Abstract
This paper considers the objective function of the Additive (Pareto–Koopmans) Data Envelopment Analysis model and finds a flaw which is induced by the manner in which units invariance is sought. This flaw is rectified, in the spirit of the original proposal, but the consequence is a non-linear objective function for the efficiency calculations. However, the resulting non-linear programming problems solve readily with existing NLP software.
Keywords: