This paper introduces a novel method to incorporate categorical non-discretionary variables in Data Envelopment Analysis (DEA) models. While solutions to this problem have been introduced before, they have rarely been employed in applied work. We surmise that existing solution concepts pose...