Ranking DEA Efficient Units with the Most Compromising Common Weights

One may employ Data Envelopment Analysis (DEA) to discriminate decision-making units (DMUs) into efficient and inefficient ones base upon the multiple inputs and output performance indices. In this paper we consider that there is a centralized decision maker (DM) who ‘owns’ or ‘supervises’ all the DMUs. In such intraorganizational scenario the DM has an interest in discriminating the efficient DMUs (eDMUs). This paper presents a new method that determines the most compromising set of weights for the indices’. The total of the new efficiency scores of eDMUs with the most compromising set of indices’ weights has the least total gaps to the compromised datum. The eDMUs that have efficiency score equal to one are located on the datum. The other eDMUs are either located above or below the datum. The approach is analog to the ordinary least-square method (OLS) of the residuals in statistical regression analysis. We compare the results of an example with multiple inputs and single output under the proposed approach and regression analysis.