A fairer assessment of DMUs in a generalised two-stage DEA structure

In Data Envelopment Analysis (DEA), a variety of approaches have been used in the context single-stage and basic serial two-stage systems to attain fairness evaluation decision-making units (DMUs). Little work, however, has done address this challenge generalised structure featuring additional inputs second stage proportion first-stage outputs as final outputs. paper, we argue that context, is enhanced by increasing measures related discriminatory power weighting scheme method. We describe mechanism gives prominence more contemporary concept fairness, incorporating diversity inclusion minority opinions. These aspects have, our knowledge, not yet received explicit attention methodological development DEA. propose novel combination an additive self-efficiency aggregation model, minimax secondary goal CRiteria Importance Through Inter-criteria Correlation (CRITIC) method, order promote these thus achieve better degree cooperation between stages DMU among DMUs. The model chosen over alternative multiplicative approach for reasons relating emphasis on intermediate products exchanged simplification. offers peer which each aims evaluate worst others best possible light. Application CRITIC method DEA addresses problem within cross-efficiency concept. Practical applications could include supporting determination training needs job rotation manufacturing, or sustainable supply chains. paper includes description numerical experiment, illustrating approach.