Content uploaded by Leslie Maaike Helmus
Author content
All content in this area was uploaded by Leslie Maaike Helmus on Feb 23, 2014
Content may be subject to copyright.
HIGH RISK SEX OFFENDERS
1
High Risk Sex Offenders May Not Be High Risk Forever
R. Karl Hanson
1
, Andrew J. R. Harris
2
, Leslie Helmus
3
, & David Thornton
4
1. Public Safety Canada, Ottawa, Ontario, Canada
2. Forensic Assessment Group, Ottawa, Ontario, Canada
3. Carleton University, Psychology Department, Ottawa, Ontario, Canada
4. Sand Ridge Secure Treatment Center, Mauston, Wisconsin, United States
Author Note
The views expressed are those of the authors and not necessarily those of Public Safety
Canada or the Wisconsin Department of Health Services.
We would like to thank the following researchers for granting us permission to use their
data: Alfred Allan, Susanne Bengtson, Jacques Bigras, Sasha Boer, Jim Bonta, Sébastien
Brouillette-Alarie, Franca Cortoni, Jackie Craissati, Margretta Dwyer, Reinhard Eher, Doug
Epperson, Tina Garby, Randolph Grace, Steve Gray, Andy Haag, Andreas Hill, Steve Johansen, Ray
Knight, Niklas Långström, Terry Nicholaichuk, Kevin Nunes, Jean Proulx, Martin Rettenberger,
Rebecca Swinburne Romine, Daryl Ternowski, Robin Wilson, and Annie Yessine.
Correspondence concerning this article should be addressed to R. Karl Hanson, Corrections
and Criminal Justice Directorate, Public Safety Canada, 10
th
floor, 340 Laurier Avenue West,
Ottawa, ON, Canada, K1A 0P8. Email: karl.hanson@ps-sp.gc.ca
RDIMS# 783489
Journal of Interpersonal Violence (in press, November 3, 2013)
HIGH RISK SEX OFFENDERS
2
Abstract
This study examined the extent to which sexual offenders present an enduring risk for sexual
recidivism over a 20 year follow-up period. Using an aggregated sample of 7,740 sexual offenders
from 21 samples, the yearly recidivism rates were calculated using survival analysis. Overall, the
risk of sexual recidivism was highest during the first few years after release, and decreased
substantially the longer individuals remained sex offence-free in the community. This pattern was
particularly strong for the high risk sexual offenders (defined by Static-99R scores). Whereas the 5
year sexual recidivism rate for high risk sex offenders was 22% from the time of release, this rate
decreased to 4.2% for the offenders in the same static risk category who remained offence-free in
the community for 10 years. The recidivism rates of the low risk offenders were consistently low
(1% to 5%) for all time periods. The results suggest that offence history is a valid, but time
dependent, indicator of the propensity to sexually reoffend. Further research is needed to explain
the substantial rate of desistance by high risk sexual offenders.
Keywords: sex offenders, risk assessment, desistance, recidivism
HIGH RISK SEX OFFENDERS
3
High Risk Sex Offenders May Not Be High Risk Forever
Of all people who commit serious transgressions, sexual offenders are perceived as the least
likely to change. The widespread implementation of long-term social controls that uniquely apply to
sexual offenders (e.g., lifetime community supervision, registration) indicates that policy-makers,
and the public that they represent, expect the risk posed by this population to persist almost
indefinitely. The reasons that sexual offenders are treated differently from other offenders are not
fully known. Contributing factors could include the particularly serious harm caused by sexual
victimization (Browne & Finkelhor, 1986; Resick, 1993), and the belief that there is “no cure” for
deviant sexual interests (e.g., Colorado Sex Offender Management Board, 2011). In certain public
discussions, the special status of sexual offenders is sometimes justified by reference to a
perceived high recidivism rate (see Ewing, 2011, p. 78).
Our belief that sexual offenders are intractable is in contrast to our openness to accept
change among other offenders. Although certain restrictions and prejudices apply to all persons
with a criminal record, the criminal justice systems of most Western democracies are predicated on
the assumption that virtually all offenders could and should be re-integrated into society as law-
abiding citizens. As articulated by Maruna and Roy (2007), the notion of personal reinvention by
“knifing off” an old self is deeply rooted in the American psyche, and, quite likely, many other
societies. It is an option, however, that is elusive to sexual offenders.
Sexual offenders vary in their risk for sexual recidivism. Previous meta-analyses have found
that the average sexual recidivism rates of identified sexual offenders are in the 7% to 15% range
after 5 to 6 years follow-up (Hanson & Morton-Bourgon, 2005; Helmus, Hanson, Thornton,
Babchishin, & Harris, 2012). In contrast, sex offenders defined as high risk by the Violence Risk
Scale – Sexual Offender Version (VRS-SO) have 10 year sexual recidivism rates between 56% and
70% (Beggs & Grace, 2010; Olver, Wong, Nicholaichuk, & Gordon, 2007).
Even if certain subgroups of sexual offenders can be identified as high risk, they need not
be high risk forever. Risk-relevant propensities could change based on fortunate life circumstances,
HIGH RISK SEX OFFENDERS
4
life choices, aging, or deliberate interventions (such as attending treatment). It is not necessary,
however, to prove that an offender has changed in order to revise a risk assessment. New
information could also be used to downgrade (or upgrade) an individual’s risk, even when the
reasons for the change are uncertain. Some of this information could be potentially available at the
time of the index sex offence (e.g., psychopathy scores), whereas other information is only
available later. In this paper, we focus on one objective indicator of post-index behaviour that
could be used to revise risk assessments: the length of time that individuals do not reoffend when
given the opportunity to do so.
General offenders are at greatest risk for new criminal behaviour immediately after release
(Blumstein & Nakamura, 2009; Bushway, Nieubeerta, & Blokland, 2011; Howard, 2011). The
longer they remain offence-free in the community, the lower their likelihood of ever again coming
in contact with the criminal justice system. Blumstein and Nakamura (2009) introduced the
concept of a redemption period, defined as the time at which an offender’s risk has declined
sufficiently that it is indistinguishable from the risk posed by men with no prior criminal record.
Similarly, Harris and Rice (2007) found that for most forensic psychiatric patients, the risk for
violent recidivism declined the longer they remained offence-free in the community. The reduction
in risk, however, was relatively modest, and did not apply to the highest risk offenders (defined by
Violence Risk Appraisal Guide [VRAG] bins of 7, 8 or 9).
Preliminary studies suggest that the overall time offence-free also applies to the risk of
sexual recidivism among sexual offenders. Harris and Hanson (2004) compared the recidivism
rates of a large sample of sexual offenders from the U.S., U.K., and Canada (n = 4,724) beginning
at four start dates: time of release, and after 5, 10, and 15 years offence-free in the community. In
their study, offence-free was defined as no new sexual or violent offences. They found that the
five-year recidivism rates were 14.0% from time of release, compared to 7.0% after 5 years, 5.4%
after 10 years, and 3.7% after 15 years offence-free. Similarly, Howard (2011) observed that the
risk of sexual recidivism declined over the four year follow-up period in his study. Neither Howard
HIGH RISK SEX OFFENDERS
5
(2011) nor Harris and Hanson (2004) examined whether the time-free effect applied equally to
sexual offenders at different initial risk levels.
Time-free adjustments for different risk levels (Static-99 risk categories) were presented by
Harris, Phenix, Hanson, and Thornton (2003; Appendix I). For each category of risk, the longer
they remained offence-free in the community (2 to 10 years), the lower their recidivism rate. For
example, the 5 year sexual recidivism for the Static-99 high risk group (scores of 6+) was 38.8%
from time of release but only 13.1% after 4 years offence-free. The decline, however, was not
completely consistent. For certain groups, the risk after 10 years offence-free was greater than the
risk after 6 years. Given the modest sample size (n < 30 for some cells), it was difficult to know
whether the observed variation was meaningful. Apart from Harris et al.’s (2003) preliminary
analyses by risk level, none of the previous studies have examined potential moderators of the
time-free effect, such as age and victim type (rapist/child molester).
The purpose of the current study was to examine the effects of time offence-free in the
community on the recidivism risk of sexual offenders. The study used an aggregate sample of
7,740 sexual offenders drawn from 21 different samples. Sexual recidivism rates were estimated
from time of release, and then after 5 years and 10 years sexual offence-free in the community.
Based on Static-99R scores (Helmus, Thornton, Hanson, & Babchishin, 2012), the sample was
divided into three risk categories: low, moderate (or typical), and high. As well, we examined a
number of other potential moderators of the time-free effect, including age at release, country of
origin, victim type (rapist/child molester), and exposure to treatment.
Method
Measures
Static-99R. Static-99R is a 10-item actuarial scale that assesses the recidivism risk of adult
male sex offenders. The items and scoring rules are identical to Static-99 (Hanson & Thornton,
2000; see also www.static99.org) with the exception of updated age weights (Helmus, Thornton, et
al., 2012). The 10 items cover demographics, sexual criminal history (e.g., prior sex offence), and
general criminal history (e.g., prior non-sexual violence).
HIGH RISK SEX OFFENDERS
6
Static-99/R are the most widely used sexual offender risk tools in mental health and
corrections (Archer, Buffington-Vollum, Stredny, & Handel, 2006; Interstate Commission for Adult
Offender Supervision, 2007; McGrath, Cumming, Burchard, Zeoli, & Ellerby, 2010). Static-99R has
high rater reliability (ICC = .89; McGrath, Lasher, & Cumming, 2012) and a moderate ability to
discriminate between sexual recidivists and non-recidivists (AUC = .69, 95% CI [.66, .72], k = 22,
n = 8,033; Helmus, Hanson, et al., 2012).
Rather than use the standard four risk categories (see Harris et al., 2003), only three risk
categories were used in order to maximize the sample size in each group (and increase the stability
of the results). The three risk categories were created based on percentile ranks (Hanson, Lloyd,
Helmus, & Thornton, 2012): specifically, scores one standard deviation below the population mean
were considered “low” (-3, -2, -1), scores one standard deviation above the mean were considered
“high” (5 and higher), and the remaining scores were considered “moderate” (0, 1, 2, 3, 4).
Samples
Twenty-one samples were selected from those used by Helmus and colleagues to re-norm
the Static-99/R (Helmus, 2009; Helmus, Hanson, et al., 2012; Helmus, Thornton, et al., 2012); of
the 23 samples with Static-99R data available, one was excluded because it did not have the
information needed to compute survival analyses, and one was excluded because it was identified
as a statistical outlier in previous research (Helmus, Hanson, et al., 2012). The data retained for
analysis contained 7,740 offenders from 21 samples. A brief description of the included studies can
be found in Table 1.
Overview of Analyses
The recidivism rates were estimated using life table survival analysis (Singer & Willet, 2003;
Soothill & Gibbens, 1978). In this approach, the follow-up time is divided into discrete time
intervals (12 months), and the proportion failing (reoffending) in each time interval is calculated.
This quantity is referred to as a hazard rate, or the probability of reoffending in a specific time
interval given that the individual has survived (not reoffended) up to that time.
HIGH RISK SEX OFFENDERS
7
The only type of recidivism examined in the current study was sexual recidivism.
Consequently, statements concerning the length of time that individuals were “offence-free” should
be interpreted as meaning that no new sexual offences were detected during that time period.
The 95% confidence interval for the observed proportions were calculated using Wald’s
method: CI ± 1.96(p(1-p)/n)
1/2
(Agresti & Coull, 1998). Proportions were interpreted as different
when their 95% confidence intervals did not overlap, which corresponds to a difference test of
approximately p < .01 (Cumming & Finch, 2005).
Results
Without controlling for time at risk, the observed sexual recidivism rate for all cases was
11.9% (n = 7,740), 2.9% for the low risk cases (n = 890), 8.5% for the moderate cases (n =
4,858), and 24.2% for the high risk cases (n = 1,992). The average follow-up period was 8.2 years
(SD = 5.2, range of 0.01 to 31.5).
Figure 1 plots the cumulative survival rates over time for the three risk categories. The
survival curves were truncated when there were fewer than 50 offenders at the end of the at-risk
period (between 20 & 25 years). As can be seen from Figure 1, the risk of reoffending was highest
in the first few years following release, and declined thereafter. This pattern was particularly strong
for the high risk offenders. During the first year after release, 7% reoffended, and during the first 5
years after release, a total of 22% reoffended. . In contrast, during the next five years (between
year 5 and year 10), the survival curve descended only 7% (from 78% to 71%) representing
yearly rates in the 1% to 2% range. No high risk sexual offender in this sample reoffended after 16
years offence-free (126 high risk cases started year 17, of which 61 were followed for 5 years or
more). The cumulative survival function indicated that the long-term recidivism rate for the high
risk offenders was approximately 32% starting from time of release.
Figure 2 and Figure 3 plot the cumulative survival rates for offenders who remained sexual
offence-free for 5 years or 10 years, respectively. Summaries of the data from Figures 1 through 3
are presented in Table 2. The high risk offenders still reoffended more quickly than the other
groups, but the recidivism rates for all groups were substantially lower than for offenders at time of
HIGH RISK SEX OFFENDERS
8
release. Whereas the 10 year sexual recidivism rate of the high risk offenders from time of release
was 28.8%, the rate declined to 12.5% for those who remained offence-free for 5 years, then
6.2% for those who remained offence-free for 10 years (see Table 2). A 10 year sexual recidivism
rate of 6.2% for the high risk group (10 years offence-free) was less than the expected rate of
moderate risk offenders from time-at-release (10.4%).
Inspection of Table 2 indicates that the expected recidivism rates were approximately cut in
half for each 5 years that the offender was sexual offence-free in the community. For example, the
5 year sexual recidivism rate of the high risk groups was 22.0% at release, 8.6% after 5 years,
and 4.2% after 10 years offence-free. The same pattern applied to the moderate risk offenders
(and the full sample). In contrast, the recidivism rates for the low risk offenders were consistently
low (1% - 5%), and did not change meaningfully based on years offence-free. For example, the
10 year sexual recidivism rate for the low risk offenders was 3.1% from time of release and 3.4%
for those who remained offence-free in the community for 10 years.
Table 3 compares the observed recidivism rate for the first five years with the recidivism
rates for years 6-10 and years 11-15. These comparisons are reported as risk ratios, with the rates
for subsequent 5-year periods divided by the rate for the first five years after release. For example,
a risk ratio of 0.50 would indicate that the recidivism rate was cut in half, and a rate of 0.25 would
indicate that the recidivism rate was ¼ the initial rate. All rate estimates were created from life
table survival analysis.
As can be seen in Table 3, the time-free effect was similar across the various subgroups
examined, including those defined by age at release, treatment involvement, pre-selected high
risk/high need, country, year of release, and victim type (adults, children, related children). As
expected, there were meaningful differences in the initial recidivism rates; however, the relative
risk reductions were similar across all subgroups. The risk ratios comparing the rates for years 6-
10 to years 1–5 were tightly clustered between 0.33 and 0.59 (median of 0.46). The risk ratios
comparing years 11 – 15 to years 1- 5 varied between 0.07 and 0.36, with the exception of the low
risk group, which had a risk ratio of 0.78 (median of 0.28).
HIGH RISK SEX OFFENDERS
9
Discussion
The purpose of this study was to examine the extent to which high risk sexual offenders
remain high risk over time. As has been found for general offenders and violent offenders, the risk
of sexual recidivism was highest in the first few years after release, and then decreased the longer
they remained offence-free in the community. The decline in hazard rates was greatest for sexual
offenders who had been identified as high risk at time of release. For low risk offenders, time free
had little influence: their risk was consistently low (1% to 5%). The same relative risk reductions
were observed for subgroups categorized by age at release, treatment involvement, country, and
victim type.
The current findings indicate static risk factors (e.g., prior offences, victim characteristics)
are valid, but time dependent, markers for risk relevant propensities. If high risk sexual offenders
do not reoffend when given the opportunity to do so, then there is clear evidence that they are not
as high risk as initially perceived. The current study found that, on average, their recidivism risk
was cut in half for each 5 years that they remained offence-free in the community.
Risk predictions describe lives that have yet to be fully lived; consequently, the more we
know of an offender’s life, the easier it is to predict the remainder. At the time of release, the best
estimate of the likelihood of recidivism is the base rate for the group that the offender most closely
resembles (i.e., offenders with the same risk score). Once given the opportunity to reoffend, the
individuals who reoffend should be sorted into higher risk groups, and those who do not reoffend
should be sorted into lower risk groups. This sorting process can result in drastic changes from the
initial risk estimates. Based on the current results, for example, 22 out of 100 high risk offenders
would be expected to be charged or convicted of a new sexual offence during the ten years
following release. In contrast, the rate would be 4 out of 100 for those who survive sexual offence-
free for 10 years. This low recidivism rate among the survivors suggests that their initial
designation as “high risk” sexual offenders was either incorrect, or that something has changed.
The current study did not address the reasons for the strong empirical association between
years crime-free and desistance. There are several different mechanisms that could lead to this
HIGH RISK SEX OFFENDERS
10
effect. The study did not directly address whether the offenders remaining offence-free were
different individuals from the recidivists. Consequently, any apparent “effect” of time offence-free
could be attributed to pre-existing differences between offenders. Given that criminal history
variables (including Static-99R scores) are fallible indicators of risk relevant propensities, some
individuals who have a conviction for a sexual offence (or even a high Static-99R score) may never
have had an enduring propensity toward sexual crime in the first place.
It is also possible that certain high risk offenders genuinely changed. All the offenders in the
current study had been convicted of at least one sexual offence, which would indicate a non-
negligible risk at one time. Furthermore, it would be difficult to get a high score (5+) on Static-99R
without an extended period of engaging in sexual and general crime. Nevertheless, a substantial
portion of the high risk offenders survived throughout the complete follow-up period without any
new crimes being detected. Given that it is likely that at least some of the offenders changed in a
prosocial direction, further research is needed to increase our capacity to distinguish between
desisters and future recidivists.
The only type of recidivism examined in the current study was sexual recidivism (as
measured by charges and convictions). Consequently, it is quite likely that evaluators would have
increased capacity to discriminate recidivists from non-recidivists by monitoring ongoing
involvement in non-sexual crime, and by measuring indicators of commitment to prosocial goals. In
particular, structured methods for evaluating sexual offenders’ criminogenic needs have been
demonstrated to be incremental to Static-99/R in the prediction of sexual recidivism for prison
samples (Beggs & Grace, 2010; Knight & Thornton, 2007; Olver et al., 2007) and community
samples (McGrath et al., 2012).
Even if the reasons for the reduced risk over time are not fully known, the current results
have clear implications for the community supervision of sexual offenders. Following Andrews and
Bonta’s (2010) risk principle, high risk sexual offenders should receive the most intensive service
and monitoring during the early part of their community sentence. Subsequently, the intensity of
interventions could decline to the level normally applied to moderate risk individuals when
HIGH RISK SEX OFFENDERS
11
offenders who were initially high risk remain offence-free for several years.
The current findings also suggest that certain long-term supervision and monitoring policies
(e.g., lifetime registration) may be being applied to a substantial number of individuals with a low
risk for sexual offending. Although the moral consequences of sexual offending may last forever,
the current results suggest that sexual offenders who remain offence-free could eventually cross a
“redemption” threshold in terms of recidivism risk, such that their current risk for a sexual crime
becomes indistinguishable from the risk presented by non-sexual offenders.
Previous large sample studies have found that the likelihood of an “out of the blue” sexual
offence to be committed by offenders with no history of sexual crime is 1% to 3%: 1.1% after 4
years (Duwe, 2012); 1.3% after 3 years (Langan, Schmitt, & Durose, 2003); 3.2% after 4.5 years
(Wormith, Hogg, & Guzzo, 2012). In comparison, only 2 of 100 moderate risk sexual offenders in
the current study committed a new sexual offence during a five year follow-up period if they were
able to remain 10 years offence-free in the community. The high risk offenders in the current
sample, however, never fully resembled nonsexual offenders. Although their recidivism rates
declined substantially when they were 10 years offence-free, the five year recidivism rate of the
initially high risk offenders (4.2%) was still higher than the expected rate for nonsexual offenders
(1%-3%).
Limitations
The current results were predicated on the assumption that release to the community
provided opportunities for offending. However, it is possible that certain forms of conditional
release are sufficiently confining as to meaningfully limit opportunities (e.g., house arrest). The
nature of the supervision conditions of the offenders in the current study were not fully known;
however, given the typical practices in the jurisdictions for these time periods, it would be likely
that the offenders had real opportunities to reoffend once released to the community.
Some evidence that supervision practices may moderate the time-free effect is provided in
a recent study by Zgoba et al. (2012). This follow-up study of 1,789 adult sex offenders from four
states (Minnesota, New Jersey, Florida, and South Carolina) did not find that risk declined with
HIGH RISK SEX OFFENDERS
12
time in the community. Overall, there was a constant hazard rate of 1% per year for first 10 years
(e.g., 5% after 5 years; 10% after 10 years). The reasons for the constant hazard rate is not
known, but could be related to strict supervision practices and high rates of technical breaches
observed in these samples.
Another limitation is that recidivism was measured by officially recorded charges or
convictions. It is well known that official records as an indicator of recidivism have high specificity
(those identified are most likely guilty) but low sensitivity (many offences are undetected). Even if
the detection rate per offence is low, however, the detection rate per offender could be high if
offenders commit multiple offences. As well, the most serious offences are those most likely to be
reported to the police (Fisher, Daigle, Cullen, & Turner, 2003).
Conclusions
This study found that sexual offenders’ risk of serious and persistent sexual crime decreased
the longer they had been sex offence-free in the community. This pattern was particularly evident
for high risk sexual offenders, whose yearly recidivism rates declined from approximately 7%
during the first calendar year, to less than 1% per year when they have been offence-free for 10
years or more. Consequently, intervention and monitoring resources should be concentrated in the
first few years after release, with diminishing attention and concern for individuals who remain
offence-free for substantial periods of time.
HIGH RISK SEX OFFENDERS
13
References
References marked with an asterisk indicate studies included in the analyses. Wilson et al.
(2007a & b) both have an asterisk, although they were combined into one sample for the
current study.
Agresti, A., & Coull, B. A. (1998). Approximate is better than “exact” for interval estimation of
binomial proportions. The American Statistician, 52, 119-126.
doi:10.1080/00031305.1998.10480550
*Allan, M., Grace, R. C., Rutherford, B., & Hudson, S. M. (2007). Psychometric assessment of
dynamic risk factors for child molesters. Sexual Abuse: A Journal of Research and
Treatment, 19, 347-367. doi:10.1007/s11194-007-9052-5
Andrews, D. A., & Bonta, J. (2010). The psychology of criminal conduct (5
th
ed.). Newark, NJ:
LexisNexus/Anderson.
Archer, R. P., Buffington-Vollum, J. K., Stredny, R. V., & Handel, R. W. (2006). A survey of
psychological test use patterns among forensic psychologists. Journal of Personality
Assessment, 87, 84-94. doi:10.1207/s15327752jpa8701_07
*Bartosh, D. L., Garby, T., & Lewis, D., & Gray, S. (2003). Differences in the predictive validity of
actuarial risk assessments in relation to sex offender type. International Journal of Offender
Therapy & Comparative Criminology, 47, 422-438. doi:10.1177/0306624X03253850
Beggs, S. M., & Grace, R. C. (2010). Assessment of dynamic risk factors: An independent
validation study of the Violence Risk Scale: Sexual Offender Version. Sexual Abuse: A
Journal of Research and Treatment, 22, 234-251. doi:10.1177/1079063210369014
*Bengtson, S. (2008). Is newer better? A cross-validation of the Static-2002 and the Risk Matrix
2000 in a Danish sample of sexual offenders. Psychology, Crime & Law, 14, 85-106.
doi:10.1080/10683160701483104
*Bigras, J. (2007). La prédiction de la récidive chez les délinquants sexuels [Prediction of
recidivism among sex offenders]. Dissertations Abstracts International, 68 (09). (UMI No.
NR30941).
HIGH RISK SEX OFFENDERS
14
Blumstein, A., & Nakamura, K. (2009). Redemption in the presence of widespread criminal
background checks. Criminology, 47, 327-359. doi:10.1111/j.1745-9125.2009.00155.x
*Boer, A. (2003). Evaluating the Static-99 and Static-2002 risk scales using Canadian sexual
offenders. Unpublished master’s thesis, University of Leicester, Leicester, United Kingdom.
*Bonta, J., & Yessine, A. K. (2005). [Recidivism data for 124 released sexual offenders from the
offenders identified in The National Flagging System: Identifying and responding to high-
risk, violent offenders (User Report 2005-04). Ottawa: Public Safety and Emergency
Preparedness Canada]. Unpublished raw data.
*Brouillette-Alarie, S., & Proulx, J. (2008, October). Predictive and convergent validity of
phallometric assessment in relation to sexual recidivism risk. Poster presented at the annual
conference for the Association for the Treatment of Sexual Abusers, Atlanta, GA.
Browne, A., & Finkelhor, D. (1986). Impact of child sexual abuse: A review of the research.
Psychological Bulletin, 99, 66-77. doi:10.1037/0033-2909.99.1.66
Bushway, S. D., Nieubeerta, P., & Blokland, A. (2011). The predictive value of criminal background
checks: Do age and criminal history affect time to redemption? Criminology, 49, 27-60.
doi:10.1111/j.1745-9125.2010.00217.x
Colorado Sex Offender Management Board (CSOMB). (2011). Standards and Guidelines for the
Evaluation, Assessment, Treatment, and Supervision of Juveniles who have Committed
Sexual Offenses. Colorado Department of Public Safety, Division of Criminal Justice.
Available at
http://dcj.state.co.us/odvsom/sex_offender/SO_Pdfs/2012%20ADULT%20STANDARDS%20
FINAL%20C.pdf
*Cortoni, F., & Nunes, K. L. (2007). Assessing the effectiveness of the National Sexual Offender
Program (Research Report No. R-183). Unpublished report, Correctional Service of Canada.
*Craissati, J., Bierer, K., & South, R. (2011). Risk, reconviction, and “sexually risky behaviour” in
sex offenders. Journal of Sexual Aggression, 17, 153-165.
doi:10.1080/13552600.2010.490306
HIGH RISK SEX OFFENDERS
15
Cumming, G., & Finch, S. (2005). Inference by eye: Confidence intervals and how to read pictures
of data. American Psychologist, 60, 170-180. doi:10.1037/0003-066X.60.2.170
Duwe, G. (2012). Predicting first-time sexual offending among prisoners without a prior sex offense
history: The Minnesota Sexual Criminal Offending Risk Estimate (MnSCORE). Criminal
Justice and Behavior, 39, 1436-1456. doi:10.1177/0093854812453911
*Eher, R., Rettenberger, M., Schilling, F., & Pfafflin, F. (2009). [Data from sex offenders released
from prison in Austria]. Unpublished raw data.
*Epperson, D. L. (2003). Validation of the MnSOST-R, Static-99, and RRASOR with North Dakota
prison and probation samples. Unpublished Technical Assistance Report, North Dakota
Division of Parole and Probation.
Ewing, C. P. (2011). Justice perverted: Sex offense law, psychology, and public policy. New York:
Oxford University Press.
Fisher, B. S., Daigle, L. E., Cullen, F. T., & Turner, M. G. (2003). Reporting sexual victimization to
the police and others: Results from a national-level study of college women. Criminal Justice
and Behavior, 30, 6-38. doi:10.1177/0093854802239161
*Haag, A. M. (2005). [Recidivism data from 198 offenders detained until their warrant expiry date.
From: Do psychological interventions impact on actuarial measures: An analysis of the
predictive validity of the Static-99 and Static-2002 on a re-conviction measure of sexual
recidivism. Dissertations Abstracts International, 66 (08), 4531B. (UMI No. NR05662)].
Unpublished raw data.
*Hanson, R. K., Harris, A. J. R., Scott, T., & Helmus, L. (2007). Assessing the risk of sexual
offenders on community supervision: The Dynamic Supervision Project (Corrections
Research User Report No. 2007-05). Ottawa, ON, Canada: Public Safety Canada.
Hanson, R. K., Lloyd, C., Helmus, L., & Thornton, D. (2012). Developing non-arbitrary metrics for
risk communication: Percentile ranks for the Static-99/R and Static-2002/R sexual offender
risk tools. International Journal of Forensic Mental Health, 11, 9-23.
doi:10.1080/14999013.2012.667511
HIGH RISK SEX OFFENDERS
16
Hanson, R. K., & Morton-Bourgon, K. E. (2005). The characteristics of persistent sexual offenders:
A meta-analysis of recidivism studies. Journal of Consulting and Clinical Psychology, 73,
1154-1163. doi:10.1037/0022-006X.73.6.1154
Hanson, R. K., & Thornton, D. (2000). Improving risk assessments for sex offenders: A comparison
of three actuarial scales. Law and Human Behavior, 24, 119-136.
doi:10.1023/A:1005482921333
Harris, A. J. R., & Hanson, R. K. (2004). Sex offender recidivism: A simple question (User Report
2004-03). Ottawa, ON: Public Safety and Emergency Preparedness Canada.
Harris, A. J. R., Phenix, A., Hanson, R. K., & Thornton, D. (2003). Static-99 coding rules: Revised
2003. Ottawa, ON: Solicitor General Canada.
Harris, G. T., & Rice, M. E. (2007). Adjusting actuarial violence risk assessments based on aging or
the passage of time. Criminal Justice and Behavior, 34, 297-313.
doi:10.1177/0093854806293486
Helmus, L. (2009). Re-norming Static-99 recidivism estimates: Exploring base rate variability
across sex offender samples (Master’s thesis). Available from ProQuest Dissertations and
Theses database. (UMI No. MR58443). Also available from www.static99.org
Helmus, L., Hanson, R. K., Thornton, D., Babchishin, K. M., & Harris, A. J. R. (2012). Absolute
recidivism rates predicted by Static-99R and Static-2002R sex offender risk assessment
tools vary across samples: A meta-analysis. Criminal Justice and Behavior, 39, 1148-1171.
doi:10.1177/0093854812443648
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: Journal of Research and Treatment, 24, 64-101 doi:10.1177/1079063211409951
*Hill, A., Habermann, N., Klusmann, D., Berner, W., & Briken, P. (2008). Criminal recidivism in
sexual homicide perpetrators. International Journal of Offender Therapy and Comparative
Criminology, 52, 5-20. doi:10.1177/0306624X07307450
Howard, P. (2011). Hazards of different types of reoffending. Ministry of Justice Research Series
HIGH RISK SEX OFFENDERS
17
3/11. London, UK: UK Ministry of Justice. Available at
www.justice.gov.uk/publications/research.htm
Interstate Commission for Adult Offender Supervision. (2007). Sex offender assessment
information survey (ICAOS Documents No. 4-2007). Lexington, KY: Author.
*Johansen, S. H. (2007). Accuracy of predictions of sexual offense recidivism: A comparison of
actuarial and clinical methods. Dissertations Abstracts International, 68 (03), B. (UMI No.
3255527).
*Knight, R. A., & Thornton, D. (2007). Evaluating and improving risk assessment schemes for
sexual recidivism: A long-term follow-up of convicted sexual offenders (Document No.
217618). Submitted to the U.S. Department of Justice.
Langan, P. A., Schmitt, E. L., & Durose, M. R. (2003). Recidivism of sex offenders released from
prison in 1994. Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics.
*Langström, N. (2004). Accuracy of actuarial procedures for assessment of sexual offender
recidivism risk may vary across ethnicity. Sexual Abuse: A Journal of Research and
Treatment, 16, 107-120. doi:10.1177/107906320401600202
Maruna, S., & Roy, K. (2007). Amputation or reconstruction? Notes on the concept of “knifing off”
and desistance from crime. Journal of Contemporary Criminal Justice, 23, 104-124.
doi:10.1177/1043986206298951
McGrath, R. J., Cumming, G. F., Burchard, B. L., Zeoli, S., & Ellerby, L. (2010). Current practices
and emerging trends in sexual abuser management: The Safer Society 2009 North
American survey. Brandon, VT: Safer Society Foundation.
McGrath, R. J., Lasher, M. P., & Cumming, G. F. (2012). The Sex Offender Treatment Intervention
and Progress Scale (SOTIPS): Psychometric properties and incremental predictive validity
with Static-99R. Sexual Abuse: A Journal of Research and Treatment, 24, 431-458.
doi:10.1177/1079063211432475
*Nicholaichuk, T. (2001, November). The comparison of two standardized risk assessment
instruments in a sample of Canadian Aboriginal sexual offenders. Paper presented at the
HIGH RISK SEX OFFENDERS
18
annual Research and Treatment Conference of the Association for the Treatment of Sexual
Abusers, San Antonio, TX.
Olver, M. E., Wong, S. C. P., Nicholaichuk, T., & Gordon, A. (2007). The validity and reliability of
the Violence Risk Scale – Sexual Offender Version: Assessing sex offender risk and
evaluating therapeutic change. Psychological Assessment, 19, 318-329. doi:10.1037/1040-
3590.19.3.318
Resick, P. A. (1993). The psychological impact of rape. Journal of Interpersonal Violence, 8, 223-
255. doi:10.1177/088626093008002005
Singer, J. D., & Willet, J. B. (2003). Applied longitudinal data analysis: Modeling change and event
occurrence. Oxford: Oxford University Press.
Soothill, K. L., & Gibbens, T. C. N. (1978). Recidivism of sexual offenders: A re-appraisal. British
Journal of Criminology, 18, 267-276.
*Swinburne Romine, R., Dwyer, S. M., Mathiowetz, C., & Thomas, M. (2008, October). Thirty years
of sex offender specific treatment: A follow-up Study. Poster presented at the conference
for the Association for the Treatment of Sexual Abusers, Atlanta, GA.
*Ternowski, D. R. (2004). Sex offender treatment: An evaluation of the Stave Lake Correctional
Centre Program. Dissertations Abstracts International, 66 (06), 3428B. (UMI No. NR03201).
*Wilson, R. J., Cortoni, F., & Vermani, M. (2007a). Circles of support and accountability: A national
replication of outcome findings (Report No. R-185). Ottawa, ON: Correctional Service of
Canada.
*Wilson, R. J., Picheca, J. E., & Prinzo, M. (2007b). Evaluating the effectiveness of professionally-
facilitated volunteerism in the community-based management of high-risk sexual offenders:
Part two – A comparison of recidivism rates. The Howard Journal, 46, 327-337.
doi:10.1111/j.146
Wormith, J. S., Hogg, S., & Guzzo, L. (2012). The predictive validity of a general risk/needs
assessment inventory on sexual offender recidivism and an exploration of the professional
override. Criminal Justice and Behavior, 39, 1511-1538. doi:10.1177/0093854812455741
HIGH RISK SEX OFFENDERS
19
Zgoba, K. M., Miner, M., Knight, R., Letourneau, E., Levenson, J., & Thornton, D. (2012). A multi-
site recidivism study using Static-99R and Static-2002 risk scores and tier guidelines from
the Adam Walsh Act. Report submitted to the US National Institute of Justice.
https://www.ncjrs.gov/pdffiles1/nij/grants/240099.pdf
HIGH RISK SEX OFFENDERS
20
Table 1
Descriptive Information
Study
n
Age
M (SD)
Country
5-Year
Recid
(%)
Recidivism
Criteria
Type of Sample
Mostly
Treated
Release
Period
Mdn
Year
Release
Allan et al. (2007)
492
42 (12)
NZ
9.8
Charges
Prison treatment
Yes
1990-2000
1994
Bartosh et al. (2003)
186
38 (12)
U.S.
11.8
Charges
Routine correctional
-
1996
1996
Bengtson (2008)
311
33 (10)
Denmark
19.6
Charges
Forensic psychiatric
-
1978-1995
1986
Bigras (2007)
483
43 (12)
Canada
7.4
Charges
Correctional Service of Canada
Mixed
1995-2004
1999
Boer (2003)
299
41 (12)
Canada
3.7
Conviction
Correctional Service of Canada
-
1976-1994
1990
Bonta & Yessine (2005)
133
40 (10)
Canada
17.3
Conviction
Preselected high risk
Mixed
1992-2004
1999
Brouillette-Alarie & Proulx
(2008)
228
36 (10)
Canada
14.2
Conviction
Prison & community treatment
-
1979-2006
1996
Cortoni & Nunes (2007)
73
42 (12)
Canada
0.0
Charges
Prison treatment
Yes
2001-2004
2003
Craissati et al. (2008)
209
38 (12)
U.K.
6.7
Conviction
Community supervision
Mixed
1992-2005
1998
Eher et al. (2008)
706
41 (12)
Austria
4.9
Conviction
European prison
-
2000-2005
2003
Epperson (2003)
177
37 (13)
U.S.
11.3
Charges
Routine correctional
-
1989-1998
1995
Haag (2005)
198
37 (10)
Canada
19.7
Conviction
Preselected high risk
Mixed
1995
1995
Hanson et al. (2007)
702
42 (13)
Canada
8.7
Charges
Community supervision
-
2001-2005
2002
Hill et al. (2008)
86
39 (11)
Germany
9.6
Conviction
Sexual homicide perpetrators
-
1971-2002
1989
Johansen (2007)
273
38 (11)
U.S.
5.5
Charges
Prison treatment
Yes
1994-2000
1996
Knight & Thornton (2007)
466
36 (11)
U.S.
23.3
Charges
Civil commitment evaluation
-
1957-1986
1970
Långström (2004)
1,278
41 (12)
Sweden
5.4
Conviction
Routine European prison
No
1993-1997
1995
Nicholaichuk (2001)
281
35 (9)
Canada
26.3
Conviction
High intensity treatment
Yes
1983-1998
1992
Swinburne Romine et al.
(2008)
680
38 (12)
U.S.
8.8
Conviction
Community treatment
Mixed
1977-2007
1988
Ternowski (2004)
247
44 (13)
Canada
6.5
Charges
Prison treatment
Yes
1994-1998
1996
Wilson et al. (2007a & b)
232
42 (11)
Canada
12.4
Charges
Preselected high risk
-
1994 -2007
2002
Total
7,740
40 (12)
10.1
1957-2007
1996
Note. 5 year sexual recidivism rates were obtained from survival analysis. All samples had >50 cases at the beginning of the 5-year interval.
HIGH RISK SEX OFFENDERS
21
Table 2
Sexual Recidivism Rates From Survival Analyses (Including Confidence Intervals)
N at start of
follow-up
5 Years Follow-up
10 Years Follow-up
15 Years Follow-up
%
95% CI
(n)
%
95% CI
(n)
%
95% CI
(n)
Complete Sample
From Release
7,740
10.1
9.4
10.8
(4,735)
14.2
13.3
15.2
(1,847)
16.6
15.4
17.9
(755)
5 Years offence free
4,735
4.6
3.9
5.4
(1,847)
7.3
6.1
8.5
(755)
9.0
7.5
10.5
(420)
10 Years offence free
1,847
2.8
1.8
3.8
(755)
4.6
3.1
6.0
(420)
4.8
3.3
6.3
(102)
Low (scores of -3 to -1)
From Release
890
2.2
1.2
3.2
(601)
3.1
1.8
4.4
(234)
4.7
2.1
7.4
(88)
5 Years offence free
601
0.95
0.12
1.8
(234)
2.6
0.12
5.1
(88)
4.3
0.23
8.4
(53)
10 Years offence free
234
1.7
0.0
4.1
(88)
3.4
0.0
7.4
(53)
-
-
-
-
Moderate (scores of 0 to 4)
From Release
4,858
6.7
5.9
7.4
(3,081)
10.4
9.3
11.4
(1,175)
12.6
11.1
14.0
(496)
5 Years offence free
3,081
4.0
3.1
4.8
(1,175)
6.3
4.9
7.7
(496)
8.0
6.1
9.8
(280)
10 Years offence free
1,175
2.4
1.2
3.6
(496)
4.2
2.4
5.9
(280)
4.5
2.7
6.4
(69)
High (scores of 5+)
From Release
1,992
22.0
20.1
24.0
(1,053)
28.8
26.4
31.1
(438)
31.8
29.0
34.5
(171)
5 Years offence free
1,053
8.6
6.6
10.6
(438)
12.5
9.6
15.3
(171)
14.3
10.8
17.7
(87)
10 Years offence free
438
4.2
2.0
6.4
(171)
6.2
3.1
9.3
(87)
-
-
-
-
Notes. - Indicates insufficient numbers to make useful estimates (< 50 cases per cell). Each column presents information for a specified
follow-up period (i.e., 5, 10, or 15 years). The rows denote when the follow-up period starts. For example, the second row of data is for
offenders in the overall sample who did not commit a sex offence in the first five years. The 5-year follow-up data for these offenders starts
after their five years of offence-free survival in the community (i.e., it reflects recidivism rates 10 years from their initial release date).
HIGH RISK SEX OFFENDERS
22
Table 3
Relative reductions in sexual recidivism based on comparing the rate during the first 5 years in the community with the 5-year
rates starting after 5 and 10 offence-free years in the community.
Sample size
at start of
follow-up
Initial
5-Year Recidivism Rate
(Years 1-5)
Relative rate
after 5 years
offence-free
(Years 6-10)
Relative rate
after 10 years
offence-free
(Years 11 -15)
%
(n)
Risk Ratio
(n)
Risk Ratio
(n)
Complete Sample
7,740
10.1
(4,735)
0.46
(1,847)
0.28
(755)
Risk Level (Static-99R scores)
Low (scores of -3 to -1)
890
2.2
(601)
0.44
(234)
0.78
(88)
Moderate (scores of 0 to 4)
4,858
6.7
(3,081)
0.59
(1,175)
0.36
(496)
High (scores of 5+)
1,992
22.0
(1,053)
0.39
(438)
0.19
(171)
Age at Release
Immature (18 to 30 years)
1,818
13.74
(1,130)
0.46
(524)
0.31
(260)
Young (30 to 50 years)
4,434
10.07
(2,719)
0.44
(1,051)
0.21
(411)
Prime of Life (50+ years)
1,488
5.44
(866)
0.52
(272)
0.31
(84)
Sample Type
Routine Correctional
4,040
6.73
(2,248)
0.55
(671)
-
Pre-selected Treatment
1,920
8.85
(1,442)
0.46
(642)
0.32
(420)
Pre-selected High Risk/Needs
1,621
20.42
(963)
0.37
(491)
0.16
(294)
Country
United States
1,782
12.70
(1,318)
0.33
(810)
0.15
(552)
Canada
2,875
11.10
(1,298)
0.48
(379)
0.16
(55)
Other
3,082
7.63
(2,118)
0.60
(658)
-
Year of Release (sample median)
1970 - 1995
4,268
11.38
(3,278)
0.42
(1,628)
0.24
(734)
1996 - 2003
3,472
8.40
(1,457)
0.47
(219)
-
Table continues
HIGH RISK SEX OFFENDERS
23
Table 3 continued
Victim Type
Adults (rape)
2,182
9.95
(1,262)
0.45
(443)
0.24
(102)
Children (all child molesters)
3,188
8.59
(1,887)
0.42
(807)
0.19
(351)
Related children (incest)
1,539
4.17
(985)
0.50
(418)
0.07
(179)
Notes. In the two right-hand columns the “rate” represents the 5-year recidivism percentage observed in either the “after 5
years” or “after 10 years” offence free in the community{as seen in Table 2} divided by the observed recidivism rate in the first
5 years in the community. Using the “Moderate” Static-99R row as an example, the expected 5-year recidivism rate for the
initial sample (n = 4,858) is 6.68%. For those who did not reoffend in the first five years (n = 3,081), between the 6
th
and 10
th
year of follow-up the recidivism rate for this group is 3.96%. The 5-year recidivism rate for those who survived the first 5
years (3.96%) is then divided by the initial 5-year recidivism rate (6.68%) (3.96/6.68 = 0.59) which is the risk ratio included in
the table. This finding indicates that the recidivism rate for men with “Moderate” Static-99R scores during the period between
years 6 and 10 of follow-up has reduced to about 60% of what it was during the first 5 years of release. This method of
calculation is used throughout Table 3.
HIGH RISK SEX OFFENDERS
24
Figure 1. Time to Sexual Recidivism by Risk Level
0.5
0.55
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
P
r
o
p
o
r
t
i
o
n
Years of Follow-up
Low Moderate High
Low n = 890 Moderate n = 4,858 High n = 1,992
HIGH RISK SEX OFFENDERS
25
Figure 2. Time to Sexual Recidivism after Five Years Sex Offence-Free in the
Community by Risk Level
0.5
0.55
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
P
r
o
p
o
r
t
i
o
n
Years of Follow-up
Low Moderate High
Low n = 601 Moderate n = 3,081 High n = 1,053
HIGH RISK SEX OFFENDERS
26
Figure 3. Time to Sexual Recidivism after Ten Years Sex Offence-Free in the Community
by Risk Level
0.5
0.55
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95
1
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
P
r
o
p
o
r
t
i
o
n
Years of Follow-up
Low Moderate High
Low n = 234 Moderate n = 1,175 High n = 438
