A Comparison of Stochastic Frontier Approaches to Estimating Inefficiency and Total Factor Productivity: An Application to Irish Dairy Farming

This paper compares standard stochastic frontier models for panel data with a number of recently developed models designed to remove unobserved heterogeneity from the inefficiency component. Results are used to construct a generalised Malmquist total factor productivity (TFP) index. We conclude that the choice of approach makes little difference where the purpose of the study is to analyse aggregate trends in TFP and its components. However, where inefficiency estimates and their dispersion are of interest, attention should be paid to how the analyst’s interpretation of inefficiency relates to the underlying assumptions of the model that is used.