This is the comment I wrote (awaiting moderation):
________
Enos' article contains multiple irregularities in several datasets, for example:
1) Deletion of nearly 850 precincts (roughly 1/3 of total) from official Chicago (aggregate, precinct-level) elections data, for no apparent reason; deletions are non-random and involve observations with turnout/Republican vote values that directly contradict the author's hypotheses.
2) Modification of data values for an additional number of precincts (data values in author's datasets directly contradict official values).
3) Certain summary statistics, such as the number of non-voters in Enos' (individual-level) voter data file (a subset of the VCS Illinois voter data file), are sharply at odds with official turnout statistics, including a mathematically impossible number of non-voters.
Issues 1) and 2) were never investigated by AJPS, nor addressed by Enos.
Issue 3) was recently investigated by AJPS. Enos posted a response claiming that there are reliability issues in the original dataset, and that he did delete some observations for various reasons, such as addresses he was unable to geocode. However, he did not post the code he used to extract the Chicago data from the Illinois voter file and geocode addresses, which would have enabled an independent verification of his claims, as well as determining whether his deletions introduced biased in his estimates. AJPS closed their investigation prematurely, taking Enos' explanation at face value , without requesting any data or code, and without asking any of the natural follow-up questions or giving the complainant an opportunity to react to Enos' response. Most critically, regardless of any reliability issues in the original data, it is still the case that Enos' estimate values changed over time across several paper versions, for inexplicable reasons, and all changes (amounting to thousands of observations) are favorable to his hypotheses.
Full report and R code available here:
https://www.file.io/SDos/download/9Xa98d9DFtzP
This report was submitted by anonymous complainants to Harvard in 2018, and swept under the rug by FAS dean Claudine Gay (whose work also promotes the same racial threat theory as Enos and has been under suspicion for research fraud for 20 years) with an illiterate response (that the allegations do not fall within the purview of the FAS administrative entity formally tasked with investigating research misconduct).
____