Determinants of technical efficiency of cocoa production in Liberia. A case of Lofa and Nimba Counties, Liberia

Abstract

This study presents the use of a stochastic frontier production function to examine the Technical efficiency of cocoa production in Liberia. The study draws on data from a field survey administered to 303 smallholder cocoa farmers in four major cocoa farming districts in Lofa and Nimba Counties in Liberia. They are Voinjama, Salayea, Twah River and Bull Yao. The districts were purposively selected because they are major cocoa growing area in Liberia and consist of 15-20 percent predominant activities. Productive efficiency was analyzed using stochastic frontier analysis with a Cobb-Douglas production function while assuming a truncated-normal distribution for the inefficiency term. The output variable was total quantity of Cocoa produced, while input variables were quantity or value of inputs used in the production process, namely labor, farm size, amount of pesticide applied in liter, amount of fertilizer applied in kg and age of cocoa bush. Based on the maximum likelihood stochastic result, the value of lambda (λ) indicates that there is a 140% variation in output due to technical inefficiency. This led to rejection the second alternative hypotheses of this study and maintained that smallholder cocoa farmers in Lofa and Nimba are not technically efficient. The mean age of cocoa bush and cost of labour were significantly and negatively contributing to frontier output. Cocoa farmer’s education, age of cocoa bush, type of cocoa, household size, and employed elsewhere significantly affected and reduced technical efficiency in the study areas and efficiency was relatively higher among smallholder farmer that were member of associations.