A new interval efficiency measure in data envelopment analysis based on efficiency potential

This study proposes a new approach for evaluating the efficiency of decision-making units (DMUs) with interval data. We define a potential-based measure of efficiency (PBM), analogous with the concepts of the vector field and the potential function in Physics that projects the DMUs onto the best and the worst frontiers defined by the best and the worst values of the inputs and outputs of the DMUs. This contrasts with the standard and arguably unfair approach that benchmarks the best performance of a DMU against the worst performance of all other DMUs or the worst performance of a DMU against the best performance of all other DMUs. We argue that it is more sensible to assume that all the DMUs perform at their best or at their worst. Then, in this way, a narrower efficiency interval is determined for each DMU. We show how to derive a numerical efficiency, and hence, how to rank DMUs from the efficiency interval. Our methodology is illustrated using a dataset that relates to the performance of the branches of an insurance company in Iran. We compare our methodology with other existing methods from the literature.

Automated summary

This study proposes a new approach for evaluating the efficiency of decision-making units with interval data, introducing a potential-based measure of efficiency. The methodology is demonstrated using a dataset related to the performance of branches of an insurance company in Iran and compared with existing methods.