Application of a stochastic production frontier to slave agriculture: an extension

Abstract

Free, small slave and large slave farms in the antebellum American South are compared using a stochastic production frontier approach, which allows a distinction to be drawn between technological superiority and technical efficiency. Previous studies using this approach had only considered slave farms. In addition, cotton share is included as an independent variable in the production function in order to control for variations in crop mix across farm types that might make slave farms look more efficient. I find that the gang system did make large slave farms technologically superior, but that without the gang system, slave agriculture was not superior. In contrast to previous findings, large slave farms were not less technically efficient than small slave farms. Overall, slave farms were more efficient than free farms.