Resource Allocation of a Parallel System with Interaction Consideration Using a DEA Approach: an Application to Chinese Input-Output Table.

Beibei Xiong,Jie Wu,Qingxian An,Junfei Chu,Liang
DOI: https://doi.org/10.1080/03155986.2017.1335046
2017-01-01
Abstract:Resource allocation is a popular and important issue in the enterprise management. Recently, data envelopment analysis (DEA) as a non-parametric method for measuring the performance of decision-making units (DMUs) has brought a new flavour to this issue. However, most of resource allocation works by DEA focused on single stage system or consider the internal production process of the system as a 'black box.' With the competition and relation among economic entities enhance, the system becomes more and more complex and interactive. To go inside the 'black box', in this paper, we propose a new DEA approach to allocate the resource in a bidirectional interactive parallel system. We consider not only the resource allocation of a certain DMU, but also the resource allocation of all DMUs for a centralized decision maker through centralized models. Moreover, the leader-follower relationship between two subunits is studied by a non-cooperative model as a theoretical extension. Finally, the approach is applied to Chinese input-output table in the cooperation scenario. We compare our approach with the traditional approach and find that it can obtain more potential gains.
What problem does this paper attempt to address?