Incorporating quality into data envelopment analysis: A stochastic dominance approach

The study is concerned with the incorporation of information on differences in quality into DEA. Three approaches for incorporation of quality in efficiency evaluations and corresponding criteria for efficiency are suggested. The three criteria are operationalized for use in empirical analyses in a number of LP-models in the DEA-tradition and the links between the models and the microeconomic production theory are established. A relationship between dominance criteria designed for an evaluation of quality and quantity performance and the notion of weak disposability is shown to exist within the context of estimation of production possibility sets. Preliminary results of an efficiency analysis of public primary schools in Denmark based upon the suggested framework for quality-controlled efficiency evaluations are reported.