Flexible measure in the presence of the partial input to output impacts process

Precise recognition of the nonparametric measurement approach in the production process and proper application of accurate techniques to categorise the variables play a key role in the process of improving performance of decision-making units (DMUs). The classical data envelopment analysis (DEA) models require that the status of all inputs and outputs measures be precisely specified in advance. However, there are situations where a performance measure can play input role for some DMUs and output role for the others. This paper introduces an approach to determine the situation of such flexibility where in the presence of resource sharing among subunits, the partial input will impact output in DEA. As a result, DMUs have a fair evaluation when compared to each other. Likewise, the maximum improvement is obtained in aggregate efficiency due to partial input to output impacts. The proposed approach is applied to a set of real data collected from 30 branches of an Iranian bank.