The fixed weighting nature of a cross-evaluation model

Cross-evaluation has been touted as a powerful extension of Data Envelopment Analysis that provides, not only a unique ordering among the Decision Making Units (DMUs), but also eliminates unrealistic weighting schemes without requiring the elicitation of weight restrictions from application area experts. The goal of this paper is to prove, in the single-input, multiple-output case, cross-evaluation implicitly uses a single fixed set of weights. We demonstrate how this unseen fixed set of weights may still be unrealistic.