The Effect of Crop Specialization on Farms' Performance: A Bayesian Non-neutral Stochastic Frontier Approach

A Bayesian non-neutral stochastic input distance function model is used to examine whether output specialization has an impact on the economic performance of vegetable producers in Benin. Specialization is assumed to have an effect on the production frontier itself, as well as on the distance of each producer's observed data to this frontier (technical efficiency). A derivative-based measure of economies of scope is obtained by exploiting the duality between the shadow cost and the input distance functions. In this study, we also control for spatial heterogeneity of vegetable production by including a soil fertility variable in the production function at the farm level. The technology is found to exhibit no economies of scope, indicating that vegetable producers have no incentive for specialization or diversification. However, the degree of specialization has a positive effect on technical efficiency. From a policy perspective, the findings imply that policies to encourage specialization may lead to higher performance.