Stochastic Frontier Models in Sports Economics

Young Hoon Lee

This paper is intended to introduce a variety of stochastic frontier econometric models to applied economists working in the field of sports economics. First, it discusses the characteristics and assumptions of individual stochastic frontier models that should be used in empirical studies. For example, it distinguishes “preferred” from “not-preferred” models based on the size of the panel data sample or the research purposes. Second, it discusses the characteristics of the sports industry and differentiates it from manufacturing industry, which is necessary when using stochastic frontier models that have been developed for the study of general industries such as manufacturing. The third purpose of the paper is to introduce ready-made program codes for various stochastic frontier models. Information about the available FRONTIER4.1, STATA commands and GAUSS codes may help widen the domain from which empirical sports economists can select their regression models for use in efficiency studies.