I wrote a GAMS code for a non-radial DEA model i.e. a multiplier Slack Based Model (SBM) with undesirable outputs, presented by WEI et al. (2012). Why do the model results in negative scores, despite the set Constrains 1(j) (see the GAMS code)?! Where did I go wrong? I attached the required files.
Thanks in advance.
Wei, C., Ni, J., & Du, L. (2012). Regional allocation of carbon dioxide abatement in China. China Economic Review, 23(3), 552-565.
Dear Morteza,
I have found a Gams model on the web which gives the same results as Banxia Frontier Software. I attach the model so you can insert your data and check if it also gives negative results.
Hi,
Try defining Constraints 2-4 for each input/output separately, e.g.:
Const2(i).. (1/m)*(1/xo(i))- v(i)=l=0;
TB
Due to elimination of a variable, the efficiency scores are actually 1+E, where E is a negative number resulting from the linear problem in the paper. See Section 13.2.2 in Cooper, Seiford, and Tone (2006).