Journal of Experimental Criminology
https://doi.org/10.1007/s11292-021-09480-z

The effectiveness of Sex Offender Registration
and Notification: A meta-analysis of 25
years of findings
Kristen M. Zgoba 1

& Meghan

M. Mitchell 2

Accepted: 1 July 2021/Published online: 21 September 2021
# The Author(s), under exclusive licence to Springer Nature B.V. 2021

Abstract
Objectives Examine 25 years of Sex Offender Registration and Notification (SORN)
evaluations and their effects on recidivism.
Methods We rely on methodology guidelines established by the Campbell Collaboration for meta-analyses to systematically synthesize results from 18 research articles
including 474,640 formerly incarcerated individuals. We estimate the effect of SORN
policies on recidivism from 42 effect sizes and determine if the effect of SORN varies
by sexual or non-sexual recidivism when examining arrest or conviction as outcomes.
Results The random-effects meta-analysis model demonstrated that SORN does not
have a statistically significant impact on recidivism. This null effect exists when
examining a combined model and when disaggregating studies by sexual or nonsexual offenses, or conceptualizing recidivism by arrest or conviction.
Conclusions SORN policies demonstrate no effect on recidivism. This finding holds
important policy implications given the extensive adoption and net-widening of penalties related to SORN.
Keywords Evaluation . Megan’s Law . Meta-analysis . Sex Offender Registry and

Notification . SORN . RSOs

* Kristen M. Zgoba
kzgoba@fiu.edu
Meghan M. Mitchell
mmitchell@ucf.edu

1

Department of Criminology and Criminal Justice, Florida International University, 11200 SW 8th
Street, Miami, FL 33199, USA

2

Department of Criminal Justice, University of Central Florida, 12805, Pegasus Dr, Florida,
FL 32816, USA

~ Springer

K. M. Zgoba, M. M. Mitchell

Introduction
Official statistics on sexual assault indicate approximately 500,000 assault incidents
take place a year (Bierie & Davis-Siegel, 2015; NCJA-SOMAPI, 2017), although
crime victimization surveys hint at higher numbers (Morgan & Truman, 2020). As a
result of these heightened frequencies, and in the absence of empirical validation,
policies governing individuals convicted of sexual offenses have been established
(NCJA-SOMAPI, 2017). The importance of and reliance upon these policies, known
as Sex Offender Registration and Notification (SORN) laws, has been firmly
established in the American criminal justice system over the last two and a half
decades. These laws have been found to be widely supported by the public, law
enforcement, policymakers, and sometimes even those convicted of sexual offenses
(Bierie, 2016; Levenson, D’Amora, & Hern, 2007a; NCJA-SOMAPI, 2017; Tewksbury & Lees, 2007). Whether through media prominence, political touting, or community sentiment, it is clear that SORN laws are a mainstay in American culture.
As the year 2021 marks the 25th anniversary of the federal passage of Megan’s Law,
the past two and a half decades have been witness to numerous modifications and
additions to the law (Call, 2018; Lobanov-Rostovsy, 2015; Zgoba et al., 2018). While
the objective behind SORN policies is to both provide law enforcement a list of
registered sex offenders (RSOs) for investigation and to make the public aware of an
individual registrant’s address, SORN laws have spurred the growth of other sexual
offense policies. These variations have included laws that focused on registrant residence restrictions, civil commitment, polygraph, internet activity, GPS tracking, and
international travel (Lobanov-Rostovsy, 2015). In addition to the spill-over effect of
these laws into other areas of registrants’ lives, registries have grown in size and
duration of the registration period. The most recent numbers available from the
National Center for Missing and Exploited Children (NCMEC) indicated that over
935,000 people were registered and living in the USA (Bierie, n.d.)1. The escalation of
laws and penalties focusing on sexual offenses, as well as the growing number of
individuals placed on registries, highlights the recognition that SORN policies are
likely here to stay.
As time passes and SORN laws continue to expand in breadth and net-widen to
include more registrants for longer periods of time (Harris et al., 2020a), it becomes
imperative to examine the comprehensive effectiveness of the law. Although informative, the extant literature lacks a singular defining trend or consensus on its efficacy
(Zgoba et al., 2018). Despite widespread adoption, studies have not consistently
demonstrated that SORN is associated with a decrease in recidivism or an increase in
public safety (Bierie, 2016; NCJA-SOMAPI, 2017), which leads many to wonder, have
we put the cart before the horse? Moreover, no individual study exists that combines
the statistical effects and results of numerous studies into one comprehensive examination of the overall trend of SORN. This study aims to fill this gap in both the
literature framework and the empirical analyses through a meta-analysis of 25 years of
independent studies that examine the effectiveness of SORN policies. While Call
1

This number was released via personal communication with Dr. David Bierie from the US Marshals Service.
NCMEC data were utilized to create this count and the USMS approved the release and publication of the
information.

~ Springer

The effectiveness of Sex Offender Registration and Notification: A...

(2018) recently published a systematic review, and Pawson (2002) conducted one
approximately 20 years ago, the current study represents the only meta-analysis of
SORN effectiveness on recidivism.

History of Sex Offender Registration and Notification legislation
During the preceding 25 years, states, as well as the federal government, have continuously modified and, in many cases, strengthened sex offense legislation as it pertained to
the practice of registering individuals and notifying the public of their whereabouts (Harris
et al., 2020a). While registration and notification are separate constructs, collectively, they
are known as sex offender registration and notification laws. Since 1996, they have
functioned in tandem as one law (i.e., Megan’s Law and SORN). Registration refers to
the statutory requirement that upon release, individuals convicted of sex offenses register
their personal information with local law enforcement authorities, while notification refers
to the process by which the public is informed of the released individual’s address and
personal information. While intrinsically linked, the two policies have been born from
separate objectives—sex offender registration was intended to facilitate police apprehension of recidivists by maintaining a pool of potential suspects, as well as the deterrence of
subsequent offending due to increased scrutiny. In contrast, the objective of notification
was to improve the community’s ability to protect themselves from RSOs, through altered
personal behavior or reports to law enforcement regarding suspicious behavior by registrants (Matson & Lieb, 1997; Ragusa-Salerno & Zgoba, 2012).
While much of the focus has been on recent legislation for sexual offenses, original
versions of the law were adopted many decades before, as California became one of the
first states in 1947 to develop a sex offense registry (Call, 2018; Logan, 1999; Logan,
2009). Following a number of publicized sex crimes, Washington state gained national
attention in 1990 for implementation of a contemporary version of the law known as
the Community Protection Act (Matson & Lieb, 1997). Shortly thereafter in 1994,
registration was extended to all 50 states when the Wetterling Act was enacted into
federal law through the Title XVII of the Violent Crime Control and Law Enforcement
Act (Logan, 2009). The Wetterling Act required that by 1997 all states establish a sex
offense registry or risk losing federal criminal justice funding. Subsequently, the 1994
murder of Megan Kanka occurred a few months prior to the enactment of the
Wetterling Act and changed the content of the original version of the Act to include
a community notification requirement (Logan, 2009). One month later in October
1994, New Jersey enacted the Registration and Community Notification Law, more
commonly known as “Megan’s Law” (Petteruti & Walsh, 2008; Zgoba et al., 2018). In
1996, the federal version of Megan’s Law was later endorsed as an amendment to the
Wetterling Act (Kabat, 1998). As such, from 1996 to the present, registration and
notification policies have existed in combination as one law.
As it was written, the federal law did not mandate standardization of implementation
and states took the liberty to adopt individualized provisions for notifying the public
about those they deemed to carry the most risk. As this occurred, the notification
stipulations evolved in a variety of ways, for instance, producing different procedures
for the tiering of RSOs or implementing broad notification procedures without a
distinction between high and low risk (Chajewski & Mercado, 2009; Levenson,
~ Springer

K. M. Zgoba, M. M. Mitchell

Brannon, et al., 2007b; Lobanov-Rostovsy, 2015; Vasquez et al., 2008; Zgoba et al.,
2016; Zgoba et al., 2018). These variations in SORN procedures among the states
catalyzed Congress to enact the Adam Walsh Child Protection and Safety Act (AWA)
(see Logan, 2009, for a more thorough discussion). The Adam Walsh Act sought to
standardize SORN provisions with the objective of creating a widespread national
system for the registration and notification of persons convicted of a sexual crime or a
crime against a child (Zgoba et al., 2016). However, since the AWA functions similarly
to the Wetterling Act by withholding funding to states that do not implement, currently
only 17 states and three US territories have successfully implemented the provisions set
forth by the AWA (Harris & Lobanov-Rostovsky, 2010; Harris et al., 2020b; Petteruti
& Walsh, 2008; Zgoba et al., 2018).
Research exploring the efficacy of SORN
Research exploring the efficacy of SORN policies has focused primarily on individual
states, with a small number of studies examining multiple state effects (NCJASOMAPI, 2017). Within this research, with limited exceptions (Barnoski, 2005;
Bierie, 2016; Duwe & Donnay, 2008; Freeman, 2012), little evidence of a SORN
impact has been found for either first-time sexual offending or reoffending (Adkins
et al., 2000; Agan, 2011; Maddan et al., 2011; Sandler et al., 2008; Schram & Milloy,
1995; Vasquez et al., 2008; Zgoba et al., 2008; Zgoba et al., 2010; Zgoba et al., 2018).
One of the first examinations of the effectiveness of SORN legislation was published
by Schram and Milloy (1995, prior to federal passage of the law). They used a sample
of 180 high-risk individuals in Washington State who were either subjected or not to
community notification under Washington’s Community Protection Act. The authors
determined that individuals subjected to community notification were rearrested more
quickly for new sexual offenses, compared with those offenders who were not subject
to notification; however, overall recidivism rates between the two groups did not differ.
Similar state studies in New Jersey and Iowa also found negligible results associated
with SORN (Adkins et al., 2000; Tewksbury et al., 2012; Zgoba et al., 2008; Zgoba
et al., 2010), while a time series analysis in New York found that 95% of all sexual
offense arrests were committed by first-time offenders (Sandler et al., 2008).
Adkins et al. (2000) examined the effects of registration in Iowa. A preregistration
group (n = 201) and a post-registration group (n = 233) comparison indicated that sex
offense registration appeared to have varied effects on recidivism rates over the followup period of 4.3 years. Sex offense recidivism was low at 3% for the registry sample
and 3.5% for the pre-registry sample. Of those who were convicted of sex offenses, the
registry sample had a lower volume of recidivism per person than the pre-registry
sample; however, the differences in recidivism were not found to be statistically
significant. Similarly, researchers in New Jersey followed a sample of 550 individuals
released both pre- and post–SORN implementation for an average period of approximately 7 years (Zgoba et al., 2010). The authors determined that multiple measures of
recidivism, in addition to community tenure (the amount of time spent in the community before a rearrest) and harm reduction (decreased number of victims), were not
significantly different between the cohorts. Building upon these results, Tewksbury
et al. (2012) created matched samples of sex offenders in New Jersey released pre- and
post–Megan’s Law (N = 495). The groups were matched on a number of variables
~ Springer

The effectiveness of Sex Offender Registration and Notification: A...

theoretically linked to recidivism including age, race, education, family history, marriage, employment, substance use, mental health issues, sexual offense characteristics,
and criminal history. The timeframe was extended to allow for an 8-year post-release
examination. No significant differences were noted between the groups for both general
and sexual reoffending. Sandler et al. (2008) utilized a time series analysis to examine
differences in sexual offense arrest rates before and after the enactment of New York
State’s Sex Offender Registration Act. The authors found no support for the effectiveness of SORN laws in reducing sexual offending by rapists, child molesters, sexual
recidivists, or first-time sex offenders. In fact, the authors discovered that over 95% of
all sexual offense arrests were first-time offenses, thereby creating uncertainty that
SORN is an effective method to target those who repeatedly commit sexual offenses.
A study by Petrosino and Petrosino (1999) approached an evaluation of SORN from
a slightly different angle by determining how many repeat sex crimes may potentially
have been avoided if SORN was in effect in Massachusetts. The authors examined
criminal records of “sexual psychopaths” (N = 136) and found that 27% of the sample
had a prior conviction that met the requirements of the Massachusetts Registry Law
before their most recent sex crime. The authors conclude and that a small number of
cases would have potentially been prevented if SORN was in effect. Similarly,
Levenson and Zgoba (2016) examined the rate of repeat sex crime arrests in Florida
using aggregate data for the period 1990 to 2010. The average annual sexual repeat
arrest rate prior to and after the implementation of sexual offender registration laws in
1997 was 4.9% and 7.5%, respectively, indicating a statistically significant increase
post–SORN implementation.
Similar to the discussed quasi-experimental analyses, time series analyses of sexual
offending rates pre- and post–SORN enactment also find limited to no evidence of
effectiveness. Vasquez et al. (2008) compared rates of forcible rape in 10 states. One of
the particular strengths of this study was the inclusion of the multiple states and the
increased ability to generalize beyond a single state. Overall, the authors concluded that
SORN had no influence on the number of rapes reported after the law was implemented, although several states demonstrated a non-significant increase in the number of
rapes, a significant reduction in rates was only found in three states studied. Comparably, Agan (2011) used the Bureau of Justice Statistics data from 15 states that
contained information on the subsequent arrests of previously convicted sex offenders
released from prison in 1994. The results yield small, insignificant differences in
recidivism for those required to register and those who were not, causing the
researcher to conclude that registration does not appear to reduce reoffending rates
among RSOs.
Although the majority of prior research has found the effectiveness of SORN to be
futile in reducing sexual recidivism, of note are the few studies that have found limited
positive outcomes associated with SORN. For example, Barnoski (2005) examined the
criminal histories of 8359 individuals released before and after the passage of Washington state’s SORN policy using three measures of recidivism. The author discovered
that both violent and sexual felony recidivism decreased since the passage of the 1997
SORN statute; however, he notes that a causal link cannot be determined as other
factors, such as decreasing crime rates and increasing incarceration, can be a
contributing factor. Zgoba et al. (2008) discovered a similar finding in New Jersey,
where sexual offense rates steadily decreased after the implementation of SORN;
~ Springer

K. M. Zgoba, M. M. Mitchell

however, the finding cannot be isolated to SORN alone, as all crime rates decreased
during similar years. That same year, Duwe and Donnay (2008) compared three groups
of people in Minnesota: high-risk previously convicted sex offenders released with
SORN provisions, individuals released prior to implementation of SORN, and RSOs
released after SORN implementation, but not subject to community notification.
Although a comparison of the notification group and the prenotification group suggested a reduction in general and non-sexual recidivism, there was no reduction in
recidivism for the notification and non-notification group. The authors conclude that
community notification yielded a specific deterrent effect for those offenders registered
under its provisions only.
A small number of studies have also documented different outcomes associated with
the law. A subsequent analysis of more than 300,000 sex offenses in 15 states found
that registration reduced the frequency of sex offenses because it provided law enforcement with information on local sex offenders (Prescott & Rockoff, 2011). Similar to
Duwe and Donnay (2008), these results demonstrated a narrow effect—a decrease in
crime concentrated among “local” victims (e.g., friends, acquaintances, neighbors),
with little evidence of a decrease in crimes against strangers. Conversely, Freeman
(2012) found that notification status had an effect on recidivism, as well as a potential
positive benefit from SORN’s effects on timing to rearrest in New York. Among each
of the samples examined, individuals were rearrested for non-sexual offenses more
quickly after the implementation of SORN policies.
Current study
Given the widespread adoption and use of SORN policies and variation that exists
between studies, the present study seeks to examine the comprehensive effectiveness of
SORN through standardization of findings. Although the research reviewed above is
informative, it lacks a singular defining theme or a consensus among the findings.
Currently, no individual study exists that synthesizes the statistical effects and results of
numerous studies into one inclusive examination on the overall trend of SORN. This
study fills this gap through a meta-analysis using a number of independent studies that
examine the effectiveness of SORN policies. Accordingly, this study represents the first
meta-analysis of SORN effectiveness2.

Methods
A meta-analysis is conducted to examine the effect of SORN policies on recidivism and
to systematically synthesize results from numerous research articles. This approach is
well recognized for providing a “transparent, objective, and replicable framework” for
systematic and quantitative reviews (Borenstein et al., 2009, p. xxiii). To do so, we rely
2

It is important to note that, while important, a small number of the studies outlined in the research review are
not included in the meta-analysis due to expiration of data storage rules and data output reporting styles that
inhibited calculation of an effect size. Efforts were made with many authors to retrieve all relevant data for
inclusion in the current study. Some of the studies not included in the analysis are Adkins et al., (2000);
Petrosino and Petrosino (1999); Sandler et al., (2008); Schram and Milloy (1995); Vasquez et al., (2008); and
Zgoba et al., (2010).

~ Springer

The effectiveness of Sex Offender Registration and Notification: A...

on methodology guidelines established by the Campbell Collaboration (for review see
Campbell Collaboration, n.d.).
Inclusion and exclusion criteria
To examine the effect of SORN on recidivism, several criteria were used to gather
eligible studies. First, the target population was limited to adults who had been
incarcerated for a sexual offense and released from a correctional facility—this included both prisons and jails3. As juveniles are treated differently from adults in the
criminal justice system and subjected to varying forms of registration and notification,
studies examining their recidivism were excluded from this analysis. Second, the
outcome variable had to measure “recidivism” as indicated by failure to register4,
revocation, arrest, charges, conviction, or incarceration. Third, searches were limited
to published journal articles, books, dissertations, technical reports, and other grey
literature. Fourth, eligible studies had to provide the common statistics or raw data
necessary for the calculation of effect sizes. It is important to note that studies were not
excluded based on the rigor of their methodological design. Fifth, research articles were
limited to those written in the English language. Sixth and finally, publication dates
were limited from 1996 to 2020, as the federal version of SORN was implemented in
1996. Studies were excluded if they did not include the relevant statistics for inclusion
(authors were contacted to provide statistical data for inclusion) and if they did not meet
the inclusion criteria.
Strategies for searching the literature
Systematic procedures were used to conduct an exhaustive search of studies while
limiting bias. Keywords were searched using Google Scholar. The keyword Boolean
phrase used was “sex offender registration and notification” OR “sexual offender
registration and notification” AND registration OR sex offender registration OR
Megan’s Law OR Adam Walsh Act OR prisoner OR adult prisoner OR offender OR
incarceration OR inmate OR corrections OR sex offender NOT juvenile. Although
historically other databases were relied upon for meta-analyses, Google Scholar has
become remarkedly robust and commonly relied upon by scholars as a comprehensive
resource. We used this database for a number of reasons (Campbell, 2010; Haddaway
et al., 2015; Kendall, 2020; Krug, 2020; Merguerian, 2020). First, it has a robust ability
to search grey literature including dissertations/theses, conference proceedings, books,
and reports in addition to research articles. Second, Google Scholar searches the full
text of articles for keywords as opposed to simply searching the title, abstract, and
tagging information. Third, the database searches across all disciplines; hence, it is not
limited to specific areas of inquiry. Lastly, the materials that appear on the search
3

No exclusions were made solely based on this category. Meaning, no study was excluded because they had
studied parole or probation released offenders in exclusion. Parolees were included in this analysis because in
many states sex offenders are held on parole supervision for life (PSL) or community supervision for life
(CSL) after release from prison. Individuals who received only probation would not have been; however, no
studies were excluded for this reason.
4
FTR is a felony in many states and with the federal government. It was not included in any recidivism
analyses, however, because the studies were excluded for various other reasons.

~ Springer

K. M. Zgoba, M. M. Mitchell

engine are curated by a machine and algorithm as opposed to journals selected by
humans.
The first round of the online keyword searches was conducted between August 24,
2020, and August 25, 2020. These searches produced 1280 related publications, which
were later reduced to 228 after removing unrelated topics based on a review of the title
and summary from Google Scholar. Of this number, only six duplicates existed and
those were removed resulting in an article pool of 222. Following a review of the
abstracts, only 40 relevant articles remained. A review of article references generated an
additional 24 articles; however, 22 articles were removed after a careful review of their
methods, resulting in 42 articles. After thorough examinations of each publication by
both authors, 18 articles contained the appropriate inclusion criteria necessary for the
meta-analysis.
It should also be noted that the authors utilized their extensive foundation of
colleagues who work in legal practices, clinical and therapeutic offices, and research
institutions to provide suggestions of studies that may not have been on the list
originally (e.g., file drawer papers). Furthermore, references were searched and published authors were contacted to review the sample and ensure that studies were not
missed.
Coding procedures
To determine the effect of SORN policies on recidivism, 18 studies were manually
coded for the effect sizes and outcome variables.5 To examine variations, two outcome
variables were categorically coded and included in this analysis. Public concern
surrounding sex offense recidivism is centered on a person’s likelihood to reoffend
for another sexual offense (Przybylski, 2015); consequently, an offense type variable
was created to measure sexual and non-sexual recidivism. Additionally, some studies
combined both types of offenses; hence, a combined indicator for both was created.
Some forms of recidivism may be more or less likely among individuals (Hepburn &
Griffin, 2004); thus, an indicator of the type of recidivism (i.e., revocation, arrest,
charge, convicted, reincarcerated, or multiple indicators, which could include two or
more of the aforementioned outcomes) was also used.
Effect size coding and meta-analytic strategy
In the current study, we first calculated the mean effect of SORN on recidivism based
on 18 studies. Overall, 42 effect sizes were available throughout the studies because
some studies had multiple dependent variables. For example, a study could have a
5

Meta-analysis procedures were originally developed to demonstrate a bivariate and isolated effect of a
treatment on an outcome. These effects are ideal in experimental and quasi-experimental designs; however,
those study designs are not always feasible in social sciences which has resulted in researchers relying on
multivariate designs to control for confounding effects. Within multivariate analyses, controls may be
operationalized or measured differently resulting in an increase in bias and inability to compare controls
consistently across studies (Borenstein et al., 2009). Due to this difference in statistics and operationalizations,
multi-variate effects were not included in this meta-analysis. These studies include Ackerman et al., 2012;
Freeman & Sandler, 2010; Maurelli & Ronan, 2013; Park et al., 2014; Vasquez et al., 2008; and Zgoba et al.,
2018.

~ Springer

The effectiveness of Sex Offender Registration and Notification: A...

unique effect for rearrest for any charge, rearrest for a sex offense, and conviction of a
sex offense. Within this example, this one study could provide three effects.6 Although
substantial heterogeneity existed in offense type and type of recidivism, those variations existed within-study and did not commonly exist between studies. If the heterogeneity would have been between studies, moderator analyses would have been ideal;
however, a sufficient number of studies did not exist within each offense type and type
of recidivism to accurately examine each outcome and its moderating effect. For
example, we could only examine the moderating effects for type of recidivism for
arrest, conviction, and charge—with limitations when examining conviction and charge
because they only have a k of 3 and 2 respectively—but we cannot include multiple
types of recidivism or reincarceration because both of those only had a k of 1. Thus, any
attempt at moderation would be incomplete. Additionally, when conducting metaanalyses, it is important not to include multiple effects from one study within one
meta-analysis because those effects will be highly correlated with one another. Moreover, a multilevel approach is not advisable given the small number of level 2 clusters
(i.e., k = 18) and the small number of effect sizes within each of those clusters (i.e., n =
42 effect sizes). Thus, to examine variation in study outcomes, we instead conducted
four additional meta-analyses to determine SORN effects based on certain outcomes.
All analyses were conducted using The Comprehensive Meta-Analysis Program.
Random-effects models were estimated for all models.7 These models were utilized
because it is theoretically unlikely that all studies included in the analysis are identical,
due to their variation in sample size, follow-up period, and recidivism measures.
Moreover, this type of study is useful when effect sizes are drawn from a subset of
the population in examination (Borenstein et al., 2009). Based on this fluidity, it is not
viable to assume that the true effect size is consistent for the entire sample of studies.
Moreover, the I2 statistic indicates that a high degree of heterogeneity exists between
studies.8

Results
As reported in Table 1, 18 studies and effect sizes were included in the analysis, and
these studies included a total of 474,640 offenders who were subjected to SORN and
had their recidivism tracked longitudinally. Regarding study specifics for the combined
effects (N = 18), six studies measured sexual recidivism, one examined non-sexual
recidivism, and 11 had a combined recidivism measure of both sexual and non-sexual
6

In instances where multiple effects existed, if the authors could combine those effects to create an combined
study effect, they did so. In studies where effects could not be combined, the authors chose one effect for the
main model and noted that in Table 1.
7
The authors recognize that when the number of studies is small, it is difficult to properly apply the randomeffects model given the large variability in effect size (Borenstein et al., 2009). Thus, we also provide the
fixed-effects for each model although we feel confident that the random-effects models are most appropriate
for our data given theoretical and statistical rationales.
8
Heterogeneity demonstrates the variation that exists in the true effect size underlying a certain population.
This is essentially the effect that would exist for an infinite number of cases (Borenstein et al., 2009). I2
statistics are interpreted as a ratio and demonstrate inconsistency across studies by determining the extent to
which confidence intervals overlap—determining the true effect underlying studies (Higgins et al., 2003;
Higgins & Thompson, 2002).

~ Springer

K. M. Zgoba, M. M. Mitchell
Table 1 Primary study characteristics, random-effects sizes, and confidence intervals for 18 SORN and
recidivism studies (N = 474,640)
Study

Offense type

Recidivism type

Sample size

Odds ratio

95% CI

Adkins et al., 2000

Both

Multiple

434

0.648

[0.426–0.984]

Agan, 2011

Both

Arrest

9623

1.141

[1.051–1.239]

Barnoski, 2005

Both

Conviction

8359

1.288

[1.169–1.418]

Bouffard & Askew, 2019

Sexual

Charge

424

0.182

[0.126–0.263]

Carr, 2015

Both

Conviction

2005

0.846

[0.647–1.106]

Cohen & Spidell, 2016

Both

Arrest

93,524

0.463

[0.426–0.504]

Duwe & Donnay, 2008

Both

Arrest

310

0.559

[0.351–0.890]

Freeman, 2012

Both

Arrest

17,165

0.641

[0.602–0.682]

Fundack, 2019

Sexual

Arrest

1985

25.503

[20.588–31.591]

Letourneau et al., 2010

Sexual

Charge

6064

1.149

[1.048–1.258]

Levenson & Zgoba, 2016

Sexual

Arrest

180

1.964

[1.149–3.355]

Maddan et al., 2011

Both

Arrest

2920

0.544

[0.458–0.647]

Prescott & Rockoff, 2011

Sexual

Arrest

328,260

1.499

[0.737–3.048]

Tewksbury & Jennings, 2010

Sexual

Conviction

1582

1.034

[0.864–1.236]

Tewksbury et al., 2012

Both

Arrest

495

0.754

[0.525–1.083]

Zevitz, 2006

Both

Reincarceration

213

0.982

[0.514–1.876]

Zgoba et al., 2018

Non-sexual

Arrest

547

0.802

[0.495–1.301]

Zgoba et al., 2008

Both

Arrest

550

0.716

[0.511–1.004]

Note: Odds ratio = 0.962; CI= [0.685–1.351]; Z = −0.222; p = 0.824; I2 = 98.89; Q = 1527.37, p < 0.05.

recidivism. When operationalizing recidivism, one study did so using multiple indicators, two did so using a charge, 11 relied on arrest, three used conviction, and one
measured reincarceration.
Individual and combined effect sizes
The effect sizes and confidence intervals for each of the 18 studies are included in
Table 1 and visually depicted in Figure 1. The effects from 11 studies were statistically
significant at p-value < 0.05 and one study was very close to this arbitrary cut point
with a p-value of 0.053. Within these studies, seven of them found that SORN
decreased recidivism and this effect ranged from an 82% reduction in recidivism
(Bouffard & Askew, 2019) to a 28% reduction (Zgoba et al., 2008). Other statistically
significant studies (n = 5) reported an increase in recidivism as a result of SORN. These
increases ranged from a 14% increase (Agan, 2011) to an increase in the odds by 25.50
(Fundack, 2019).9 The remaining six studies had non-significant findings of which an
equal number of studies has odds ratios above and below one. The Q-statistic for the
model was significant (Q(18) = 1527.37, p < 0.05), indicating heterogeneity of effect
sizes throughout studies (I2 = 98.89). The random-effects model demonstrated that

9

Although this study had an effect size that was much larger than our sample, we retained it for analyses.

~ Springer

Statistics for each study
Odds
ratio

Adkins et al. , 2000
Agan , 2011
Barnoski , 2005
Bouffard & Askew, 2019
Carr, 2015
Cohen & Spidell , 2016
Duwe & Donnay, 2008
Freeman, 2012
Fundack 2019
Letourneau et al. , 2010
Levenson & Zgoba , 2016
Maddan et al. , 2011
Prescott & Rockoff, 2011
Tewksbury & Jennings , 2010
Tewksbury, Jennings, & Zgoba , 2012
Zevitz, 2006
Zgoba , Jennings, & Salerno , 2018
Zgoba , Witt, & Dalessandro, 2008

Odds ratio and 95% Cl

Lower Upper
limit
limit Z-Value p-Value

0.648 0.426 0.984 -2.033
1.141
1.051 1.239
3.143
5.136
1.288 1.1 69 1.418
0.182 0.126 0.263 -9.057
0.846 0.647 1.106 -1 .223
0.463 0.426 0.504 -18.012
0.559 0.351 0.890 -2.450
0.641
0.602 0.682 -14.045
25.503 20.588 31 .591 29.650
1.149 1.048 1.258
2.971
1.964 1.149 3.355
2.469
0.544 0.458 0.647 -6.897
1.119
1.499 0.737 3.048
1.034 0.864 1.236
0.363
0.754 0.525 1.083 -1 .527
0.982 0.514 1.876 -0.056
0.802 0.495 1.301 -0.893
0.716 0.511 1.004 -1 .938
0.962 0.685 1.352 -0.222

0.042
0.002
0.000
0.000
0.221
0.000
0.014
0.000
0.000
0.003
0.014
0.000
0.263
0.717
0.127
0.955
0.372
0.053
0.824

I

Jt5)

1•
■

---I
■

.
0.01

.,,
S·

.

Favours A

Notes. Favours A associates with reductions in recidivism, whereas Favours B corresponds with increases in recidivism.
Figure 1 Overall random-effects model evaluating SORN on recidivism

I ■

T

0.1

V,

~

The effectiveness of Sex Offender Registration and Notification: A...

Study name

1

10
Favours B

100

K. M. Zgoba, M. M. Mitchell
Table 2 Study characteristics, random-effects sizes, and confidence intervals for SORN and sexual and nonsexual offense type recidivism studies
Study

Recidivism type

Sample size

Odds ratio

95% CI

Sex offensea
Adkins et al., 2000

Multiple

434

0.858

[0.296–2.490]

Agan, 2011

Arrest

9623

1.009

[0.837–1.216]

Barnoski, 2005

Conviction

8359

0.553

[0.452–0.676]

Bouffard & Askew, 2019

Charge

424

0.182

[0.126–0.263]

Carr, 2015

Conviction

2005

0.766

[0.361–1.628]

Cohen & Spidell, 2016

Arrest

93,524

0.550

[0.454–0.667]

Duwe & Donnay, 2008

Arrest

310

0.297

[0.129–0.684]

Freeman, 2012

Arrest

17,165

0.768

[0.682–0.864]

Fundack, 2019

Arrest

1985

25.503

[20.588–31.591]

Letourneau et al., 2010

Charge

6064

1.149

[1.048–1.258]

Levenson & Zgoba, 2016

Arrest

180

1.964

[1.149–3.355]

Maddan et al., 2011

Arrest

2920

0.867

[0.661–1.137]

Prescott & Rockoff, 2011

Arrest

328,260

1.499

[0.737–3.048]

Tewksbury & Jennings, 2010

Conviction

1582

1.034

[0.864–1.236]

Tewksbury et al., 2012

Arrest

495

0.719

[0.410–1.259]

Zevitz, 2006

Arrest

213

1.720

[0.723–4.090]

Zgoba et al., 2008

Arrest

550

0.747

[0.413–1.352]

Non-sex offenseb
Adkins et al., 2000

Multiple

434

0.642

[0.416–0.992]

Agan, 2011

Arrest

9623

1.140

[1.046–1.243]

Carr, 2015

Arrest

2005

0.986

[0.742–1.311]

Cohen & Spidell, 2016

Arrest

93,524

0.213

[0.168–0.270]

Duwe & Donnay, 2008

Arrest

310

0.763

[0.471–1.235]

Freeman, 2012

Arrest

17,165

0.686

[0.645–0.730]

Levenson & Zgoba, 2016

Arrest

180

1.969

[1.152–3.364]

Tewksbury et al., 2012

Arrest

495

0.873

[0.613–1.242]

Zgoba et al., 2018

Arrest

547

0.802

[0.495–1.301]

Note: a Odds ratio = 0.978; CI= [0.609–1.572]; Z = −0.091; p = 0.928; I2 = 98.51; Q = 1074.52, p < 0.05.
b Odds ratio = 0.768; CI= [0.555–1.063]; Z = −1.594; p = 0.111; I2 = 96.39; Q = 221.46, p < 0.05.

SORN reduced the mean of recidivism by 3.8%, (OR = 0.962 [0.685 − 1.351]) which
was not statistically significant (Z18 = −0.222, p = 0.824).10
Effect sizes for sexual or non-sexual offense types
The effect sizes and confidence intervals for studies by sexual and non-sexual offense
type are included in Table 2. Seventeen studies provided estimates for sexual offenses
10

The fixed-effects model demonstrated that SORN reduced the mean of recidivism by 13.5%, (OR = 0.865
[0.837–0.894]) which was statistically significant (Z18 = −8.612, p = 0.000).

~ Springer

The effectiveness of Sex Offender Registration and Notification: A...

with various types of recidivism (i.e., arrest, conviction, charge, or multiple types).
Within this grouping of studies, eight studies were statistically significant at p-value <
0.05. All but three of the studies found that SORN decreased recidivism. The effects
ranged from an 82% reduction in recidivism (Bouffard & Askew, 2019) to a 23%
reduction (Freeman, 2012). Statistically significant increases were found to range from
a 15% increase (Letourneau et al., 2010) to an increase in the odds by 25.50 (Fundack,
2019). The remaining nine studies had non-significant findings. The Q-statistic for the
model was significant (Q(17) = 1074.52, p < 0.05) with a model I2 of 98.51. The
random-effects model demonstrated that (see Figure 2), for sexual offenses, SORN
reduced the mean of recidivism by 2.2% (OR = 0.978 [0.609 − 1.572]), which was not
statistically significant (Z17 = −0.091, p = 0.928).11
For non-sexual offenses, nine studies provided estimates with various types of
recidivism (i.e., arrest or multiple types). Five of the nine studies were statistically
significant at p-value < 0.05. Three studies found that SORN decreased non-sexual
recidivism with effects ranging from a 79% reduction in recidivism (Cohen & Spidell,
2016) to a 31% reduction (Freeman, 2012). Statistically significant increases were
found for two studies and those effects ranged from 14% (Agan, 2011) to an increase of
97% (Levenson & Zgoba, 2016). The remaining four studies had non-significant
findings. The Q-statistic for the model was significant (Q(9) = 221.46, p < 0.05) with
a model I2 of 96.39. For non-sexual offense types (see Figure 3), the random-effects
model demonstrated that SORN reduced the mean of recidivism by 23.2% (OR = 0.768
[0.555 − 1.063]), which was not statistically significant (Z9 = −1.594, p = 0.111).12
Effect sizes for arrested or convicted recidivism types
The effect sizes and confidence intervals for studies disaggregated by recidivism type
are presented in Table 3. When examining arrest as a recidivism outcome, thirteen
studies were included of which seven had significant findings (p-value < 0.05).
Significant and negative effects ranged from a 54% reduction in recidivism (Cohen
& Spidell, 2016) to a 36% reduction (Freeman, 2012). Conversely, statistically significant increases in arrests were found to range from a 13% increase (Agan, 2011) to an
increase in the odds by 25.50 (Fundack, 2019). The remaining six studies had nonsignificant findings. The Q-statistic for the model was significant (Q(13) = 1320.40, p <
0.05) with a model I2 of 99.09. The random-effects model (see Figure 4) demonstrated
that SORN increased the mean of recidivism by 11.9%, (OR = 1.19 [0.697 − 1.797])
which was not statistically significant (Z13 = 0.467, p = 0.641).13
Five studies examined conviction as an outcome using various type of recidivism
with three of them being statistically significant at p-value < 0.05. All but one of the
studies found that SORN decreased convictions with all three effects hovering around a
46% reduction. The sole study that reported a statistically significant increase found
11

The fixed-effects model demonstrated that for sexual offenses SORN increased the mean of recidivism by
6.9%, (OR = 1.069 [1.014–1.126]) which was statistically significant (Z17 = 2.475, p = 0.013).
12
For non-sexual offense types, the fixed effects model demonstrated that SORN reduced the mean of
recidivism by 21.8%, (OR = 0.782 [0.746–0.819]) which was statistically significant (Z9 = −10.255, p =
0.000).
13
The fixed-effects model demonstrated that SORN decreased the mean of recidivism by 22.4%, (OR = 0.776
[0.746–0.807]) which was statistically significant (Z13 = −12.646, p = 0.000).

~ Springer

I~

.,,...

ti)

s·

~
...

Statistics for each studl

Study name
Odds
ratio

Lower Upper
lirrit
limit Z-Value p-Value

0.296 2.490
0.837 1.216
0.452 0.676
0.126 0.263
0.361 1.628
0.454 0.667
0.129 0.684
0.682 0.864
20.588 31 .591
1.048 1.258
1.149 3.355
0.661 1.137
0.737 3.048
0.864 1.236
0.410 1.259
0.723 4.090
0.413 1.352
0.609 1.572

-0.281
0.092
-5.771
-9.057
-0.693
-6.103
-2.852
-4.369
29.650
2.971
2.469
-1 .032
1.119
0.363
-1.154
1.227
-0.963
-0.091

0.779
0.926
0.000
0.000
0.488
0.000
0.004
0.000
0.000
0.003
0.014
0.302
0.263
0.717
0.249
0.220
0.336
0.928

I

I•

I

I

0.01

0.1

I

■

~

Favours A

Notes. Favours A associates with reductions in recidivism, whereas FavoursB corresponds with increases in recidivism.
Figure 2 Random-effects model evaluating SORN on sexual offense type recidivism

l

1

10
Favours B

100

K. M. Zgoba, M. M. Mitchell

Adkins et al. , 2000
0.858
Agan , 2011
1.009
0.553
Barnoski , 2005
Bouffard & Askew, 2019
0.182
0.766
Carr, 2015
0.550
Cohen & Spidell , 2016
0.297
Duwe & Donnay, 2008
0.768
Freeman , 2012
25.503
Fundack 2019
1.149
Letourneau et al. , 2010
1.964
Levenson & Zgoba, 2016
0.867
Maddan et al. , 2011
1.499
Prescott & Rockoff, 2011
1.034
Tewksbury & Jennings, 2010
0.719
Tewksbury, Jennings, & Zgoba , 2012
1.720
Zevitz, 2006
0.747
Zgoba , Witt, & Dalessandro, 2008
0.978

Odds ratio and 95% Cl

Statistics for each study

Odds ratio and 95%CI

The effectiveness of Sex Offender Registration and Notification: A...

Study name

Odds Lower Upper
ratio limit limit Z-Value p-Value

Adkins et al , 2000

0.642

0.416 0.992

-1 .997

0.046

Agan, 2011

1.140

1.046 1.243

2.992

0.003

Carr, 2015

0.986

0.742

-0.099

0.922

Cohen & Spidell, 2016

0.213

0.168 0.270 -12. 772

0.000

Duwe& Donnay, 2008

0.763

0.471

-1 .101

0.271

Freeman, 2012

0.686

0.645 0.730 -11 .909

0.000

1.311
1.235

Levenson &Zgoba, 2016

1.969

1152 3364

2.479

001 3

Tewksbury, Jennings, &Zgoba, 2012

0.873

0.613

1.242

-0.756

0.449

Zgoba, Jennings, & Salerno, 2018

0.802

0.495

1.301

-0.893

0.372

0.782

0.746 0.819 -10.255

0.000

I

I

I

I

0.01

0.1

-

Favours A
Notes. Favours A associates with reductions in recidivism, whereas Favours B corresponds with increases in recidivism.
Figure 3. Random-effects model evaluating SORN on non-sexual offense type recidivism

1t5)

.,,

V,

~...

t
1

10

Favours B

100

K. M. Zgoba, M. M. Mitchell
Table 3 Recidivism type study characteristics, random-effects sizes, and confidence intervals for SORN and
arrested or convicted recidivism studies
Study

Sexual or non-sexual recidivism Sample size Odds ratio 95% CI

Arresteda
Agan, 2011

Both

9623

1.132

[1.041–1.230]

Carr, 2015

Both

2005

0.846

[0.647–1.106]

Cohen & Spidell, 2016

Both

93,524

0.463

[0.426–0.504]

Duwe & Donnay, 2008

Both

310

0.559

[0.351–0.890]

Freeman, 2012

Both

17,165

0.641

[0.602–0.682]

Fundack 2019

Sexual

1985

25.503

[20.588–31.591]

Levenson & Zgoba, 2016

Non-sexual

180

1.969

[1.152–3.364]

Maddan et al., 2011

Both

2920

0.544

[0.457–0.646]

Prescott & Rockoff, 2011

Sexual

328,260

1.499

[0.737–3.048]

Tewksbury et al., 2012

Both

495

0.754

[0.525–1.083]

Zevitz, 2006

Sexual

213

1.720

[0.723–4.090]

Zgoba et al., 2018

Non-sexual

547

0.802

[0.495–1.301]

Zgoba et al., 2008

Both

550

0.716

[0.511–1.004]

Agan, 2011

Sexual

9623

1.115

[0.851–1.461]

Barnoski, 2005

Both

8359

1.501

[1.357–1.660]

Duwe & Donnay, 2008

Both

310

0.535

[0.333–0.857]

Maddan et al., 2011

Both

2920

0.542

[0.454–0.646]

1582

1.034

[0.864–1.236]

Convictedb

Tewksbury & Jennings, 2010 Sexual

Note: a Odds ratio = 1.119; CI= [0.697–1.797]; Z = 0.467; p = 0.641; I2 = 99.09; Q = 1320.40, p < 0.05.
ratio = 0.886; CI= [0.569–1.380]; Z = −1.534; p = 0.593; I2 = 96.29; Q = 107.91, p < 0.05

b Odds

that SORN increased convictions by 50% (Barnoski, 2005). The remaining two studies
had non-significant findings. The Q-statistic for the model was significant (Q(5) =
221.46, p < 0.05) with a model I2 of 96.39. For studies examining conviction as an
outcome, the random-effects model demonstrated that SORN reduced the mean of
recidivism by 11.4% (OR = 0.886 [0.569 − 1.380]), which was not statistically
significant (Z5 = −1.534, p = 0.593) (see Figure 5).14
Publication bias
With any meta-analysis, the possibility for publication bias exists (Lipsey & Wilson,
2000) given the likelihood of significant findings to be published by journals over nonsignificant findings (Rosenthal, 1979). As previously mentioned, we relied heavily on
informal conversations with a network of experts to ensure that any “file drawer” studies
were included in the sample; however, we also wanted to examine the publication bias
14

For studies examining conviction as an outcome, the fixed-effects model demonstrated that SORN
increased the mean of recidivism by 11.7%, (OR = 1.117 [1.037–1.203]) which was statistically significant
(Z5 = 2.905, p = 0.004).

~ Springer

Odds ratio and 95%CI

Statistics for each study_

The effectiveness of Sex Offender Registration and Notification: A...

Study_ name

Odds Lower Upper
ratio limit
limit Z-Value p-Value

~an, 2011
Carr, 2015
Cohen & Spidell, 2016
Duwe & Donnay, 2008
Freeman, 2012
Fundack 2019
Levenson & Zgoba, 2016
tv'addan et al., 2011
Prescott & Rockoff, 2011
Tewksbury, Jennings, & Zgoba, 2012
Zevitz, 2006
Zgoba, Jennings, & Salerno, 2018
Zgoba, Witt, & Dalessandro, 2008

1.132 1.041 1.230 2.912
0.846 0.647 1.106 -1 .223
0.463 0.426 0.504 -18.012
0.559 0.351 0.890 -2.450
0.641 0.602 0.682 -14.045
25.503 20.588 31 .591 29.650
1.969 1.152 3.364 2.479
0.544 0.457 0.646 -6.907
1499 0.737 3.048 1.119
0.754 0.525 1083 -1 .529
1.720 0.723 4.090 1.227
0.802 0.495 1.301 -0.893
0.716 0.511 1.004 -1 .938
1.119 0.697 1.797 0.467

0.004
0.221
0000
0.014
0000
0.000
0.013
0.000
0.263
0.126
0.220
0.372
0.053
0.641

I

+

■

I

0.01

.,,

V,

~...

Figure 4 Random-effects model evaluating SORN on arrest recidivism

■

0,1

10

Favours A

Favours B

Notes. Favours A associates with reductions in recidivism, whereas Favours B corresponds with increases in recidivism.
1t5)

...

100

K. M. Zgoba, M. M. Mitchell
Study name

Statistics for each study

Odds ratio and 95% CI

Odds Lower Upper
ratio
limit
limit Z-Value p-Value

Agan, 2011

1.115

0.851

1.461

0.791

0.429

Barnoski, 2005

1.501

1.357

1.660

7.892

0.000

Duwe & Donnay, 2008

0.535

0.333

0.857

-2.598

0.009

Maddan et al., 2011

0.542

0.454

0.646

-6.823

0.000

Tewksbury & Jennings, 2010

1.034

0.864

1.236

0.363

0.717

0.886

0.569

1.380

-0.534

0.593
0.01

0.1

Favours A

1

10

100

Favours B

Notes. Favours A associates with reductions in recidivism, whereas Favours B corresponds with increases in recidivism.

Figure 5 Random-effects model evaluating SORN on conviction recidivism

statistically. Thus, we examined any bias for our combined model (Table 1 and Figure 1)
using a number of techniques. First, of the 18 studies included in this analysis, statistically
significant and non-statistically significant findings were found in both positive and
negative directions, suggesting that there is not a bias in terms of what types of findings
are published on this topic. Second, a funnel plot was used to chart asymmetry of risk
ratios and standard errors (Sterne & Egger, 2001). Studies were distributed around the
mean suggesting that bias may not exist. To further test the asymmetry based on the
intercept of a linear regression against the reciprocal of standard errors, the Egger et al.
(1997) was used. Asymmetry was not present, and the model had an intercept of 2.22
(95% CI = −5.56, 10.00), p = 0.278), suggesting that publication bias does not exist.
Fourth, the Orwin’s Fail Safe N (Orwin, 1983) was used and equaled 28. This estimate
indicates that 28 studies with a mean risk ratio of more than 1 would be needed to change
the cumulative effect to be greater than a significance level of 0.05. Considering only 18
studies were included in this meta-analysis in total—all with various effects—increasing
the number by 28 would be substantial. Fifth and finally, trim and fill estimates (Duval &
Tweedie, 2000) were computed to estimate the effect of publication bias, on the outcome
of the meta-analyses. The trim-and-fill procedure imputes the missing data and recalculates the combined effect should publication bias exist. As demonstrated in Figure 6,
0.0

Standard Error

0.1

0

0

0.2

0
0.3

0
0.4
-4

-3

-2

-1

0

1

Log odds ratio

Figure 6. Funnel plot for asymmetry of risk ratios and standard errors

~ Springer

2

3

4

The effectiveness of Sex Offender Registration and Notification: A...

no imputed studies were needed for symmetry. Based on the aforementioned bias
statistics, publication bias is negligent and not likely to substantially influence the mean
effect size between SORN and recidivism.

Discussion
Study findings
The goal of the present study was to conduct a meta-analysis of SORN evaluations.
Relying on the methodology guidelines established by the Campbell Collaboration, we
systematically synthesized results from 18 research articles spanning approximately 25
years, from 1996 (the year of federal implementation of SORN) to 2020 including
474,640 formerly incarcerated sex offenders. To date, this is the most extensive review
period of SORN and the only meta-analysis. We examined three different outcomes—
combined recidivism, offense type by sexual recidivism and non-sexual recidivism, and
recidivism type by arrest and conviction. The random-effects meta-analysis model
demonstrated that the SORN does not have a statistically significant impact on
combined recidivism. This null effect also exists when disaggregating studies by sexual
or non-sexual recidivism, or conceptualizing recidivism by arrest or conviction. In sum,
this finding indicates that the implementation of SORN over the last 25 years has
demonstrated no effect on the deterrence or decrease of adult offending.
While these findings are consistent with previous research on the topic, they
create a number of legislative and policy-related dilemmas that require attention
and warrant a discussion on reform. First, the evolution of SORN policies over
the last two and a half decades has had a net-widening effect to make more
registrants eligible for registry placement and for considerably longer periods of
time. These expansions are counterintuitive to the empirical research that demonstrates individuals convicted of sexual offenses age out of crime as other
individuals do (Hanson, 2002; Hanson et al., 2014; Hanson et al., 2018). This
link between age and general criminal behavior is firmly established in the field
of criminology and desistance from criminal offending is “nearly inevitable”
across offender types (Gottfredson & Hirschi, 1990; Hanson, 2018). In general,
the risk of sexual re-offense decreases with advancing age. Specifically, recidivism rates for individuals in their 40s and 50s consistently demonstrate declines, with subsequent steep declines after the age of 60. Additionally, in large
studies, the sexual recidivism rates of those over the age of 60 were shown to be
less than 4% (Hanson, 2002; Helmus et al., 2012). In light of the aging registrant
population, these findings suggest that longer registration periods under SORN
policy changes may be inefficient and they have the potential to misallocate
resources and supervision to those at varying and sometimes lower-risk levels
(Booth, 2016). Since not all individuals are equally likely to reoffend, and
because risk wanes with age, the heightened number of individuals remaining
on the registry forces individuals into a “one size fits all” solution that may make
it difficult for the public to discern true risk (Kahn et al., 2017; Zgoba et al.,
2016). While well-intentioned, over-allocation of management strategies may
also encourage collateral consequences among the population of registered sex
~ Springer

K. M. Zgoba, M. M. Mitchell

offenders, as lower-risk individuals may be forced into homelessness and unemployment. Accordingly, risk levels may increase for these individuals as a
misuse of policies creates a barrier to successful community reintegration
(Levenson, D’Amora & Hern, 2007).
In addition to desistance due to age, a growing body of research demonstrates that
recidivism risk declines the longer individuals remain offense-free in the community.
Hanson et al. (2018) found that individuals reach a “desistance threshold,” whereby
recidivism is decreased by half with every 5 years in the community offense-free, and
that by 20 years, even high-risk registered individuals are less likely to sexually
recidivate than a non-sexual offender who commits an “out of the blue” sexual offense
(see also Hanson, 2018; Hanson et al., 2014). These findings indicate that certain
individuals convicted of a sexual offense will present no significant risk at the time of
sentencing, and if offense-free in the community for 10 to 15 years, even the highestrisk individuals will pose no significant risk. In essence, the risk of a sexual offense is
“reset” to that of someone who has no history of a sexual offense. This is important
since many SORN policies have what is referred to as “time bars” or lifetime registration requirements, both of which are unrelated to dynamic risk and duration offensefree in the community. These time-based requirements suggest that rule and lawabiding behavior are irrelevant and life consequences are set according to one singular
criminal act, even if this offense occurred decades before. This policy response
promotes the fundamental idea that individuals convicted of sexual offenses maintain
high and enduring risk across their lifespan (Hanson et al., 2018; Kahn et al., 2017).
Decades of empirical research fails to support this notion and experts in the field posit
that communities would be better served by implementing strategies based on riskneed-responsivity paradigms, whereby resources are funneled to those most in need—
individuals with a high risk of sexual recidivism (Bonta & Andrews, 2007). By not
doing so, we dilute the efficacy of important management strategies and cannot be
surprised when research indicates that the policies are ineffective, when in fact they are
being applied ineffectively. It is also important to remember that as a society we run the
risk of creating new problems, by misdiagnosing current problems. Maintaining low or
reduced risk individuals on registries for a lifetime, or barring them from petitioning for
termination, may promote collateral consequences in the form of housing instability,
employment instability, and internet bans, thereby further increasing their risk of
reoffending. As SORN laws currently exist, any public safety benefits we accumulate
will likely not outweigh the collateral harm they promote (Hanson, 2018; Mercado
et al., 2008).
Lastly, one of the cornerstones of SORN is that identification as an individual
convicted of a sexual offense will deter reoffending. The notion here is that if we
apply a label to an individual and tell everyone else to watch this person’s behavior,
they will desist from the unwanted behavior. However, research has not supported this
assumption. The majority of empirical studies indicate that those convicted of a sexual
offense have no previous sexual offense on their record, and in one instance over 95%
of all sexual offense arrests were committed by first-time offenders (Hepburn & Griffin,
2002; Sandler et al., 2008). In addition, many have posited that individuals may
internalize criminal labels, subsequently creating self-fulfilling prophecies. Braithwaite
(1989) built upon labeling theory by arguing that the application of labels increases
crime in certain circumstances, particularly when efforts are not made to reintegrate the
~ Springer

The effectiveness of Sex Offender Registration and Notification: A...

person back into society, when they are rejected or are labeled on a long-term basis. All
of these notions ring true under modern-day SORN policies. These empirical and
anecdotal research findings minimize the strength of one of the more basic principles
of SORN legislation.
In sum, despite their futility and potential damage, SORN policies are likely here to
stay due to their widespread public, law enforcement, and political support. This
situation presents the metaphorical dilemma of shutting the stable door after the horse
has left the barn. Support for the laws is unwavering, yet 25 years of empirical data
point to the demonstrable ineffectiveness of SORN policies; how do we then reign the
horse back in?
Since the implementation of the AWA, SORN policies have expanded in breadth
and scope and show little likelihood of being scaled back. The NCMEC indicates that
by now almost one million individuals in the USA are listed on the sex offense registry
(Bierie, n.d.), and many for extended periods of time due to registration increases. It is
imperative that researchers take steps beyond academic publishing to inform the public,
law enforcement, and policy makers that governmental oversight of registrant presence
is not a feasible solution to protecting potential sexual abuse victims. As upwards of
ninety percent of victims and offenders know and prey on one another, we need to
confront the uncomfortable truth that those who commit sexual offenses are usually not
strangers. In addition to the overuse of tier levels and registration periods, a database
and notification system to alert the public of stranger identification and addresses is
inherently flawed and we should therefore not be surprised by the null effects of SORN
policies.
Limitations
This study is not without limitations. Because many studies included in this analysis
relied on official measures of recidivism, the findings must be interpreted with caution
regarding undercounting. Recidivism is defined by offenses that are known to the
police and appear on the criminal histories that are utilized for research purposes.
Previous studies show that sexual offenses may be underreported to police and that the
official records used in the majority of research may underestimate the number of
sexual offenses that are actually occurring (Furby et al., 1989). This underreporting will
subsequently affect base rate counts of offending, follow-up periods, and time at large
(Langevin et al., 2004). However, it bears noting that official records are the most
accepted measure of recidivism, as anecdotal, imputation techniques, and qualitative
measures introduce ambiguity, particularly with sexual victimization. Additionally,
utilizing a meta-analytic technique alleviates some of the low base rate concerns as
well.
It should also be recognized that some advocates of registries believe they hold
purpose in various areas of crime reduction that should not be overlooked or
underestimated (Bierie, 2016). Particularly, limited research has shown that different
SORN components have aided law enforcement investigations in the capture and
identification of perpetrators, as well as a quicker time to arrest and specific deterrence
among individuals on the registry (Bierie, 2016; Duwe & Donnay, 2008; Prescott &
Rockoff, 2011). Researchers should have a balanced conversation about these potential
benefits, as well as the null findings and collateral consequences cited in other research,
~ Springer

K. M. Zgoba, M. M. Mitchell

as we move toward having an informed debate on efficacy. As noted by Bierie (2016),
it may be worth looking at what good can come out of the law if it works some of the
time for some of those placed on it. It may be possible that modifications (i.e., targeted
enforcement of registry placement, reduction in lifetime registration, and utilization of
actuarial risk assessment tools to name a few) may be made that will still yield value in
the law (see the work on Grand Challenges by Levenson et al., 2016).
Lastly, meta-analyses, though extremely useful in summarizing the effects of SORN
and recidivism, are not without limitations. A small number of related studies could not
be included in this meta-analysis because they did not contain the necessary information to calculate an effect size (see footnotes one and three). Even with reaching out to
the authors of these papers, we were not able to gather the appropriate statistics.
Related, only 18 studies were included in the analysis. Although this k is more than
suitable for a meta-analysis, it makes examining variation between studies challenging
which is related to the next point, moderating analyses. We were not able to examine
moderation and subgroup analyses, which may reveal alternative findings. Finally, we
did not exclude studies based on methodological rigor—doing so has the potential to
produce various findings. Despite these limitations, this study provides the most
expansive and up-to-date evaluation of SORN policies.

Conclusion
The present study adds to the existing research by completing an objective and
replicable framework for a meta-analysis of 25 years of sex offense registration and
notification evaluations. The study findings provide comprehensive evidence that
SORN policies have no effect on sexual and non-sexual crime commission over their
period of existence, thereby failing to deliver on the intention of increasing public
safety. Given the vast support that exists for the laws, their lack of efficacy will likely
create a false sense of security for the public and may ultimately create more harm than
benefit. Communities rely on the government for determination of risk, regardless of its
acuity, and those placed on registries may experience harm and increased risk of reoffense through societal disenfranchisement and loss of jobs and housing. As previous
research has consistently found that clinical and actuarial assessments are better
predictors of recidivism risk than state legislation, public safety would likely be
maximized by focusing limited resources on the highest-risk individuals, rather than
utilizing a one size fits all law (Andrews et al., 2011; Zgoba et al., 2018). Furthermore,
this call for reform of SORN policies, and lifetime requirements and offense bars
specifically, comes at a time when the country is experiencing both a decarceration and
criminal justice reform movement. For much of the last four to five decades, incarceration growth and criminal justice sanctions were accepted as a necessary consequence
of high rates of crime or a perceived deterrence (Zgoba & Clear, 2021). Recently,
however, examinations of the connection between crime and criminal justice sanctions
have questioned their effect on crime prevention (for a review, see National Research
Council, 2014), while at the same time recognizing a broad range of undesirable
consequences such as burgeoning prison populations, offense management strategies,
and criminal justice sanctions (Clear, 2007; Western, 2007; Zgoba & Clear, 2021). It is
important to the pursuit of science and empirical inquiry that our conversations for
~ Springer

The effectiveness of Sex Offender Registration and Notification: A...

reform begin to consider those who have committed low-level acts of violence,
including low-risk sexual offenses. Due to the dynamic nature of risk across time,
offense, and age, recognition that SORN policies may prove to be more hurtful than
helpful is a necessary conversation that does not automatically signal society has gone
soft on crime. Allocating resources to the highest-risk individuals will likely allow
treatment providers and law enforcement to focus their efforts on those who need it the
most, not on the diffusion of all individuals. It is time that we work as an empirically
informed community, unhindered by emotion, to find a solution to reining the horse
back into the barn.

References
Ackerman, A. R., Sacks, M., & Greenberg, D. F. (2012). Legislation targeting sex offenders: Are recent
policies effective in reducing rape? Justice Quarterly, 29(6), 858–887.
Adkins, G., Huff, D., & Stageberg, P. (2000). The Iowa sex offender registry and recidivism. Iowa Department
of Human Rights.
Agan, A. Y. (2011). Sex offender registries: Fear without function? The Journal of Law and Economics, 54(1),
207–239.
Andrews, D. A., Bonta, J., & Wormith, J. S. (2011). The risk-need-responsivity (RNR) model: Does adding
the good lives model contribute to effective crime prevention? Criminal Justice and Behavior, 38(7),
735–755.
Barnoski, R. P. (2005). Sex offender sentencing in Washington State: Has community notification reduced
recidivism? Washington State Institute for Public Policy.
Bierie, D. M. (2016). The utility of sex offender registration: A research note. Journal of Sexual Aggression,
22(2), 263–273.
Bierie, D. M. Personal Communication on June 22, 2021. United States Marshals Service.
Bierie, D. M., & Davis-Siegel, J. C. (2015). Measurement matters: Comparing old and new definitions of rape
in federal statistical reporting. Sexual Abuse, 27(5), 443–459.
Bonta, J., & Andrews, D. A. (2007). Risk-need-responsivity model for offender assessment and rehabilitation.
Rehabilitation, 6(1), 1–22.
Booth, B. D. (2016). Elderly sexual offenders. Current Psychiatry Reports, 18(4), 34.
Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. R. (2009). Introduction to meta-analysis.
Wiley.
Bouffard, J. A., & Askew, L. N. (2019). Time-series analyses of the impact of sex offender registration and
notification law implementation and subsequent modifications on rates of sexual offenses. Crime &
Delinquency, 65(11), 1483–1512.
Braithwaite, J. (1989). Crime, shame and reintegration. Cambridge University Press.
Call, C. (2018). Megan’s Law 20 years later: A systematic review of the literature on the effectiveness of sex
offender registration and notification. Journal of Behavioral and Social Sciences, 5, 2015–2217.
Campbell, H. (2010). Google Scholar: A credible database? The Clever Researcher https://beryliveylibrary.
wordpress.com/2010/10/21/google-scholar/. Accessed 29 Jan 2020.
Campbell Collaboration. (n.d.). So you want to write a Campbell systematic review? The Campbell
Collaboration: Resource Center. http://www.campbellcollaboration.org/resources/research/Write_a_
systematic_review.php. Accessed 29 Jan 2020.
Carr, J. (2015). The effect of sex offender registries on recidivism: Evidence from a natural experiment. https://
webapps.krannert.purdue.edu/sites/Home/DirectoryApi/Files/98f4dcf8-3431-4b5f-b0a3-83ae043dc2a1/
Download. Accessed 8 Jan 2021.
Chajewski, M., & Mercado, C. C. (2009). An evaluation of sex offender residency restriction functioning in
town, county, and city-wide jurisdictions. Criminal Justice Policy Review, 20, 44–61.
Clear, T. R. (2007). Imprisoning communities: How mass incarceration makes disadvantaged neighborhoods
worse. Oxford University Press.
Cohen, T. H., & Spidell, M. C. (2016). How dangerous are they: An analysis of sex offenders under federal
post-conviction supervision. Federal Probation, 80, 21.

~ Springer

K. M. Zgoba, M. M. Mitchell
Duval, S., & Tweedie, R. L. (2000). Trim and fill: A simple funnel-plot-based method of testing and adjusting
for publication bias in meta-analysis. Biometrics, 56(2), 455–463.
Duwe, G., & Donnay, W. (2008). Impact of Megan’s Law on sex offender recidivism: The Minnesota
experience. Criminology, 46, 411–446.
Egger, M., Davey Smith, G., Schneider, M., & Minder, C. (1997). Bias in meta-analysis detected by a simple,
graphical test. British Medical Journal, 315(7109), 629–634.
Freeman, N. J. (2012). The public safety impact of community notification laws: Rearrest of convicted sex
offenders. Crime & Delinquency, 58, 539–564.
Freeman, N. J., & Sandler, J. C. (2010). The Adam Walsh Act: A false sense of security or an effective public
policy initiative? Criminal Justice Policy Review, 21(1), 31–49.
Fundack, A. (2019). Maryland Sex Offender Registry and Sex Offender Recidivism: Time Series
Correlational Study(Doctoral dissertation, Walden University).
Furby, L., Weinrott, M. R., & Blackshaw, L. (1989). Sex offender recidivism: a review. Psychological
Bulletin, 105(1), 3.
Gottfredson, M. R., & Hirschi, T. (1990). A general theory of crime. Stanford University Press.
Haddaway, N. R., Collins, A. M., Coughlin, D., & Kirk, S. (2015). The role of Google Scholar in evidence
reviews and its applicability to grey literature searching. PLoS One, 10(9), e0138237.
Hanson, R. K. (2002). Recidivism and age: Follow-up data from 4,673 sexual offenders. Journal of
Interpersonal Violence, 17(10), 1046–1062.
Hanson, R. K. (2018). Long-term recidivism studies show that desistance is the norm. Criminal Justice and
Behavior, 45(9), 1340–1346.
Hanson, R. K., Harris, A. J. R., Helmus, L., & Thornton, D. (2014). High-risk sex offenders may not be high
risk forever. Journal of Interpersonal Violence, 29, 2792–2813. https://doi.org/10.1177/
0886260514526062.
Hanson, R. K., Harris, A. J., Letourneau, E., Helmus, L. M., & Thornton, D. (2018). Reductions in risk based
on time offense-free in the community: Once a sexual offender, not always a sexual offender. Psychology,
Public Policy, and Law, 24(1), 48.
Harris, A. J., Lobanov-Rostovsky, C., & Levenson, J. S. (2010). Widening the net: The effects of transitioning
to the Adam Walsh Act’s federally mandated sex offender classification system. Criminal Justice and
Behavior, 37(5), 503–519.
Harris, A. J., Kras, K., Lobanov-Rostovsky, C., & Ann, Q. (2020a). Information sharing and the role of Sex
Offender Registration and Notification. (Document Number 254680). US Department of Justice.
Harris, A. J., Kras, K. R., Lobanov-Rostovsky, C., & Ann, Q. (2020b). States’ SORNA implementation
journeys: Lessons learned and policy implications. New Criminal Law Review, 23(3), 315–365.
Harris, A. J., & Lobanov-Rostovsky, C. (2010). Implementing the Adam Walsh Act’s sex offender registration
and notification provisions: A survey of the states. Criminal Justice Policy Review, 21(2), 202–222.
Helmus, L., Thornton, D., Hanson, R. K., & Babchishin, K. M. (2012). Improving the predictive accuracy of
Static-99 and Static-2002 with older sex offenders: Revised age weights. Sexual Abuse: A Journal of
Research and Treatment, 24(1), 64–101.
Hepburn, J., & Griffin, M. (2002). An analysis of risk factors contributing to the recidivism of sex offenders on
probation. Washington, DC: US Department of Justice, NCJ Number 203905.
Hepburn, J., & Griffin, M. (2004). An analysis of risk factors contributing to the recidivism of sex offenders on
probation. US Department of Justice.
Higgins, J. P. T., & Thompson, S. G. (2002). Quantifying heterogeneity in a meta-analysis. Statistics in
Medicine, 21, 1539–1558.
Higgins, J. P. T., Thompson, S. G., Deeks, J. J., & Altman, D. G. (2003). Measuring inconsistency in metaanalyses. BMJ, 327(6), 557–560.
Kabat, A. R. (1998). Scarlet letter sex offender databases and community notification: Sacrificing personal
privacy for a symbol’s sake. The American Criminal Law Review, 35, 333–370.
Kahn, R. E., Ambroziak, G., Hanson, R. K., & Thornton, D. (2017). Release from the sex offender label.
Archives of Sexual Behavior, 46(4), 861–864.
Kendall, S. (2020). LibGuides: PubMed, Web of Science, or Google Scholar? A behind-the-scenes guide for
life scientists: Which one is best: PubMed, Web of Science, or Google Scholar? https://libguides.lib.msu.
edu/c.php?g=96972&p=627295. Accessed 29 Jan 2020.
Krug, A. (2020). Library resources & services: Use Google Scholar: library databases vs. Google
scholar. https://howardcc.libguides.com/c.php?g=52041&p=428654. Accessed 29 Jan 2020.
Langevin, R., Curnoe, S., Federoff, P., Bennett, R., Langevin, M., Peever, C., Pettica, R., & Sandhu, S.
(2004). Lifetime sex offender recidivism: A 25-year follow-up study. Canadian Journal of Criminology
and Criminal Justice, 46, 531–552.

~ Springer

The effectiveness of Sex Offender Registration and Notification: A...
Letourneau, E. J., Levenson, J. S., Bandyopadhyay, D., Sinha, D., & Armstrong, K. S. (2010). Effects of
South Carolina’s sex offender registration and notification policy on adult recidivism. Criminal Justice
Policy Review, 21(4), 435–458.
Levenson, J. S., & Zgoba, K. M. (2016). Community protection policies and repeat sexual offenses in Florida.
International Journal of Offender Therapy and Comparative Criminology, 60(10), 1140–1158.
Levenson, J. S., D’Amora, D. A., & Hern, A. L. (2007a). Megan’s Law and its impact on community re-entry
for sex offenders. Behavioral Sciences & the Law, 25, 587–602.
Levenson, J. S., Brannon, Y. N., Fortney, T., & Baker, J. (2007b). Public perceptions about sex offenders and
community protection policies. Analyses of Social Issues and Public Policy, 7(1), 137–161.
Levenson, J. S., Grady, M. D., & Leibowitz, G. (2016). Grand challenges: Social justice and the need for
evidence-based sex offender registry reform. Journal of Sociology & Social Welfare, 43, 3.
Lipsey, M. W., & Wilson, D. (2000). Practical meta-analysis (applied social research methods). Sage.
Lobanov-Rostovsy, C. (2015). Adult sex offender management. United States Department of Justice. Office of
Justice Programs Office of Sex Offender Sentencing, Monitoring, Apprehending, Registering, and
Tracking, SOMAPI Research Brief, 1-6.
Logan, W. A. (1999). Liberty interests in the preventive state: Procedural due process and sex offender
community notification laws. The Journal of Criminal Law and Criminology, 89, 1167–1233.
Logan, W. A. (2009). Knowledge as power: Criminal registration and community notification laws in
America. Stanford University Press.
Maddan, S., Miller, J. M., Walker, J. T., & Marshall, I. H. (2011). Utilizing criminal history information to
explore the effect of community notification on sex offender recidivism. Justice Quarterly, 28(2), 303–
324.
Matson, S., & Lieb, R. (1997). Megan’s Law: A review of state and federal legislation. Washington State
Institute for Public Policy (Document No. 97-10-1101).
Maurelli, K., & Ronan, G. (2013). A time-series analysis of the effectiveness of sex offender notification laws
in the USA. The Journal of Forensic Psychiatry & Psychology, 24(1), 128–143.
Mercado, C. C., Alvarez, S., & Levenson, J. (2008). The impact of specialized sex offender legislation on
community reentry. Sexual Abuse, 20(2), 188–205.
Merguerian, K. (2020). Subject guides: Google scholar, databases and research: using Google Scholar. https://
subjectguides.lib.neu.edu/c.php?g=336121&p=2262644. Accesse 29,2021.
Morgan, R. E. & Truman, J. L. (2020). Criminal Victimization, 2019. Bureau of Justice Statistics, NCJ
255113.
National Criminal Justice Association (NCJA) (2017). Sex Offender Management Assessment and Planning
Initiative (SOMAPI).
National Research Council. (2014). The growth of incarceration in the United States: Exploring causes and
consequences. The National Academies Press. https://doi.org/10.17226/18613.
Orwin, R. G. (1983). A fail-safe N for effect size in meta-analysis. Journal of Educational Statistics, 8(2),
157–159.
Park, J. H., Bandyopadhyay, D., & Letourneau, E. (2014). Examining deterrence of adult sex crimes: A semiparametric intervention time-series approach. Computational Statistics & Data Analysis, 69, 198–207.
Pawson, R. (2002). Does Megan’s Law work?: A theory-driven systematic review. ESRC UK Centre for
Evidence Based Policy and Practice.
Petrosino, A. J., & Petrosino, C. (1999). The public safety potential of Megan’s Law in Massachusetts: An
assessment from a sample of criminal sexual psychopaths. Crime & Delinquency, 45, 140–158.
Petteruti, A., & Walsh, N. (2008). Registering harm: How sex offense registries fail youth communities.
Justice Policy Institute.
Prescott, J. J., & Rockoff, J. E. (2011). Do sex offender registration and notification laws affect criminal
behavior? Journal of Law and Economics, 54, 161–206.
Przybylski, R. (2015). Recidivism of adult sexual offenders. United States Department of Justice. Office of
Justice Programs Office of Sex Offender Sentencing, Monitoring, Apprehending, Registering, and
Tracking, SOMAPI Research Brief, 1-6.
Ragusa-Salerno, L. M., & Zgoba, K. M. (2012). Taking stock of 20 years of sex offender laws and research:
An examination of whether sex offender legislation has helped or hindered our efforts. Journal of Crime
and Justice, 35(3), 335–355.
Rosenthal, R. (1979). The file drawer problem and tolerance for null results. Psychological Bulletin, 86(3),
638–641.
Sandler, J. C., Freeman, N. J., & Socia, K. M. (2008). Does a watched pot boil? A time-series analysis of New
York State’s sex offender registration and notification law. Psychology, Public Policy, and Law, 14, 284–
302.

~ Springer

K. M. Zgoba, M. M. Mitchell
Schram, D. D., & Milloy, C. D. (1995). Community notification: A study of offender characteristics and
recidivism. Washington Institute of Public Policy.
Sterne, J. A. C., & Egger, M. (2001). Funnel plots for detecting bias in meta-analysis: Guidelines on choice of
axis. Journal of Clinical Epidemiology, 54(10), 1046–1055.
Tewksbury, R., & Lees, M. B. (2007). Perceptions of punishment: How registered sex offenders view
registries. Crime & Delinquency, 53(3), 380–407.
Tewksbury, R., & Jennings, W. G. (2010). Assessing the impact of sex offender registration and community
notification on sex-offending trajectories. Criminal Justice and Behavior, 37(5), 570–582.
Tewksbury, R., Jennings, W. G., & Zgoba, K. (2012). A longitudinal examination of sex offender recidivism
prior to and following the implementation of SORN. Behavioral Sciences & the Law, 30, 308–328.
Vasquez, B. E., Maddan, S., & Walker, J. T. (2008). The influence of sex offender registration and notification
laws in the United States. Crime & Delinquency, 54, 175–192.
Western, B. (2007). Punishment and inequality in America. Russell Sage.
Zevitz, R. G. (2006). Sex offender community notification: Its role in recidivism and offender reintegration.
Criminal Justice Studies, 19(2), 19–208.
Zgoba, K. M., & Clear, T. (2021). A Review of the Reality of Violent Offending and the Administration of
Justice. Criminal Justice Policy Review, 32(4), 352–373.
Zgoba, K., Witt, P., Dalessandro, M., & Veysey, B. M. (2008). Megan’s Law: Assessing the practical and
monetary efficacy (Document No. 225370). U.S. Department of Justice.
Zgoba, K., Veysey, B. M., & Dalessandro, M. (2010). An analysis of the effectiveness of community
notification and registration: Do the best intentions predict the best practices? Justice Quarterly, 27,
667–691.
Zgoba, K., Miner, M., Letourneau, E., Levenson, J., Knight, R., & Thornton, D. (2016). A multi-state
recidivism study using Static-99 and Static-2002 risk scores and tier guidelines from the Adam Walsh
Act. Sexual Abuse: A Journal of Research and Treatment, 28, 722–740.
Zgoba, K. M., Jennings, W. G., & Salerno, L. M. (2018). Megan’s law 20 years later: An empirical analysis
and policy review. Criminal Justice and Behavior, 45(7), 1028–1046.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.
Kristen M. Zgoba is an Assistant Professor of Criminology and Criminal Justice at Florida International
University. Prior she served as the research supervisor for the New Jersey Department of Corrections. Dr.
Zgoba has studied sexual offense legislation in the United States and overseas through the National Institute of
Justice and a Fulbright Research Scholar Award to the United Kingdom. She has published in the Prison
Journal, Race and Justice, Criminology & Public Policy, Sexual Abuse: A Journal of Research and
Treatment, Justice Quarterly and Criminal Justice and Behavior. She is also a member of the Research
Council within the American Correctional Association and is a founding member of the Sex Offense Policy &
Research group. Dr. Zgoba also serves on the Board of Directors at the Edna Mahan State Correctional
Facility for women in New Jersey.
Meghan Mitchell is an Assistant Professor of Criminal Justice at the University of Central Florida. She
received her PhD from Sam Houston State University. Her research examines corrections, subcultures, reentry,
and research methodologies. She has published in the British Journal of Criminology, Justice Quarterly,
Sociological Methods & Research, and the Journal of Criminal Justice, among others. She is a Co-PI on the
CCWORK Study—an interdisciplinary evaluation of correctional workers’ wellbeing, organizations, roles,
and knowledge, and is the PI for the translation to the United States. She is also the Research Coordinator for
the Florida Prison Education Project.

~ Springer