Inputs and outputs classification in integer-valued data envelopment analysis

In conventional data envelopment analysis (DEA) models, data are usually considered as continuous measures while the role of measures from input and/or output aspects is known. However, there are situations in the real world that a measure can play either input or output roles, a flexible measure, while it can only take integer values (e.g., nurse trainees and medical interns in assessing the performance of hospitals). In the current paper, integer DEA models with flexible measures from two individual DMUs and aggregate efficiency perspectives are proposed to determine the relative efficiency of units and the role of flexible measures. Indeed, slacks are addressed and a slacks-based measure (SBM) of efficiency is defined for calculating the performance of decision making units (DMUs) in the presence of flexible and integer factors. The majority decision rule is also used to specify the role of flexible measures from the individual DMU point of view. Two examples including banks and hospitals are utilized to clarify and demonstrate the suggested approach. (C) 2019 Elsevier Ltd. All rights reserved.