CENTRALIZED ALLOCATION OF FLEXIBLE MEASURES: A NON-RADIAL DIRECTIONAL DISTANCE FUNCTION DEA APPROACH

In the traditional data envelopment analysis (DEA), the classification of inputs and outputs is certain. However, in some real-world applications, the role of performance measures as inputs or outputs is uncertain and flexible. Moreover, resource and revenue management is crucial and is controlled by a central policy-maker in various processes. Therefore, this paper proposes DEA-based methods to deal with resource and revenue allocation problems with flexible measures. One of the models introduced does not include uncontrollable variables, and another incorporates them. Then the proposed approach is extended to address resource and revenue allocation plans when negative and controllable flexible measures are present. To illustrate, non-radial directional distance models are introduced to allocate centralized resources and revenues where flexible measures are considered. Numerical examples and an application of the banking industry are used to clarify and demonstrate the proposed methods. Based on the findings, the provided techniques are reliable and useful to address centralized allocation processes with flexible measures.