| Summary: | In this paper, empirical maximum likelihood (ML) and weighted least squares (WLS) estimates of socio-economic factors influencing the demand for private veterinary services in the high potential agricultural areas of Kenya are derived using probit analysis. While ML estimates explain 65 and 56 percent of the variation in demand for artificial insemination and clinical services, respectively, WLS estimates explain 74 and 72 percent of the variation after correcting for sample selection bias using the Heckman two-step procedure. Based on the ML and WLS coefficients, herd size, number of breeding cows farm income, education, knowledge of husbandry practices and time devoted to farm activities are identified as important determinants of demand and are used to construct producer demand profiles. Using the predicitive success rate together with information on the profiles, the model correctly classifies the demand for artificial insemination and clinical services to be higher than demand for vaccination and herd health services.
|
|---|