Hi Mario ..as far as I know the DEA is mainly used to evaluate the efficiency using a benchmark . Most of the time we do assimilate the score of the benchmark to a linear function( straight line) ...so the score we calculate are made comparatively to the distance of the line. Usually like you said the score has to be between 0 and 1 , but we can extend that explanation if we consider the point upper to the line positive one and under negtive and we may consider in that case absolute value to avoid confusion 😕

Nope, there is something wrong with your model. By formulation, it isn't possible to have negative efficiency scores.

Dear Dr Phu,

In my opinion, there is no error in this case. The efficiency score obtained from the SBM-O-C model can be negative, unlike the other DEA models you mentioned, which are constrained to output scores between 0 and 1.

The SBM-O-C model uses an output-oriented approach, where the inputs are constant, and the outputs are maximized. In this case, if a DMU has excessive outputs relative to its inputs, it may receive a negative efficiency score. This can occur because the SBM-O-C model does not assume any bounds on the maximum efficiency score, unlike other DEA models constrained by the range of 0 to 1.

So, a negative efficiency score in the SBM-O-C model indicates that the DMU produces too much output relative to its inputs and may need to reduce its output levels to become more efficient. However, it is important to note that negative efficiency scores in the SBM-O-C model should be interpreted cautiously and may require further analysis to understand their underlying reasons.

Best regards,