DEA and MCDA: Competing or Complementary Approaches?

Abstract

A number of authors have highlighted the similarities of the DEA and MCDA models, commenting principally on the mathematical structure and methods for solution. In this paper, however, we focus on the problematique — the nature of problems tackled and process of analysis. DEA could be described as an approach which seeks to extract as much as possible from “objective”, historical data, without resort to subjectivity. In contrast, MCDA actively seeks to elicit, understand and manage value judgements. It has been suggested that DEA is an appropriate tool for monitoring and control, whereas MCDA is most appropriate in the context of evaluation and choice. However, there are many applications which cross these boundaries and there is increasing attention in the field of DEA to the desirability of incorporating value judgements in some analyses. Thus, we ask the question: “Should DEA and MCDA be viewed as competing approaches?” We suggest that, on the contrary, there are many ways in which the two approaches can be used complementarily and that each can learn important lessons from the other.