Variations effect of intermediate products on the second stage in two-stage processes

摘要

Network data envelopment analysis (NDEA) deals with evaluating the performance of a set of homogeneous production processes (or decision making units (DMUs)) taking into account the internal structure of DMUs. A large number of studies in NDEA are based on two-stage structures with intermediate products. For a two-stage DMU, if the intermediate products increase by increasing the inputs of the first stage, then increasing of the intermediate products may lead to some variations on the outputs of the second stage. In this research, this issue is called "variations effect" of the intermediate products on the outputs of the second stage. These variations are very significant for a decision maker from the economic viewpoint because, in this case, he/she can identify variations of the outputs of the second stage and also recognize congestion by increasing the intermediate products; and then decides whether increasing the intermediate products is good or not for improving the performance of the two-stage DMU. This fact is the motivation for creating this study which hitherto, none of the studies in NDEA pay attention to this issue. The current research with the aim of tackling this issue proposes a DEA approach to determine the range of variations of the intermediate products and specify the effect of these variations on the outputs of the second stage in the presence of both negative and non-negative data. Finally, two numerical and empirical examples are provided to illustrate the potential application of the proposed approach. (C) 2018 Elsevier Ltd. All rights reserved.