 |
|
Abstract - This comment points out four severe reservations regarding Cho et al.’s (PS 2017)
finding that U.S. federal judges punish more harshly on “sleepy Mondays,” the Mondays after the start of Daylights Savings Time. First, Cho et al.'s finding pertains to only one of at least two dimensions of harshness, and the opposite result obtains in the second dimension. Second, even within the first dimension, Cho et al.'s result is statistically significant only because of a variable transformation and sample restrictions that are neither transparent in the article nor theoretically sound. Third, reanalysis of the data with superior methods reveals no significant “sleepy Monday” effect in the years 1992-2003. Fourth, sentences were on average shorter on “sleepy Mondays” out of sample,namely in 2004-2016.
Finally, the “sleepy Monday” effect is rejected out of sample. Cho et al. restricted their analysis to 1992-2003 because the USSC does not publicly disclose sentencing dates for later years. The Transactional Records Access Clearinghouse (TRAC), however, has obtained such data through Freedom of Information Act requests. Controlling for all relevant covariates available from TRAC, sentences imposed by district judges on “sleepy Mondays” are on average shorter and rarer than on other Mondays in the years 2004-2016 (models 8-10).[Research done while TRAC Fellow].
|
|
|
|
