Andrew Eggers and Justin Grimmer have posted Comment on ‘Simple tests for the extent of vote fraud with absentee and provisional ballots in the 2020 US presidential election’. Here is the abstract:
Lott (2022) introduces ‘simple tests’ of voter fraud and applies them to measure the extent of fraud in the 2020 election. Using the tests, Lott (2022) claims to have discovered 10,000 extra votes for Biden in Pennsylvania and Georgia based on mishandling of absentee votes, 6,700 extra votes for Biden in Pennsylvania from inappropriate allowances for provisional votes, and 255,000 excess votes due to artificially large turnout across several counties in six key states. All three claims are demonstrably false: the first claim relies on a specification error that Lott (2022) inadvertently reintroduced into the analysis after we identified and corrected the same error in the first public draft of this paper; the second claim is based on analysis that, by its own logic, would show larger fraud in favor of Trump; the third depends on selective reporting, as Lott (2022) shows insignificant results when differential trends in turnout across states are acknowledged. As we document in the Appendix, Lott’s (2022) analysis is riddled with errors and fails to accurately report the research that was conducted. At several points
the paper misstates the regression specification being used (as we confirm through replicating Lott’s (2022) results), misreports regression coefficients (always in a direction favorable to his argument), and incorrectly reports variables used to produce a result. After correcting these errors, we show Lott’s (2022) analysis fails to provide any evidence of voter fraud in the 2020 election.
How did Lott’s article get published in Public Choice?