![Sample our Social Sciences journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/110822/TOC-Subject-Sample-2021-Social-Sciences.jpg"})
ABSTRACT
I evaluate the 2006 Adam Walsh Act (AWA) and the sex offender registration and notification requirements that it established. I find that the law expands registry size in compliant states but does not reduce sex crimes, including those made registerable or more severe by the AWA. My results suggest that the conviction-based approach to determining inclusion on registries mandated by the AWA instead has the counterproductive effect of making registries less likely to reduce sex crimes and notification requirements are not effective at alerting prospective victims about unfamiliar offenders.
Acknowledgements
I would like to thank Alex Tabarrok, Robin Hanson, and Noel Johnson for their guidance and feedback, as well as Amanda Agan and JJ Prescott for helpful comments and suggestions and Brantly Callaway and Pedro Sant‘Anna for answering questions about their statistical method.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1. Their intuitive explanation for this finding is that sex offenders have little incentive to restrain their impulses when public notification has already turned them into outcasts.
2. As of 2008, almost all states conducted some form of actuarial risk assessment on sex offenders to determine their risk of re-offending (Daly Citation2008). Prior to SORNA and in many states that have not implemented SORNA, offenders’ risk assessment score helps determine registration and/or notification requirements. Specifically, as of 2003, slightly over half (Janus Citation2006; Logan Citation2003) and, as of 2007, approximately half of states (Freeman and Sandler Citation2010; Levenson and D’Amora Citation2007) used risk assessment to determine registration and notification requirements. Factors considered in determining the risk score of offenders include ‘treatment completion’ (Stenehjem Citation2012), prior criminal convictions, age, number of prior sex offenses, prison discipline history, offense-related sexual interests, access to victims, relationship with past victims, drug and alcohol abstinence, and employment stability. See also McGrath, Lasher, and Cumming (Citation2011), and Janus (Citation2006).
3. Due to a lag in the review process and other reasons discussed in Data, the year when states implemented SORNA in some cases preceded the year when SMART determined that they had implemented SORNA.
4. I exclude agencies whose jurisdiction does not have a population associated with it from my analysis due to the impossibility of calculating rates. This is the case, for example, for ‘national parks, colleges and universities, toll bridges and tunnels, and most state police departments’ (Federal Bureau of Investigation Citation2024c).
5. 44 states passed different versions of Jessica’s Law, which increased prison sentences for sex crimes and surveillance of sex offenders, between 2005 and 2014.
6. Or the date that legislation became effective, if different from the date when it was enacted.
7. Based on this methodology, I include Pennsylvania as a state that implemented SORNA. Even though SMART revoked its 2012 determination that the state is SORNA compliant, Pennsylvania enacted SORNA implementing legislation in 2011 and that legislation became effective in 2012 (Pennsylvania State Police Citation2018).
8. To implement the Bacon Decomposition, I use the Bacondecomp package available in Stata and R.
9. Specifically, it leads to a 41%, 33%, and 49% reduction in the sizes of the NIBRS, UCR Rape, and UCR Sex Offense Arrest datasets, respectively.
10. More specifically, I use the total number of sex crime offenses to calculate rates since an ‘incident,’ as defined by NIBRS, can involve multiple offenses as long as they were ‘committed by the same offender, or group of offenders acting in concert, at the same time and place’ (Federal Bureau of Investigation Citation2024d).
11. Specifically, registry size per 100k people grew, on average, by 50% and 22% between 2007 and 2017 in SORNA-compliant and non-compliant states, respectively.
12. Specifically, ‘abusive sexual contact’ against a minor under 13 is a Tier III offense while ‘abusive sexual contact’ against a child at least 13 years old is a Tier II offense (Adam Walsh Child Protection and Safety Act Citation2006, 5).
13. This model is influenced by Prescott and Rockoff (Citation2011) who use 1991–2005 NIBRS data to differentiate sex crimes based on the relationship between the victim and offender. They find that registries have the greatest negative effect on the incidence of sex crimes committed by ‘near’ offenders who are known to but not ‘close’ to victims. This includes neighbors and acquaintances and excludes friends, family members, and significant others (Prescott and Rockoff Citation2011, 175).
14. Most but not all of the pre-2005 registry size data was interpolated. See the note below .
15. See note 13 for the definition of ‘near’ offenders.
16. However, there is reason to doubt the seriousness of the problem of reverse causation in this model. It would not account for the difference in sign on the coefficient on registry size and the coefficient on the interaction term between SORNA and registry size. Also, sex offenders do not register until after imprisonment so there is a significant lag between when offenders commit a sex crime and when they register. Less than 3% of offenders who committed rape, sexual assault, or child molestation in 2002 were imprisoned for one year or less. As such, short-term changes in sex crime rates are unlikely to be a significant cause of short-term changes in registry size (Prescott and Rockoff Citation2011).
17. This derivation draws on Cunningham (Citation2021) and Sant’Anna and Zhao (Citation2020).
Additional information
Notes on contributors
James Freeman
James Freeman received his PhD in Economics at George Mason University in 2022. His research interests are in Law and Economics, Crime, and Econometrics. He received his Bachelor of Arts from Vassar College in 2006, his Master of Arts in Economics from the State University of New York at Buffalo in 2008, and his Master of Arts in International Relations from the Johns Hopkins School of Advanced International Studies in 2010.