Measuring decentralized network efficiency: A pre-mortem Stackelberg game approach

In this paper, we present the 'pre-mortem Stackelberg game approach' as a non-cooperative game model that can serve as a viable alternative to the conventional decentralized network DEA model, which suffers from the infeasibility problem. We identified the circumstances that give rise to the issue and recommended the use of non-radial slacks as a remedy. The rationale behind this method is provided in relation to relevant game theories as a modeling justification. To validate the effectiveness of the proposed model, we conducted an empirical study on the performance evaluation of local agents of a Korean life insurance company. Our findings show that the proposed model not only provides an alternative efficiency measure for the follower but also enables an optimal adjustment of intermediate products, for which feasible solutions do not exist under the conventional decentralized network DEA model. Moreover, the proposed model delivers efficiency measures that are equivalent to those of the conventional model when feasible solutions do exist. The proposed model is expected to facilitate the use of non-cooperative game in network DEA, particularly when the VRS assumption is appropriate for performance evaluation.