A Value Efficiency-Based Target Setting Approach in Data Envelopment Analysis

Abstract

Basic models of Data Envelopment Analysis are intrinsically preference-free, in the sense that they consider all inputs and outputs and also all decision making units of the same importance. Although this property is beneficial in many ways, it has some drawbacks simultaneously, as the decision makers' preferences are not taken into account in the process of evaluating units. To overcome this drawback many researchers have developed several techniques for incorporating the preferences into the evaluation model. One of the underlying approaches is value efficiency analysis, which evaluates decision making units by comparing them with the most preferred unit. The most preferred unit is a unit which satisfies the decision maker most. On the other hand, the issue of benchmarking is an important aspect in data envelopment analysis, as it enables the analyst to choose a target for each inefficient unit. The target unit for each unit is located on the efficient frontier and determines the path of improvement for that inefficient unit. In this paper, the issue of target setting based on the concept of value efficiency is investigated. We aim to develop a target setting model which is able to determine target units that are not only efficient but also value efficient, as well. Moreover, the targets are determined based on closest distance from the evaluated unit. Some properties of the model are also discussed. Finally, we perform the proposed model on a real data set of 42 Spanish Universities.