Stochastic Two-Stage Data Envelopment Analysis and Dissociation of Overall Efficiency into Product- Process Efficiency: Case Study of 18 Iranian Insurance Companies

The purpose of this paper is to investigate efficiency in a two-stage process that includes stochastic input/output variables where outputs of the first stage serve as inputs of the second stage. Efficiency of each stage and the overall process are independently examined initially. As a next step, a relationship between efficiency of the overall process and those of discrete stages are explored. This effort employs a simulated example where inputs, outputs, and intermediate data are stochastically random. Finally, the results of the relationally random two-stage model are compared with those of an independently random two-stage model.