Compensation analysis with additive DEA model

Abstract

Purpose

To develop a method of compensation analysis based on the additive data envelopment analysis model.

Design/methodology/approach

It is a common assumption in current target setting literature that outputs of decision making units (DMUs) are generated simultaneously. In practice, outputs of DMUs may be generated consecutively, and targets of posterior outputs may need to be adjusted to keep a DMU efficient when the anterior output cannot reach the previously set output target. The proposed method determines a compensation rate among consecutive outputs, which is applied to keep a DMU efficient when the decrease of an anterior output exceeds its stability region. This compensation analysis is also extended to consecutive inputs.

Findings

Sufficient and necessary conditions for preserving efficiency upon output compensation are presented.

Research limitations/implications

The compensation analysis when multiple anterior outputs decrease from their previously set targets is the main limitation, and needs further investigation.

Practical implications

The paper is a useful method for decision maker to set target consecutively.

Originality/value

The paper defines a new target setting (adjusting) problem, in which targets of posterior outputs may need to be adjusted to keep a DMU efficient if an anterior output of the DMU decreases from its previously set target. Theoretical analyses of the compensation analysis are provided.

Citation

He, P. and Hua, Z. (2008), "Compensation analysis with additive DEA model", Kybernetes, Vol. 37 No. 9/10, pp. 1331-1338. https://doi.org/10.1108/03684920810907625