Assessment of the Structure and Performance of the Milk Value Chain in Western Kenya

Abstract

Western Kenya Counties of Busia, Bungoma, Kakamega and Vihiga experience persistent milk deficit and low commercialization on smallholder farms. Information and knowledge on sources and points of inefficiency along the dairy value chain that contribute to low milk production and hence persistent milk deficit are scanty. The main objective of this study was to assess the structure and performance of the milk value chain in the region to determine market opportunities, variables influencing milk production, sources of inefficiencies and an upgrading strategy. An explanatory research design was used to collect both quantitative and qualitative data from 385 consumers and seven institutions in the four Counties and;400 dairy farmers, six focus groups and twelve key informants in Butula and Butere Sub Counties using a stratified proportional random sampling technique. End market tool kit and' descriptive statistics; Household Commercialization Index, Pearson's correlation and multiple linear regression were used to analyse data on markets, farm assessment and upgrading strategy respectively. The findings on markets revealed that the region's milk deficit was about 177 million litres per year with demand estimated to be 392 million litres per year against an annual production of 215 million litres. Quality and price were found to be the most important attributes significantly influencing choice of milk supplier (p<O.OO1). There was low milk supply in cooperatives with only 8.5% out of the installed cooler capacity of about 27600 litres utilized. Compared to households, hotels and institutions, cooperatives when restructured into business entities, were found to be the best milk market in an upgrading strategy. The level of commercialization in the region was found to be 39%, with low input household index of 32%. The mean milk yield per cow per day was 6.5 litres, with89% of the farms producing less than 10 litres of milk per cow per day, while yield per cow per year was 1240 litres. There was a positive and significant linear relationship between eight variables and milk production. The most important predictors explaining variations in milk production were: Fodder, dairy meal, credit, Al services, improved research technologies, group membership, policy and returns from milk sales. Collectively, they explained 63.9% of the variance, out of which 51% was explained by fodder and dairy meal combined. The model was highly significant at 1% (Fg, 291= 65.089, p<O.OOl). The main sources of inefficiencies were found to be institutional. These findings suggest a linear relationship between value chain components and milk production and hence may be useful in selecting: a) important variables b) prioritization c) estimation of impact before actual implementation d) key stakeholders in an intervention A potential model for improving the structure and performance of the milk value chain should consist of five levels: integrated input supply; dairy farmer business clusters, new structured cooperatives, contracted retailers and an enabling County government policy. The model appears feasible and could increase average yields to 10 litres per cow per day, increase annual income per cow from KES 62,000 to KES 126.500, reduce milk deficit by about 51%, with a benefit-cost of KES 14.14 for every KES 1 invested. The model has practical significance and could be adopted by the County Governments in Western Kenya to address the milk deficit problem. Further research should be carried out to validate the model.