Sex Offender Registries: Fear Without Function?
Amanda Y. Agan
Princeton University - Department of Economics
December 1, 2008
I use three separate datasets and designs to determine whether sex offender registries are effective. First, state-level panel data is used to determine whether sex offender registries or public access to them decreases the rate of rape and other sexual abuse. Second, a dataset which contains information on the subsequent arrests of sex offenders released from prison in 1994 in 15 states is used to determine if registries reduce the recidivism rate of offenders required to register compared with those who do not. Finally, I combine data on locations of crimes in Washington, D.C. with data on locations of registered sex offenders to determine whether knowing the location of sex offenders in a region help predict the locations of sexual abuse. The results from all three datasets do not support the hypothesis that sex offender registries are effective tools for increasing public safety.
Keywords: sex offenders, registration, notification, recidivism, deterrence
JEL Classification: K0, K4, K42
Date posted: July 22, 2009 ; Last revised: March 1, 2010