Make your own free website on

Prioritizing Benchmark Bank Branches: the Concept of Representativeness

Wade D. Cook and Moez Hababou,

York University, 4700 Keele Street, North York, M3J1P3, Ontario


This paper presents an approach to prioritizing efficient Decision-Making Units (DMU) in Data Envelopment Analysis with specific reference to bank branches. This approach is based on representativeness. Each inefficient branch is projected against a set of one or more efficient units. If one considers that inefficient units are not equally important to the decision maker, it is arguable that representing an inefficient unit belonging to a specific cluster is of a different degree of importance than representing a unit belonging to another one. The methodology proposed consists of four steps: (1) identify the candidate set of efficient units, (2) cluster all units, using clustering factors defined and weighted by management, (3) prioritize clusters using a hierarchical structure; for this, we propose a variant of the AHP model as a mechanism for weighting clusters; (4) rank efficient units using an optimization model. An illustrative application concerning bank branches is provided.

Key words: representativeness, efficient units, data envelopment analysis, ranking, peers


If you wish to receive a copy of this paper, e-mail

Back to the CIBC project