An analysis of Stochastic Maize production functions in Kenya

Abstract

In Kenya, agriculture governs the country’s fiscal economy, and this reliance on agriculture can cause both economic and hunger problems, a result of the country’s dependence upon rainfall for agricultural production. Kenyans must find ways to combat severe drought conditions; this can be accomplished through the adoption of inputs that decrease the probability of crop failure. The objective of this research is to determine whether variability exists in Kenyan maize yields, and whether or not specific inputs, specifically hybrid varieties, are either variance/skewness increasing or decreasing. The data used for this study was collected from a survey, designed by Egerton University’s Tegemeo Institute of Agricultural Policy and Development and Michigan State University, and administered in Kenya in the following years: 1997, 2000, 2004, and 2007. The survey identified factors of crop and field level production, such as inputs, crop mix, marketing data, and demographic information. This research makes use of only the 2007 data, comprising 1,397 households in total. The objectives of this thesis aim to go beyond the scope of typical production function regressions where yield is a function of a set of inputs, by examining further moments of yield, variance, and skewness to determine whether variability exists in Kenyan maize yields. Results indicate that variability does exist within Kenyan maize yields, often a result of differing input levels among households. In terms of overall impact of each variable on mean, variance, and skewness of maize yields, seed quantity, nitrogen use, and hybrid seed contribute the most to influencing these factors. In contrast, years of experience with hybrid maize, land tenure, terraced land and labor have the least influence on mean, variance and skewness within this research. Results also bring to light the popular debate against hybrid varieties versus open pollinated (OPV) or traditional varieties, and identify hybrid varieties as a source of variability in mean, variance and skewness of yields. Hybrid varieties should be paired with the knowledge of how to maximize yield in conjunction with other inputs, to give Kenya the opportunity to see substantial productivity gains throughout the country, especially in arid and semi-arid regions.