DEA model for partial centralization resource allocation among independent subset of DMUs

Abstract

This paper proposes the concept of partial centralization in view of the problem of partition and group governance in management practice, and based on this, an optimization model of independent subset incentive resource allocation is constructed. From the perspective of the decision makers of each independent subset, the model takes into account the different attributes of resources and incorporates the homogeneous decision-making units outside the subset into the reference set to allocate resources based on the principle of efficiency first and fairness. The purpose is to improve the output of each decision-making unit in the subset and reduce the myopia effect by rationally optimizing the allocation of resources. The constructed model can directly compute resource optimization results without model retransformation. Numerical examples show that by applying this model, the efficiency value of each decision-making unit after optimizing resource allocation is greater than or equal to the original efficiency value. Finally, it is applied to analyze the optimal allocation of resources of 30 banks in Taiwan. The results show that this method can better realize the optimal allocation of bank resources and stimulate the resource allocation of non-effective decision-making units, which has important application significance.