Merging Decision-Making Units Using a Non-Oriented DEA-Based Approach

So far, several data envelopment analysis (DEA)-based approaches have been introduced to merge decision-making units (DMUs). However, these approaches rely on input- and output-oriented DEA models, which may not effectively determine the appropriate levels of inputs and outputs for the merged DMU. To address this problem, this study proposes a non-oriented DEA approach for merging candidate DMUs. In the proposed approach, a virtual DMU is first introduced, and its position within the technology is analyzed to determine the appropriate levels of inputs and outputs for the merged DMU, ensuring it achieves its desired efficiency score. Compared to existing approaches, the proposed method enables the merged DMU to utilize fewer resources while producing more outputs. The advantages of this approach are demonstrated through managerial and economic insights, as well as four comparative and empirical examples. Proofs of the theorems are also included in the appendices.