Within-group common weights in Data Envelopment Analysis: an analysis of power plant efficiency

In many real world applications where DEA is applied, DMUs can often be put into groups, such as those which may be under a single management team. This often means that the multipliers used within a group should be common across that group's members. The case example examined in this regard is one involving a set of power plants, with each containing a set of power units under a common plant management. We develop a goal-programming model for this setting that seeks to derive such a common-multiplier set. The important feature of this multiplier set is that it minimizes the maximum discrepancy among the within-group scores from their ideal levels.