A simple MILP to determine closest targets in non-oriented DEA model satisfying strong monotonicity. (September 2018)

• Record Type: Journal Article
• Authors: Zhu, Qingyuan; Wu, Jie; Ji, Xiang; Li, Feng
• Is Part Of: Omega. Volume 79(2018)
• Page Start: 1
• Page End: 8
• Publication Date: 2018-09
• Subjects: Data envelopment analysis -- Monotonicity -- Closest targets; Management -- Periodicals; 658.4005
• DOI: 10.1016/j.omega.2017.07.003
• Languages: English
• ISSNs: 0305-0483

Abstract

Highlights: The previous literature about closest targets satisfying (strong) monotonicity is mostly based on complicated multi-stage procedure. A simple MILP is proposed to find the closest targets that satisfies strong monotonicity. The proposed approaches are applied to industrial production processes of China. Abstract: The benchmarking information targets based on Data envelopment analysis (DEA) are commonly classified into two groups: the farthest targets and the closest targets. Many methods have been introduced to derive the closest projection to the strongly efficienct frontier, most of them merely based on the computation of closest efficient targets. Although there are several measures that satisfy a set of interesting properties from an economic and mathematical point of view, they are based on output-oriented models with a multi-stage procedure and cannot be extended easily to the situation of non-orientation, or they are mostly based on a conceptual framework which could not be solved easily. In this paper, we provide an easy, well-defined efficiency measure based on non-oriented closest targets that satisfies strong monotonicity and that is calculated by a simple mixed integer linear program (MILP). A real case of industrial production processes of 30 provincial level regions in mainland China was analyzed to verify the applicability of our proposed approach.