A comparison of efficiency measurement using data envelopment analysis and regression in the utility industry

It is the intent of this paper to discuss the theoretical differences between the manner in which regression and data envelopment analysis (DEA) separate variation into that which is caused by noise and that caused by inefficiency. It examines the empirical results of independent studies, one that used regression and one that used DEA. The studies were performed one year apart by independent teams using the same data and responding to the same questions. Both studies focus on monitoring the efficiency of the electric utility providers in light of deregulation