Evaluating efficiency in nonhomogeneous environments

Abstract

The conventional DEA methodology is generally designed to evaluate the relative efficiencies of a set of comparable decision-making units (DMUs). An appropriate setting is one where all DMUs use the same inputs, produce the same outputs, experience the same operating conditions, and generally operate in similar environments. In many applications, however, it can occur that the DMUs fall into different groups or categories, where the efficiency scores for any given group may be significantly different from those of another group. Examples include sets of hospitals with different patient mixes, groups of bank branches with differing customer demographics, manufacturing plants where some have been upgraded or modernized and others not, and so on. In such settings, if one wishes to evaluate an entire set of DMUs as a single group, this necessitates modifying the DEA structure such as to make allowance for what one might deem different environmental conditions or simply inherent inequities. Such a modification is presented herein and is illustrated using a particular example involving business activities in Mexico. While we do carry out a detailed analysis of these businesses, it is important to emphasize that this paper’s principal contribution is the methodology, not the particular application to which the methodology is applied.