Yep it is retracted
Kasy and Sautmann Econometrica 2021 proven wrong
-
The Society has the right to retract a published paper in the event of (i) a breach of warranties stated in the copyright form, (ii) any substantive errors, or (iii) evidence of scientific misconduct or other fraudulent actions. A committee of the three journal Editors and current sitting President for the Society will be formed to evaluate and decide the proposed retraction. More information concerning retraction guidelines of the Committee of Publication Ethics can be found here.
-
I emailed Kasy in October and this is what he told me:
"Hi Karl! We are in the process of writing a note with a correction, it is under review at Econometrica. In our opinion the Arxiv note is partially right, unfortunately. I would very much appreciate if you hold off on your article (for what outlet?) until this has been peer reviewed and posted. Best regards, Max"
I am not really up to date since then.
-
https://maxkasy.github.io/home/files/papers/correction_adaptiveexperimentspolicy.pdf
I have read the correction, but honestly I don't understand anything: othe original paper, the critique, and the correction. I feel really stupid. It would take maybe half year to read and figure out this literature, and I am not able to make such investment.
What is disconcerting however is that this board is so lame. People work on hard problems, they make mistakes, mistakes get fixed, people move on. No reason for hate and anymosity hereon.
-
https://maxkasy.github.io/home/files/papers/correction_adaptiveexperimentspolicy.pdf
I have read the correction, but honestly I don't understand anything: othe original paper, the critique, and the correction. I feel really stupid. It would take maybe half year to read and figure out this literature, and I am not able to make such investment.
What is disconcerting however is that this board is so lame. People work on hard problems, they make mistakes, mistakes get fixed, people move on. No reason for hate and anymosity hereon.You need the proper background in this case. MK had made rcist remarks concerning east asians dominating econometrics.
-
https://maxkasy.github.io/home/files/papers/correction_adaptiveexperimentspolicy.pdf
I have read the correction, but honestly I don't understand anything: othe original paper, the critique, and the correction. I feel really stupid. It would take maybe half year to read and figure out this literature, and I am not able to make such investment.
What is disconcerting however is that this board is so lame. People work on hard problems, they make mistakes, mistakes get fixed, people move on. No reason for hate and anymosity hereon.Mistakes should be detected at the review stage. Econometrica usually has 4 referees. Given that none found the mistake implies that they were not chosen well. It does not reflect well on the editor, who is, im my view, really responsible here. The issue is that Imbens is not technically strong enough and has too much of a bias towards certain literatures that he never should havd been the econometrics editor for econometrica.
-
https://maxkasy.github.io/home/files/papers/correction_adaptiveexperimentspolicy.pdf
I have read the correction, but honestly I don't understand anything: othe original paper, the critique, and the correction. I feel really stupid. It would take maybe half year to read and figure out this literature, and I am not able to make such investment.
What is disconcerting however is that this board is so lame. People work on hard problems, they make mistakes, mistakes get fixed, people move on. No reason for hate and anymosity hereon.Mistakes should be detected at the review stage. Econometrica usually has 4 referees. Given that none found the mistake implies that they were not chosen well. It does not reflect well on the editor, who is, im my view, really responsible here. The issue is that Imbens is not technically strong enough and has too much of a bias towards certain literatures that he never should havd been the econometrics editor for econometrica.
Proofs only get checked in math journals these days. But it is good to have the critique in Arxiv. Would be nice if editors published both the Arxiv comment and a correction as a response (if it resolves the issue).
I think there are many mistakes in the published paper, some big and some small. Econ should embrace correction culture. It is not a big deal to publish corrections in stats and math. It is more rare in Econ (I don’t know why — perhaps because people on EJMR will jump on you and say that you made an error). I have never seen corrections published in Machine learning. The referees are instructed NOT to read the proofs for ML conference submissions. So funny.