Centralized resource allocation and target setting of a two‐stage production process using data envelopment analysis

Abstract Centralized resource allocation and target setting (CRATS) refers to a process in which top management has the power and responsibility to allocate limited input resources and to arrange expected output targets among its own decision‐making units (DMUs). Data envelopment analysis (DEA) has been widely used to investigate the CRATS issues by considering the production process of DMUs as a single‐stage structure, namely, a black box. This study proposes a novel centralized two‐stage network DEA because the DMUs with a two‐stage production process are a common phenomenon in real‐world cases. The proposed model is applied to determine the optimal CRATS plan for an empirical case, where 27 retailers belong to an automobile parts supplier in Taiwan.