Adoption of grade cattle technology in Kenya: A combined farm-level and spatial approach.

Abstract

The dissertation aims at identifying the determinants of the adoption of grade cattle technology in the specific case of the Kenyan smallholders. Adoption of high grade cows by smallholders is driven by the objective of increased milk production, for both home consumption and sale. Smallholders are believed to have a comparative advantage in rearing grade cows, but constraints to adoption are numerous: the cost of a grade cow is relatively high, and the dairy enterprise is risky. Risks include animal diseases and lack of reliable marketing outlets. Marketing risks are a common preoccupation for smallholders but it is particularly relevant for milk, which is bulky, highly perishable and needs to be sold daily. The main constraint to adoption is considered to be the entry cost and farmers have several ways to finance it. The author participated actively to the collection of survey data in several areas of Kenya that represent a broad range of levels of dairy productivity potential and market access. Two main analyses of the decision to rear grade cows were conducted, both theoretically and empirically. The first approach is static and analyses the decision at the time of the survey. The second approach uses a dynamic and spatial framework. GIS-derived distances are computed and introduced in a duration model in order to control for market access. Time is expected to play a key role in adoption and two time dimensions are introduced: an idiosyncratic time describing the conditions faced by the household at the beginning of the spell and historical time accounting for the changes in the external conditions. Results show that poor access to credit cannot be excluded as a reason for delaying adoption of grade cows. Policy changes over time are also found to play a role in the adoption process, as the reduced availability of reliable market channels and livestock services after liberalisation in 1992 are shown to have shifted down the adoption function.