Identifying suitable benchmarks in the way toward achieving targets using data envelopment analysis

Abstract

It was earlier claimed by the authors that the data envelopment analysis (DEA) models commonly used only implement objectives concerned with target setting, while peer identification, which is a key issue for the benchmarking, is actually a by-product. For this reason, they developed some bi-objective DEA benchmarking models that pay special attention to the selection of benchmarks among peers, in addition to setting the closest targets. Following an analogous approach, we here continue that research and propose new models that provide a response to the question of whom to emulate to achieve the targets that have been set for a given organization. Thus, the proposed approach now aims to relate the setting of targets and the identification of peers, specifically identifying peers that may serve as benchmarks for improving performance in the direction established by the targets set. As a result, the models identify the so-called leading benchmark, which is the one whose performance is most aligned with the targets. Eventually, a series of targets and peers are generated by varying the importance attached to the two objectives considered, which offer decision-makers alternatives for the selection of a course of action when planning improvements.