Measures of inefficiency in data envelopment analysis and stochastic frontier estimation

This paper discusses recent work in developing scalar measures of inefficiency which (a) comprehend all inefficiencies, including non-zero slacks, and (b) are readily interpretable and easily used in a wide variety of contexts. The opening section of the paper discusses some of the varied contexts in which uses of DEA are now being reported. This provides background for some of these measures. The closing section turns to simulation studies of DEA–regression combinations and possible inefficiency measures. Serious problems of bias in Stochastic Frontier (SF) regression approaches are identified. Extensions and modifications are suggested which can make a development of other inefficiency measures worthwhile for SF extensions to input-specific and multiple output evaluations.