Study on feasibility of variable returns-to-scale based super-efficiency data envelopment analysis model
Wenhua Li · Journal of systems engineering · 2003
A modified data envelopment analysis model -super-efficiency DEA can make a sufficient sortation on all the DMUs (decision making units). But under condition of variable returns_to_scale (VRS) it may be infeasible. Super efficient DMU and strongly super efficient DMU are firstly defined. Then the feasibility of SE-DEA (VRS) model is studied. We prove that SE-DEA (VRS) model for a certain DMU is infeasible if and only if this DMU is super efficient. Based on this, a method of using SE-DEA (VRS) model to evaluate DMUs is put forward. By this means we can rank all the DMUs from the highest to the lowest as: strongly super efficient, super efficient, efficient and inefficient. Finally a case of industries' economic evaluation is illustrated.