SIMULATING WEIGHTS RESTRICTIONS IN DATA ENVELOPMENT ANALYSIS BY MEANSOF UNOBSERVED DMUS

Data envelopment analysis (DEA) is a method for assessing the comparat ive efficiencies of decision making units (e.g., banks and schools) by relating their output to their input levels, Restrictions are often i mposed in these assessments to reflect prior judgments on the values o f input and/or output variables. This paper introduces a new approach to capturing and using value judgments in DEA, based on unobserved Dec ision Making Units. In so doing, it opens up a whole new approach for reflecting value judgments in DEA assessments, which can offer advanta ges in certain situations.