“…In the abstract, the cost of a juvenile record is the difference between the sentence under experienced sentencing policies and the sentence the individual would have received had the judge and recommended sentencing range not taken into consideration the individual's juvenile adjudications. 11 Following Piehl and Bushway (2007) and Hickert et al (2022), we use a three-step approach to estimating this counterfactual. First, we estimate a probit model with county-level random effects to generate the predicted probability of incarceration for all individuals taking into account known predictors including race-sex dyad (White male, White female, Black male, Black female), age at the time of sentence (in years), disposition (trial, plea), offense type (lesser homicide offense, other person offense, property, drugs, driving under the influence [DUI], and other), whether there were multiple offenses of conviction (yes, no), whether the offense had a mandatory minimum (yes, no), whether the sentence was conforming to the guidelines (above standard range; within standard range; below standard range), and the guidelines recommended minimum incarceration sentence (logged, in months).…”