Still no reform effect, just overfitted control functions and inflated precision
In his comment on our replication of Ciacci (2024), Ciacci (2025b) again finds large and statistically significant treatment effects on rape from Sweden’s ban on sex purchases. These estimates derive from overfitted control functions and incorrect clustering in the regression discontinuity in time analysis. We show how this happens and conclude that the comment does not alter the prior conclusions in the replications of the original article (Adema et al. 2025; Ciacci 2025).