« "Environmental Indifference" | Main | Federal district judge dismisses illegal reentry prosecution holding "Section 1326 violates the Equal Protection Clause of the Fifth Amendment" »

August 18, 2021

Authors of provocative paper retract judge-specific claims about "most discriminatory" federal sentencing judges

I expressed concerns in this post last month about a new empirical paper making claims regarding the "most discriminatory" federal sentencing judges under the title "The Most Discriminatory Federal Judges Give Black and Hispanic Defendants At Least Double the Sentences of White Defendants."  In addition to articulating some first-cut concerns in my initial post, I also solicited and published here an extended post by Prof. Jonah Gelbach about the work based on this Twitter thread criticizing the paper.  

This new Twitter thread by one of the authors reports that the paper has now been revised to remove judge-specific claims as to the "most discriminatory" sentencing judges, and it is now re-titled "Racial Disparities in Criminal Sentencing Vary Considerably across Federal Judges."  This new New Jersey Law Journal article, headlined "Backpedaling: Authors of Study on Racist Rulings Retract Their Claims Against Pennsylvania, New Jersey Judges," provides some more details:

The authors of a study that accused some federal judges of extreme racial and ethnic bias in sentencing have withdrawn their conclusions about specific jurists following criticism of their methodology.

An earlier version of the study, published in July by the Institute for the Quantitative Study of Inclusion, Diversity and Equity, said two Eastern District of Pennsylvania judges and one from New Jersey give Black and Hispanic defendants sentences that are twice as long as those they give to whites.

But a revised version of the study, posted Tuesday, asks readers to disregard the references to specific judges....  “A previous version of this work included estimates on individually identified judges. Thanks to helpful feedback, we no longer place enough credence in judge-specific estimates to make sufficiently confident statements on any individual judge.  We encourage others not to rely upon results from earlier versions of this work,” the revised version of the study said.

The study’s lead author, Christian Michael Smith, explained on Twitter that, “while our initial paper appreciated how random chance, systematic missing data patterns, and/or hidden structural factors for sentencing could affect judge rankings, we now regard the following possibility as less remote than we initially regarded it: that a judge who is actually unproblematic could end up on the extreme end of our discrimination estimates, due to random chance, systematic missing data patterns, and/or hidden structural factors for sentencing.”...

Gelbach, in an email, said of the retraction, ”I applaud the authors for removing the ranking of judges’ sentencing practices and for making clear that people should not rely on those rankings. Given the data limitations, that was the right decision for them to make.”

Prior related posts:

August 18, 2021 at 09:14 PM | Permalink


A comment cross-posted from the 8/9/21 post on the same subject: I am puzzled that the authors of the study (none of whom appears to have any actual knowledge of or experience in federal sentencing practices) do not acknowledge in their retraction the unanimous reaction of practicing lawyers in the Districts of the (formerly) named "most discriminatory" judges, that these three judges are not by any means more racist in their decisions than other district judges. Such reactions by experienced private practitioners were quoted in the ALM stories, and were elaborated a couple of days later in a detailed statement by the public defender's office. I realize that subjective reactions could be disproven and shown to be fallacious by valid statistical analysis. But questionable statistics can be just as legitimately called into doubt by the unanimous reactions of those with actual, individualized experience appearing before those judges. I am no statistician, but I knew the minute I read about the study and its results that there had to be something deeply wrong with its methodology, given the plainly inapt names that rose to the "top" in its Hall of Shame.

Posted by: Peter Goldberger | Aug 18, 2021 11:01:48 PM

It might not hold but there was a major district court ruling


Posted by: Joe | Aug 19, 2021 7:59:33 AM

Post a comment

In the body of your email, please indicate if you are a professor, student, prosecutor, defense attorney, etc. so I can gain a sense of who is reading my blog. Thank you, DAB