I am using the slack-based measure (SBM) of efficiency (additive DEA) to assess the eco-efficiency of farms. Efficiency scores range from 0 to 1. At a second stage, I would like to examine variables that influence the eco-efficiency scores, such as the education level of the farmers or the size of the farm. I have read many times that the double bootstrap procedure developed by Simar and Wilson (2007) was more approprite than e.g. Tobit regression. Is there a way to apply this procedure to any DEA model with R software? I know there are DEA functions using Simar and Wilson double bootstrap in the FEAR package for example, but I am interested in using SBM.
Pauline Delhez For SBM you have to modify the internal functions of the Simar and Wilson(2007) bootstrap function in the package
https://cran.r-project.org/web/packages/rDEA/rDEA.pdf
The other thing you can do is to get the bias-corrected efficiency scores using Simar and Wilson (1998) procedure with your DEA model and then regress them on environmental variables using bootstrap truncated regression in https://cran.r-project.org/web/packages/truncSP/truncSP.pdf
Pauline Delhez, you can check this RG conversation: https://www.researchgate.net/post/Can_anyone_guide_me_through_the_rDEA_package
- Badges
- Science topic
- Similar topics
- Education