DEA is a well-established performance benchmarking instrument; and many applications as well as theoretical developments have emerged since the famous contribution of Charnes et al. (1978). There are substantially two reasons for such notable interest. First, DEA is useful for calculating the relative efficiency of multiple input-output activities of a certain set of decision-making units. Second, and most interestingly, it is a non-parametric tool that can be applied without knowing the exact production function - this stands in sharp contrast to classical mathematical-statistical approaches.