Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/321372 
Authors: 
Year of Publication: 
2025
Series/Report no.: 
I4R Discussion Paper Series No. 245
Publisher: 
Institute for Replication (I4R), s.l.
Abstract: 
Mehmood et al. (2023) estimate the effect of Ramadan fasting hours on judicial decisions using case-level data from Pakistan and India. Using exogenous variation in fasting intensity due to the Islamic lunar calendar and latitude, the authors find that Muslim judges are significantly more likely to issue acquittals with longer fasting hours. Their main result, reported in Table 1, shows that each additional hour of fasting beyond the baseline minimum increases acquittal rates by about 10%. I successfully computationally reproduced this main result using the original authors' data and code: I found no coding errors or discrepancies in the replication package, and the point estimates and p-values in my reproduction match those reported in the published article. I then evaluated robustness for the Pakistan sample of judges using three alternative specifications. First, the result is robust to alternative inclusion of control variables: It remains stable and statistically significant whether controlling for case-level covariates, judge-level covariates, both, or none. Second, the effect persists across a different fixed effects specification that includes only district fixed effects. Finally, the result is robust to clustering standard errors at the judge level, although clustering at the month level increases the standard error and renders the estimate statistically insignificant. Overall, the authors' main finding is both computationally reproducible and robust.
Document Type: 
Working Paper

Files in This Item:
File
Size
915.96 kB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.