DOI: 10.36347/sjebm.2014.v01i11.005

Pages: 563-567

This paper measures the relative efficiency of some of milk-producing states in the United States (US) by Data Envelopment Analysis (DEA). 23 milk production states are selected with 12 factors, inclusive 8 inputs and 4 outputs. The recent unique model, Kourosh and Arash Model (KAM), is used to discriminate, rank and benchmark the states. There are 20 technically efficient states while Constant Returns to Scale (CRS) is applied. However, KAM suggests there is only one efficient company when thickness of the DEA frontier is less than 0.017 and other companies are inefficient with one millionth degree of freedom in inputs and outputs. Indeed, if only one millionth errors of the minimum values of positive factors are introduced in inputs and outputs values of the DEA frontier, the efficiency scores of technically efficient companies are sharply decreased. Reference sets for each technically efficient company as well as the rank of all companies are illustrated. Simulations are also calculated with Microsoft Excel Solver 2013.