Dear all,

I have a question concerning a comment from a reviewer and I am not quite sure how to go about it. To keep it simple, I have a paper on RCT (facilitating saving intervention among farmers) and the reviewer says that I should do an IV version of it by taking the treatment (binary) as an instrument. This would mean taking the basic treatment effect estimate and divide it by 0.2(multiplying it by 5) which would get the treatment effect among "compliers".

I am not sure even how to think and how to code this! I would be happy to have some thoughts about this issue.
Thanks!