧ି ଵ+ 0.724 ௧݁ି ଵ……………………………… (7) Where; ߂= first difference operator, et = random error term Nt is Rosecoco prices in Nairobi while Mt is Rosecoco prices in Mombasa. Equation 7 indicates ECM for Nairobi and Mombasa. The error correction mechanism as indicated by the lagged residual is significant and positive meaning that short run prices diverge away from the long run equilibrium and are unstable. Mombasa prices for Rosecoco appear to have significant long term effect on Nairobi prices. A 1 percent increase in Mombasa prices 38 causes a 0.65 percent rise in Nairobi prices. Besides, Acquah and Owusu (2012) also observed that wrong sign (positive) on the estimate of error correction coefficient of Accra plantain market showed that, short run price movement along the long run equilibrium path may be unstable. ߂ܯ௧= 0.002 + 0.522∆ܰܭ௧ି ଵ+ 0.556 ௧݁ି ଵ……………………………. (8) Equation 8 indicates ECM for Mombasa and Nakuru which shows that short run prices diverge away from long run equilibrium since the coefficient on the error correction term is positive. The positive sign on the coefficient of the error correction term may be an indication of instability in the short run price movements along the long run equilibrium. A 1 percent increase in Nakuru Rosecoco prices causes a 0.52 percent rise in Mombasa prices. Table 13 shows that Nairobi- Nakuru markets for Mwezi Moja were not co-integrated since the test statistic was less than the critical value at 5 percent. This is partly attributed to the fact that Nairobi market could be supplied by neighboring markets other than Nakuru hence the market is already flooded with Mwezi moja bean variety. Table 13: Cointegration test between Nairobi and Nakuru markets for Mwezi Moja. Market Test Statistic 1% critical value 5% critical value 10% critical value MacKinnon approximate pvalue for Z(t) Nairobi vs Nakuru -2.889 -3.750 -3.000 -2.630 0.0467 Nairobi- Nakuru and Nakuru- Mombasa markets for Mwitemania were co-integrated since the test statistics are greater than the critical value at 5 percent as shown in Table 14. This is probably because the marketing infrastructure in these three markets is good therefore low transaction cost. This means we reject the null hypothesis of markets not being integrated at α= 0.05. As also observed by Mayaka (2013), improving market infrastructure such as roads and communication facilities can greatly reduce transaction costs and improve price transmission and market efficiency hence increasing market integration across markets. Mombasa- Nairobi market for Mwitemania was not integrated. 39 Table 14: Cointegration tests between Nairobi, Mombasa and Nakuru markets for Mwitemania. Market Test Statistic 1% critical value 5% critical value 10% critical value MacKinnon approximate pvalue for Z(t) Nairobi vs Nakuru -4.177 -3.750 -3.000 -2.630 0.0007 Nairobi vs Mombasa -2.519 -3.750 -3.000 -2.630 0.1110 Nakuru vs Mombasa -3.689 -3.750 -3.000 -2.630 0.0043 ߂ ௧ܰ= −0.004 + 0.37∆ܰܭ௧ି ଵ− 0.94 ௧݁ି ଵ………………………… (9) Equation 9 indicates that Mwitemania price changes in Nairobi are dependent on price changes in Nakuru. The coefficient of the error correction term is negative and significant showing that short run price movements converge towards long run equilibrium path and may be stable. The results are consistent with Mukhtar and Rasheed (2010) who noted that correct sign (negative) on the coefficient of the error correction term is an indication of stability of the system. Absolute value of α2 (0.94) indicates that when prices are not in equilibrium, they would be restored in a period of less than half a month. This speed at which 94 percent of the disequilibria of the previous month shock is adjusted back to equilibrium is very fast implying high degree of co-integration amongst Nairobi and Nakuru markets for Mwitemania bean. Mwangi et al. (2014) also found similar findings on French beans, where they observed that negative sign on error correction indicated direction of correction towards the long run relationship and that French beans export adjusted rapidly to correct long run disequilibrium in a period of one month. ߂ܯ௧= −0.003 + 0.14∆ܰܭ௧ି ଵ+ 0.32 ௧݁ି ଵ………………………… (10) Equation 10 indicates ECM for Nakuru and Mombasa. The error correction mechanism as indicated by the lagged residual is significant and positive meaning that short run price movement along the long run equilibrium path may be unstable. A 1 percent increase in Nakuru prices causes a 0.14 percent rise in Nairobi prices. 40 CHAPTER FIVE CONCLUSION AND RECOMMENDATIONS 5.1 Conclusions This study established that majority of retailers and wholesalers who play an important role in the bean trade are women. These traders had knowledge on the most preferred varieties in the market by the consumers making them able to supply them at all times by sourcing the beans from their original or alternate source within and outside the country. Traders armed with market information on the prices and bean varieties exploited the market to their advantage. High transport cost was a major constraint for the traders. This was caused by poor roads and heavy rains and this would at times translate to inefficiencies in the supply of beans to the markets. As a result, the demand in the market exceeded the supply given the irregular supply of beans in the market. The key to enhancing integration of common bean markets is to improve the marketing infrastructure. Study findings show that the most traded varieties by the traders were Nyayo and Wairimu while Saitoti and Yellow bean were the least traded by the traders. The results indicated that prices of Rosecoco, Mwitemania and Mwezi moja varieties tended to fluctuate over time due to variation in planting and harvesting seasons. Co-integration and error correction model results revealed Cointegration of Nairobi- Mombasa and Nakuru- Mombasa Rosecoco markets and Nairobi- Nakuru and Nakuru- Mombasa Mwitemania markets. 5.2 Recommendations This study recommends that the government together with the private sector should improve infrastructure such as transport and communication services in order to enhance market integration in Kenya. Improved transport system lowers transaction costs thus farmers will produce more beans and in turn traders purchase the required quantity of beans to meet the market demand without exploiting the consumers. There is need to enhance market information by both farmers and traders freely for better price transmission and knowledge of the best bean markets to trade in. this could boost competitiveness and consequently increase market integration. The government needs to make key commitments to maintain the roads in Kenya in order to lower the transport cost of beans as this was a major hindrance in the efficiency and flow of beans from one market to the other. The government together with the private sector should make efforts at promoting different bean varieties in all the regions to both farmers and traders. 41 This would increase market demand and availability of different bean varieties in the market. Additionally, Kenya needs to maintain good relations with the neighboring countries to penetrate the market in East Africa as there is huge and growing market potential in the region. This would help reduce irregularity in bean supply and thus increase profit margin for the traders. 5.3 Suggestions for further research The study focused on market integration of three common bean varieties in Kenya. However more research should be conducted to document other bean varieties in key markets other than looking at common beans in general to allow a clear distinction between different varieties in the same markets. This would help in formulating market policies that encourage production and supply of diverse varieties in different markets. 42 REFERENCES Ackello-Ogutu, C. (1996). Methodologies for Estimating Informal Cross-Border Trade in Eastern and Southern Africa, USAID Technical Paper No. 29, USAID Africa Bureau, Washington. Ardeni, P.G. (1989). Does the Law of one price really hold for commodity prices? American Journal of Agricultural Economics, 71(3): 661-669. Barrett, C.B. (1996). Market Analysis Methods: Are our Enriched Tool Kits well suited to Enlivened Markets? American Journal of Agricultural Economics, 78: 825-829. Barrett, C.B., and Li, J.R. (2002). Distinguishing between equilibrium and integration in spatial price analysis. American Journal of Agricultural Economics. 84:292-307. Barret, C.B., and Mutambatsere, E. (2005). Agricultural markets in developing countries, In Lawrence E. Blume and Steven N. Durlauf, The New Palgrave, 2nd Edition. Barrett, C.B. (2008). Smallholder market participation: Concepts and Evidence from Eastern and Southern Africa. Food policy, 33(4): 299-317. Basu, J.P. (2006). Co-integration and Market Integration: An application to the Potato Markets in Rural West Bengal, India, Presented at the International Association of Agricultural Economists conference. Gold Coast, Australia. Baulch, B. (1997). Transfer Costs, Spatial Arbitrage, and Testing for Food Market Integration, American Journal of Agricultural Economics, 79: 477-487. Bernstein, H., and Amin. (1995). The Political Economy of Maize filiere. The Journal of Peasant Studies. 23(2/3) 120-145. Boisseleau, F., and Hewicker, C. (2002). European Electricity and its Impact on Market Integration, KEMA Consulting, Germany. Buruchara, R. (2007). Background information on Common Beans (Phaseolus vulgaris L.) in Biotechnology, Breeding and Seed Systems for African Crops. http://www.africancrops.net/rockeffeller/crops/beans/index.htm. 43 Chan, F.T.S. (2003). Performance measurement in a supply chain. International Journal of Advanced Manufactural Technology, 42(13): 534-548. Cramer, G.L., and Jensen, W. (1982). Agricultural Economics and Agribusiness, 2nd Edition. McGraw Hill Book Company, USA. 222p. De Toni, A., and Tonchia, S. (2001). Performance measurement systems models, characteristics and measures. International Journal of Operations and Production Management, 21(1/2): 46-70. Du Preez, L. (2011). A Study on the Integration of Potato Markets in South Africa. Thesis. University of the Free State, Bloemfontein. Engle, R.F., and Granger, C.W.J. (1987). Co-integration and Error Correction: Representation and Testing, Econometrica, 55: 251-276. Fackler, P.L., and Tastan, H. (2008). Estimating the Degree of Market Integration, American Journal of Agricultural Economics, 87(3): 717-734. Fafchamps, M., and Hill, R. V. (2005). Selling at farmgate or travelling to market. American Journal of Agricultural Economics, 87(3): 717. 734. FAO, (2012). Statistical database of the Food and Agriculture Organization of the United Nations, Rome, Italy. http://faostat.fao.org FAOSTAT Food and Agriculture Organization at www.fao.org Ferris, S., and E. Kaganzi (2008). Evaluating marketing opportunities for haricot beans in Ethiopia. Rural agro-enterprise development project report. Fulgence, J.M., Anna, A.T., and Joan. F. (2009). Drivers of Common Bean Trade in Tanzania: A marketing Perspective. Staff paper #09-02. Ghosh, M. (2000). Cointegration Tests and Spatial Integration of Rice Markets in India, India Journal of Agricultural Economics, 55, No. 4, October- December, Pp. 616-625. 44 Ghoshray, A. (2009). Price dynamics for different qualities of coffee: http://rmi.saeprint.com/sgi/content/refs/111/113. Goodwin, B.K., and Schroeder T.C. (1991). Co-integration Test and Spatial Price Linkages in Regional Cattle Markets. American Journal of Agricultural Economics. 73(2): 452-64. Goetz, S., and Weber M.T. Fundamentals of price Analysis in Developing Countries’ Food Systems: Training manual to accompany the Computer Software Program MSTAT. Michigan State University International Development Paper. Working Paper No. 29, (1986). Goletti, F., Ahmed, R., and Farid, N. (1995). Structural Determinants of Market Integration: The case of Rice in Bangladesh, The Development Economics, XXXIII-2. Grisley, W., and Mwesigwa, D. (1991a). Consumer preferences for fresh and dry beans in Uganda: implication for bean breeders. CIAT, Kampala, Uganda. Grisley, W., and Munene, S. (1992). Dry beans sold at retail markets in Kenya: cultivars, grain types, prices and sources. CIAT, Kampala, Uganda. Groebner, D.F., and Shannon, P.W. (2005). Business Statistics. A decision Making Approach. Pentice Hall. Harris, B. (1993).There is a Method in my Madness: or is it Vice Versa? Measuring Agricultural Marketing Performance. John Abbott(ed). Agricultural and Food Marketing in Developing Countries, UK. ISAR, (2011). Bean Program, Rwanda Agricultural Research Institute. http://www.isar.rw/spip. Article45Nations. Katungi, E., Farrow, A., Chianu, J., Sperling, L., and Beebe, S. (2009). Common bean in Eastern and Southern Africa: a situation and outlook analysis. CIAT, Kampala, Uganda. http://www.icrisat.org/what-we-do/impi/projects/t12-publications/regional-situation- outlookreports/rso-common-bean-esa.pdf. 45 Kibiego, M.B., Odhiambo, M.O., and Kimani, P.M. (2003). Analysis of bean marketing system in urban areas of Kenya. African Crop Science Society, 6: 587-590. Kohls, R.L., and Uhl J.N. (1998). Marketing of agricultural products. Fifth Edition, Macmillan, New York. Korir, M.K. (2005). Cross Border Bean Marketing Between the Northern Zone of Tanzania and Nairobi, Kenya. M.Phil. Thesis. Moi University. Eldoret. Kose, M.A., Prasad, E.S., and Terrones, M.E. (2003). Financial Integration and Macroeconomic Volatility, IMF Staff Papers, 50: 119-142. Kotler, P. (1997). Marketing Management Analysis, Planning, Implementation, and Control, 9th Edition. Upper Saddle River, NJ: Prentice Hall International. Kweka, S.O., Ndakidemi, P.A., Mushi, C.S., Nkonya, E., and David, S. (1998). Adoption of Lyamungu 85 Bean Variety in the Medium-Altitude Zone of Tanzania. Magrath, P. (1992). Methodologies for studying Agricultural marketing in developing Countries. Marketing series 2, Chart ham, UK: Natural resource Institute. Masinjila, M. (2009). Gender Dimensions of Cross Border Trade in the East African Community- Kenya/Uganda and Rwanda Burundi Border. Mauyo, L.W., Chianu, J.N., Nassiuma, B.K., and Musebe, R.O. (2010). Cross-Border Bean Market Performance in Western Kenya and Eastern Uganda: http://www.SciRP.org/journal/jssm. Mauyo, L.W., Webukhulu, R., Musebe, R.O., Kirkby, R.A., and Buruchara, R. (2011). Structure and Conduct of Cross-Border Bean (Phaseolus Vulgaris) Marketing in East Africa: The case of Western Kenya and Eastern Uganda. Journal of Emerging Trends in Economics and Management Sciences, 2(6): 478-482. Mayaka, K. V. (2013). An Assessment of Dry Beans Market Integration in Selected Markets in Kenya. Thesis. Egerton University, Njoro. 46 Mbugua, G., and Munene, S. (1997). Determination of Dry Bean Varieties grown by farmers in Kenya’s major bean producing areas. A paper presented at the First East and Central Africa Bean Research Network (ECABREN). McNair, M.P., and Harry, L. (1956). Readings in Marketing, 2nd –ed. McGraw-Hill Company, Inc. New York. McNew, K. (1996). Spatial market integration: definition, theory and evidence. Agricultural and Resource Economic Review, 25: 1-11. Mendoza, G. (1995). A Primer on marketing channels and margins. Lyme Rimer Publishers. Meyers, J. R. (2008). The Effectiveness of Spatial Maize Price Transmission in Malawi. Department of Agriculture Food and Resource Economics. Michigan State University. East Lansing, MI 48906. Minde, I., and T.O. Nakhumwa. (1996). Unrecorded cross-border trade between Malawi and her neighbours: implications on food security. Technoserve Inc. Ministry of Agriculture (MoA). (2008). Strategic Plan. Government Printers, Nairobi. Minten, B., and Kyle, S. (1999). The effect of distance and road quality on food collection, marketing margins, and traders’ wages: Evidence from the former Zaire. Journal of Development Economics, 60 (1999): 467-495. Moser, C., Barrett, C., and Minten, B. (2009). Spatial Integration at Multiple Scales: Rice Markets in Madagascar, Agricultural Economics, 40: 281-294. Mukhtar, T., and Rasheed, S. (2010). Testing Long run Relationship between Exports and Imports: Evidence from Pakistan. Journal of Economic Cooperation and Development, 31(1): 41-58. Mushtaq, K., Gafoor, A., and Dad, M. (2008). Apple Market Integration: Implications for Sustainable Agricultural Development, The Lahore Journal of Economics, 13: 129-138. 47 Mwangi, C. S., Mbatia, E.L.O., and Nzuma, M. J. (2014). Effects of Export Rate Volatility on French Beans Exports in Kenya. Journal of Agricultural Economics, Extension and Rural Development, (1)1:001-012. Mwita , J. K., and Audi, P.O. (1981). The farming system of lowland Machakos District, Kenya. Report of farm survey results from Mwala location. Technical report No.1. Negassa, A.,Myers, R., and Gabre-Madhin, E. (2003). Analyzing the Grain Market Efficiency in Developing Countries: Review of Existing Methods and Extensions to the Parity Bound Model, Market trade and institutions division, Discussion paper, No. 63. Njikam ,O., and Tchouassi, G. (2011). Women in Informal Cross-border Trade: Evidence from Cameroon. International Journal of Economics and Finance, 3(3): 202-213 Nyange, A.D. (1999). Estimation of inter-regional maize market integration in Tanzania and its determinants. http://www.foodnet.cgiar.org/post%20harvest/papers/market%20integration.David%20N yange.htm. Odipo, O.T., Bett, K. H., Lagat, K. J., and Sigei, K. G. (2014). Analysis of Market Integration: A Case of Sugar in Selected Markets in Kenya. Journal of Economics and Sustainable Development, 5(2): 2222-1700. Oladapo, M.O., and Momoh, S. (2008). Price transmission and market integration in Oyo state, Nigeria. Journal compilation, African Development Bank, London: Blackwell publishing. Pomeroy, R. S., and Trinidad A.C. (1995). Industrial Organization of Market Analysis: Fish marketing. Prices, products and people: Analyzing Agricultural Markets in Developing Countries. International Potato Center. RATIN- Regional Agricultural trade Intelligance Network (2011). Food Situation analysis in East Africa. In service of the Eastern African Grain Council. RATIN- Regional Agricultural trade Intelligance Network (2012). Food Situation analysis in East Africa. In service of the Eastern African Grain Council. 48 Ravallion, M. (1986). “Testing Market Integration.” American Journal of Agricultural Economics, 68 (1986): 88-109. Samuelson, P. (1964). Spatial equilibrium and linear programming. The American Economic Review, 42(3): 283-303. Scarborough, V., and Kydd, J., (1992). Economic analysis of Agricultural Markets. A manual marketing series 5, chartham UK: Natural resource. Serra, T., and Goodwin, B.K. (2002a). price Transmission and Asymmetric Adjustment in the Spanish Dairy Sector. Sexton, R.J., Kling C.L., and Carman H.F. (1991). Market Integration, Efficiency of Arbitrage, and Imperfect Competition: Methodology and Application to U.S. Celery. American Journal of Agricultural Economics, 73(3): 568-580. Singh, S.P. (1999). Common Bean Improvement for the Twenty-First Century. Kluwer Acad. Publ., Dordrecht, Germ. Soniia, D., and Sperling, L. (1999). Improving technology delivery mechanisms: Lessons from bean seed systems research in eastern and central Africa. Agriculture and Human Values 16:381-388. Spilsbury, J., Jagwe. J., and Wanda, K. (2004). Evaluating the Marketing Opportunities for beans and its products in the principle beans growing countries of ASARECA, draft regional report compiled by International Institute of Tropical Agriculture and Food net. Stern, L.W., Ansary, A.I.E., and Coughlan, A.T. (1996). Marketing Channels, 5th Edition, Upper Saddle River, NJ: Prentice Hall International. Tchale, H. (2002). Informal cross-border trade: The case of beans in Eastern and Southern Africa-implications on bean seed multiplication, dissemination entrepreneurship and quality. In proceedings of a bean seed workshop held January 12-14 at Arusha, Tanzania, by bean/cowpea collaborative research support program-East Africa. 49 UBOS (Uganda Bureau of Statistics). (2007), Uganda National Household Survey 2005/2006, Agricultural Module. 214pp. Website: www.ubos.org USAID (United States Agency for International Development), (2010). Staple Food Value Chain Analysis in Kenya. Chemonics International Inc. USAID (United States Agency for International Development), (1995) “Policy Determination No. 19” Bureau for Program and Policy Co-ordination, USAID, Washington DC. Van Campenhout, B. (2007). Modeling Trends in Food Market Integration; Method and an Application to Tanzanian Maize Markets, Food policy, 32: 112-127. Van Rheenen, H.A. (1979). Diversity of food beans in Kenya. Econ Bot 33:448-454. Vollrath, T.L. (2003). North American Agricultural Market Integration and its Impact on the Food and Fiber System, Agricultural Information Bulletin No. 784, Washington, DC. Voysest, O., and Dessert M. (1991).Bean cultivars, classes and commercial seed types. A; Voysest O, eds. Common beans: research for crop improvement. CAB International and CIAT, Wallingford, UK. Pp. 119-162. WFP/PDPE, (2007). Market analysis tool: Market integration. Working paper 7, 2007. World Bank 2009. Wim, V., and Mohammad, I. H. (2010). Evaluation of Rice Markets Integration in Bangladesh. Lahore Journal of Economics 26: 77-96 Wortmann, C.S., Kirkby R.A., Eledu C.A. and Allen D.J. (1998). Atlas of common bean (Phaseolus vulgaris L) in Africa. Cali, Colom. (Publication No. 297). Xavery, P. R., Kalyebara, C., and F. Ngulu (2005). The impact of improved bean varieties in Northern Tanzania. Selian Agricultural Research Institute (SARI) Tanzania in collaboration with the Pan-African Bean Research alliance (PABRA) and the International Center for Tropical Agriculture (CIAT). 46 APPENDICES APPENDIX ONE: INTERVIEW SCHEDULE FOR TRADERS AND TRANSPORTERS Traders and Transporters/Couriers Questionnaire Questionnaire Number__________________________ Question Codes Response A: Background Information 1 Interviewer name 2 Country 1=Kenya, 2=Uganda, 3=Tanzania, 4= Burundi, 5=Rwanda 3 Border site 4 Border Town 5 Name of respondent 6 Gender of the respondent 1= Male, 0= Female 7 Age of the respondent (In complete years ) 8 What is your highest level of education? 9 Are you a resident of this border area? 1= Yes, 0= No 10 If no, where do you live? 11 Country of origin of the respondent 1=Kenya, 2=Uganda, 3=Tanzania, 4= Burundi, 5=Rwanda 12 What is the main category of trade you are involved in? 1=transportation, 2= whole sale, 3= retailer, 4= Other specify………………… 13 Who is your customer? If there are multiple 1= End consumer 47 customers, please rank them in order of importance 2= Retailer 3= Middleman/aggregator (actually purchases) 4= Wholesaler 5= Broker/fixer (only trades) 6= Transporter who wholesales 7= Processor/Miller 8= Any other, please specify 14 Number of years in the started business 15 Compared to five years ago, how did the number of customers, change? (Only ask if business is more than five years old 1= Increased 2= Decreased 3= No Change 16 Do you participate in informal cross border trade 1= Yes, 0= No 17 Percentage of the trade 18 Is informal cross border trade your main source of employment? 19 Give reasons why you participate in informal cross border trade? B: Volumes and values of beans trade 48 20 Local name Trad e in it [1= Yes; 0= No] Loca l Nam es Sourc e (coun try and locali ty name ) Alternat e source of supply (country and locality name) What types of units do you buy at? 1 = Kilos, 2 = 50 kilo bags, 3 = 90 kilo bag, 4 = 110 kilo bags, 5 = tons, 6= big tins/debes Any other, please specify Purchas ing quantit y per month Purch ase price Sell ing pric e Time taken to sell Place to where they are sold (country and locality name) Packa ging Quanti ty declar ed (%) 49 21 Indicate the month when the beans are available in the market, when buying price, selling price and trade volumes are high, average or low. Codes: 1= high, 2= average, 3= low, 4= not available Bean local names Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec General availability Buying price 50 Volume of sales Sales prices Availability in usual source General availability Buying price Volume of sales Sales prices Availability in usual source General availability Buying price Volume of sales Sales prices Availability in usual source General availability Buying price Volume of sales Sales prices Availability in usual source 51 General availability Buying price Volume of sales Sales prices Availability in usual source General availability Buying price Volume of sales Sales prices Availability in usual source General availability Buying price Volume of sales Sales prices Availability in usual source General availability Buying price Volume of sales Sales prices 52 Availability in usual source 22 Are the bean varieties normally mixed 1=Yes 2=N0 23 If mixed do you sort them or market as they are 1=sort 2= minimal sorting 3=No sorting 24 What are the reasons for sorting (e.g what market requires, differential price) 25 Mode of transport 1= 7 ton truck (Cantor), 2= 12 ton truck , 3= 28 ton truck, 4= Land- cruiser taxis, 5= Pick-ups (Hilux), 6= Vans (owned or rented), 7= Matatus/buses/public transport, 8= Carts (donkeys/cows), 9= Cycles, 10= On animals (camels/cows/donkeys), 11= Wheelbarrows, 12= Humans, 13= Any other please specify 26 What are the 3 main constraints to transporting commodities to you/your clients? Please rank them in order of importance Use codes in Q 24 27 Do you have access to market information? 1= Yes, 0= No 28 If, yes, what kind of Market Information do you get? 1= On price, 2= On volume of the commodity, 3= On Quality of the commodity, 4= On Transport costs, 5= On Storage price, 6= On what the competition is doing, 7= Any other, please specify 29 Main constraints and challenges in bean trade 1= Seasons (heavy rains, etc), 2= Lack of access to credit, 3= Irregular quantities of supply by wholesalers/suppliers/producers, 4= Lack of storage, 5= High transport costs, 6= Lack of good roads, 7= Low profit 53 margins, 8= Heavy informal taxes (bribes), 9= Distances from wholesale/supply markets, 10= Heavy formal taxes (tariffs), 11= Insecurity, 12= Poor road infrastructure, 13= Not convinced that demand will last too long, 14= Any other, please specify 30 In your view, what actions should be taken by Government to address the above problems 31 In your view, what actions should be undertaken by the Private sector to address the above problems 32 Compared to five years ago, how does the volumes of trade changed? (Only ask if business is more than five years old 1= Increased, 2= Decreased, 3= No Change 33 How have the trading regimes affected bean trade in the past five years? 1= Increased, 2= Decreased, 3= No Change 34 How have customs documents changed compared to 5 years ago? 54 35 How have fees and charges at the border points customs documents changed compared to 5 years ago? 36 Are there new costs and charges introduced Are there new costs and charges introduced What amount of costs do you incur in this trade?  Handling,  Packaging  Transport  Storage and rental cost  Labour cost of work paid by traders  Insurance expenses/costs Closure-Thank you very much for your time 37 Do you have any question? 55 APPENDIX TWO: INTERVIEW SCHEDULE FOR CUSTOMS OFFICIALS Formal and Informal Exports and Imports Direction of trade: from……………………….to…………………………… Customs Station ……………………………………. Beans Types (Photo) Month Amount declared (tones/month ) Value Tax paid Quality declared % Share of not declared Destinations Informal Trade as % of total trade (officer estimate) 56