Test whether processing non-transfer-eligible juvenile arrestees as adults has any effect on their likelihood of criminal recidivism.

A regression discontinuity design is used to analyze the effect of processing juveniles as adults on a four-year felony rearrest measure using a sample of 78,142 felony drug arrests.

For the felony drug offenders in this sample, processing juveniles as adults reduced the probability of recidivism by 3 to 5 percent. Based on the rapid onset and limited change in size of these effects over the duration of a four-year follow-up as well as the concentration of the effect within a subpopulation having the least risk of incarceration, we attribute this finding to a combination of enhanced deterrence and incapacitation in the adult system.

Our results suggest that processing juveniles in the adult system may not uniformly increase offending and may reduce offending in some circumstances. Our findings also highlight the utility of quasi-experimental research designs for estimating the life-course effects of contact with the criminal justice system.