Variants of the slacks-based measure with assurance region and zeros in input-output data

TL;DR

This paper enhances the slacks-based measure (SBM) by integrating assurance regions and a closer target setting approach, enabling efficient computation and handling zeros in data without transformations.

Contribution

It introduces a hybrid SBM with assurance region that maintains desirable properties and can process zeros in data directly.

Findings

Efficiency scores computed via linear programming.

Maintains monotonicity and other desirable properties.

Handles zeros in data without transformations.

Abstract

Incorporating an assurance region (AR) into the slacks-based measure (SBM) improves practicality; however, its efficiency measure may not have desirable properties, such as monotonicity. We incorporate a closer target setting approach into the SBM with AR and a variant of the SBM with AR. We demonstrate that the efficiency measure with the hybrid approach has the same desirable properties as that without AR, and we also show that the efficiency scores can be computed by solving linear programming problems. Our proposed approach can handle zeros in the observed input-output data without any data transformation or additional model modification.