An Improved Approach for Estimating Returns to Scale in Data- Envelopment Analysis

The model presented in the paper by Biresh K Sahoo,Mohammad Khoveyni ,Robabeh Eslami and Pradipta Chaudhury(2016).( Returns to scale and most productive scale size in DEA with negative data ) for estimating Returns to Scale (RTS) in the presence of negative data cannot properly estimate RTS in most cases, even when the data are nonnegative, since it is relative to translation invariance, whereas RTS is not relative to translation invariance. By transferring the data, the model of Sahoo et al. (2016) continues to show the same RTS as before the transfer, while the type of RTS may have changed. Several counterexamples have been given to prove this claim. In addition, a model is proposed to modify theproblem of the model of Sahoo et al. (2016) along with the theorem. Then, examples are solved with other methods of estimating the RTS and the proposed method, and the results are investigated. It is observed that the model of Sahoo et al. (2016) provides an inaccurate estimate of RTS in most cases, while our proposed model correctly estimates RTS in all cases

کلیدواژه ها:

Data envelopment analysis(DEA) ، Return to scale(RTS) ، Translation invariance.