Data envelopment analysis models based on decentralized decision making

Information technologies, such as blockchain, have promoted the development of decentralized decision-making in complex systems and generated many decentralized autonomous organizations. We construct a set of decentralized Data Envelopment Analysis (DEA) models based on the basic decentralized decision mode of voting majority, and we consider the assumptions that DMUs have no prior information about other DMUs' decisions, and that they do. Then, the process of decision-making for each decision making unit (DMU) can be viewed as a social behavior over multiple time periods. Using a numerical example, we demonstrate that the new DEA models can provide acceptable efficiency results with specific weights for each DMU and can describe the details of the consensus relationship among DMUs. Similar to the market mechanism in reality, the efficiency scores of DMUs within an alliance containing more than half of the participants are overestimated, at the expense of nonallied DMUs. The models in this paper provide a possible path for efficiency measurement based on decentralized decision-making principles in complex systems.

Automated summary

This paper constructs a set of decentralized DEA models based on decentralized decision-making principles for efficiency measurement in complex systems. The research shows that these models can provide acceptable efficiency results for each decision making unit and describe the consensus relationship among units.