Assessing the Efficiency of the Electric Energy Distribution using Data Envelopment Analysis with undesirable outputs

In this paper we propose to use Data Envelopment Analysis (DEA) to assess the efficiency of twenty electricity distributors in Brazil. Among the variables chosen to perform this study, we have chosen some distributors´ undesirable results, which in DEA are called undesirable outputs that need to be included. We use and discuss some approaches to deal with these types of variables. To determine a ranking for the electricity distributors we also use Inverted Frontier and the composed efficiency index, which provides a ranking for the distributors. The results show that all approaches used to deal with undesirable outputs provide the same ranking and the inverted frontier and the composed index aids in obtaining a ranking of distributors without any subjectivity.