Non-discretionary Factors in Data Envelopment Analysis: Review and Extension

Abstract

Data Envelopment Analysis (DEA) technique uses linear programming to evaluate the relative efficiency of a homogeneous set of Decision Making Units (DMUs) in their use of multiple inputs to produce multiple outputs. The standard DEA models do not take into account non-discretionary inputs and outputs and ignore the possibility that efficiency may be correlated with the non-discretionary factors. However, one key issue in performance measurement problems is how to treat non-discretionary factors, which influence the performance of DMUs and are, at the same time, out of the control of the management. In this paper, a new model for measuring efficiency is defined such that non-discretionary factors are taken into account by the decision maker. The main contributions of this paper are fourfold: (1) we review the existing approaches for measuring efficiency scores to control non-discretionary factors in production; (2) we provide a discussion of strengths and weaknesses and highlighting potential limitations of the existing non-discretionary DEA models; (3) we propose a new approach based on relative importance of non-discretionary inputs that overcomes existing weaknesses; (4) we use a numerical example to demonstrate the feasibility and richness of the obtained solutions.