A modified super-efficiency measure based on simultaneous input-output projection in data envelopment analysis

Super-efficiency data envelopment analysis (SE-DEA) models have been developed and applied in many situations. However, under the condition of variable returns to scale (VRS), infeasibility of the SE-DEA model may occur and restrict its application. A modified SE-DEA measure based on simultaneous input-output projection is proposed as a way to systematically characterize the super-efficiency in both inputs and outputs. The modified measure overcomes the infeasibility problem while providing ease of computation and interpretation. The practicability of the proposed measure in real applications and its comparison to other super-efficiency measures are illustrated empirically using an example.