Oriented Stochastic Data Envelopment Models: Ranking Comparison to Stochastic Frontier Approach

Abstract

Results of data envelopment analysis sensitively respond to stochastic noise in the data. In this paper, by introduction of output augmentation and input reduction I extend additive models for stochastic data envelopment analysis (SDEA), which were developed by Li (1998) to handle the noise in the data. Applying the linearization procedure by Li (1998) the linearized versions of models are derived. In the empirical part of this work, the efficiency scores of Indonesian rice farms are computed. The computed scores are compared to the stochastic frontier approach scores by Druska and Horrace (2004) and weak ranking consistency with results of stochastic frontier method is observed.