Two-stage additive integer-valued data envelopment analysis models: A case of Iranian power industry

Abstract

Purpose

The purpose of this paper is to propose a modified model in multi-stage processes when there are intermediate measures between the stages and in this sense, the new efficiency scores are more accurate. Conventional data envelopment analysis (DEA) models disregard the internal structures of peer decision-making units (DMUs) in evaluating their relative efficiency. Such an approach would cause managers to lose important DMU information. Therefore, in multistage processes, traditional DEA models encounter problems when intermediate measures are used for efficiency evaluation.

Design/methodology/approach

In this study, two-stage additive integer-valued DEA models were proposed. Three models were proposed for measuring inefficiency slacks in each stage and in the system as a whole.

Findings

Three models were proposed for measuring inefficiency slacks in each stage and in the system as a whole.

Originality/value

The advantage of the proposed models for multi-stage systems is that they can accurately determine the stages with the greatest weaknesses/strengths. By introducing an applied case in the Iranian power industry, the paper demonstrated the applications and advantages of the proposed models.

Citation

Ajirlo, S.F., Amirteimoori, A. and Kordrostami, S. (2019), "Two-stage additive integer-valued data envelopment analysis models: A case of Iranian power industry", Journal of Modelling in Management, Vol. 14 No. 1, pp. 199-213. https://doi.org/10.1108/JM2-12-2017-0132