Abstract

The study investigates the performance of the traditional and improved fish smoking kilns in the Lake Chilwa basin using a stochastic production frontier. The kiln performance is measured through its efficiency in utilising inputs (firewood and labour) for an output (Smoked fish). Since fish smoking depends on the forest resource, the study also endeavours to estimate the extent of deforestation that would be avoided by using a more efficient kiln type over a specific period of time. The study uses primary data that was collected from fish smokers through field experiments and semi-structured questionnaires. The study shows that transcendental logarithmic (Translog) and Cobb-Douglas stochastic frontiers best represent data captured from traditional and improved kilns respectively. In addition, the study indicates no evidence of technical inefficiencies in both traditional and improved fish smoking kiln models. This implies that production functions for both traditional and improved fish smoking kilns have normal errors. The study further reveals that there is no significant difference in the mean technical efficiency levels between traditional and improved fish smoking kilns. The empirical results indicate that there is no difference in the amount of firewood used to smoke a given quantity of fresh fish between traditional and improved kilns. Consequently, there is no evidence of deforestation avoided by using either of the smoking methods. The study has however showed that traditional kilns use more labour than improved kilns to smoke a given quantity of fresh fish. Consequently, use of the improved kilns can divert human resources to other productive ventures.

More details

School	: School of Law, Economics and Government
Issued Date	: 2012

Download full document