Munich Personal RePEc Archive

A universal solution for units-invariance in data envelopment analysis

Xu, Jin and Zervopoulos, Panagiotis and Qian, Zhenhua and Cheng, Gang (2012): A universal solution for units-invariance in data envelopment analysis.

[img]
Preview
PDF
MPRA_paper_41633.pdf

Download (466kB) | Preview

Abstract

The directional distance function model is a generalization of the radial model in data envelopment analysis (DEA). The directional distance function model is appropriate for dealing with cases where undesirable outputs exist. However, it is not a units-invariant measure of efficiency, which limits its accuracy. In this paper, we develop a data normalization method for DEA, which is a universal solution for the problem of units-invariance in DEA. The efficiency scores remain unchanged when the original data are replaced with the normalized data in the existing units-invariant DEA models, including the radial and slack-based measure models, i.e., the data normalization method is compatible with the radial and slack-based measure models. Based on normalized data, a units-invariant efficiency measure for the directional distance function model is defined.

UB_LMU-Logo
MPRA is a RePEc service hosted by
the Munich University Library in Germany.