Efficiency estimation and its role in policy recommendations: An application to the Kenyan livestock sector

Abstract

Livestock represents 12% of Kenya's economy, and an increase in its efficiency can improve development and reduce poverty. A parametric stochastic frontier model was fitted to a large farm-level cross-sectional data set to estimate the efficiency levels and the underlying sources of the output's departure from the optimal levels. We found that the overall mean efficiency in most of the production frontier models is fairly high (above 51%) and sensitive to the model choice. The main factors that significantly reduce the inefficiency of livestock production are the length of education, gender and age of the household head, offfarm income, land ownership, access to market information and number of livestock technologies adopted. The possible policy options at the disposal of policy-makers are to increase the investment in infrastructural facilities, education and women's and youths' empowerment through capacity building and resource allocation.

Estimation of stochastic production functions: the state of the art

“…To date, the SFA has become the framework of choice of many scholars (e.g. Coelli [1995], Jondrow et al [1982], Kumbhakar-Tsionas-Sipiläinen [2009], Schmidt [2011], Mamardashvili-Bokusheva [2014], Baráth-Fertő [2015], Martinez et al [2016], Bahta et al [2018], Manyeki-Kotosz [2019]) in the estimation of TE levels for economic agents.…”

Microeconometric models of the livestock sector

“…Following these observations, it is expected that livestock farming households who have access to veterinary services have health sound animals and also widens the household's knowledge concerning the use of improved livestock production technologies and this is likely to influence market participation decision of a household positively. Additional, veterinary services and advisory services (either through extension or agricultural research) as observed in Manyeki-Kotosz (2019) study may lead to more technical inefficiency.…”

“…This study follows Rakipova et al (2003) approach and the dependent variable Y stand for tropical livestock unit (TLU) 26 and comprises of cattle representing large ruminant, and sheep and goat representing small-stock ruminate. A similar approach was also applied in the study by Lekunze-Luvhengo (2016) and Manyeki-Kotosz (2019) where the number of livestock produced per year were converted to TLUs which is equivalent to physical weights approach applied in Rakipova et al (2003) study. Different livestock species were combined into one herd, and this was necessitated because, under pastoral set-up, the three species (i.e.…”

“…A stochastic frontier latent class model (Equation 7) was applied using an extensive household farmer-level data, and the results are then compared with a model which assumes that the technology is common to all farmers. Our empirical analysis was only based on the flexible Translog functional form in order to avoid imposing unnecessary a priori restrictions on the technologies to be estimated (Orea-Kumbhakar 2004, Alvarez-del Corral, 2010, and also, from our recent study where both functional forms were tested, the flexible Translog functional forms were found to be adequate representation of the dataset (Manyeki-Kotosz 2019). Further, it also allows the relevance of input interactions when explaining production.…”

“…One of the biggest challenges to the pastoral household involvement in livestock marketing in Kenya can be associated with the nature and quantity of household factor endowment such as size of pastureland and livestock tropical units at farmer disposal (Manyeki-Kotosz 2019). For instance, the size of the pastureland is important because transaction costs are primarily fixed costs that can be spread across more output on large farms.…”