10. Context-Dependent DEA

In this chapter, the context-dependent DEA is discussed. Since a product can appear attractive in comparison with a contextual of less attractive or unattractive alternatives, the performance of firms can be influenced by the context. For an example, twenty-three Tokyo public libraries are considered and a context-dependent DEA proposed by Chen et al. (2005) is discussed. The attractiveness of each library on a particular performance level in comparison with other libraries are measured. Libraries are classified on several empirical efficient frontiers, where each frontier is used to evaluate the attractiveness. The performance of the technically efficient libraries changes as the technically inefficient libraries change their performance. The context-dependent DEA also represents another view to differentiate the performance of efficient DMUs. When DMUs in a particular level are observed as having the same performance, the attractiveness measure lets us discriminate the “equal performance” based upon the third option or the same particular evaluation context. We also develop the VBA procedure to measure the attractiveness with just one click.