Documente Academic
Documente Profesional
Documente Cultură
*
LISA L. SAMPLE
University of Nebraska at Omaha
TIMOTHY M. BRAY
University of Texas at Dallas
Research Summary:
Current legislation mandating DNA collection, civil commitment,
registration, and community notification of sex offenders is predicated
on the assumption that sex offenders are simply more dangerous than
other types of offenders in that they inevitably re-offend. Moreover,
many states are moving to expand sex offender legislation to include
non-sexual offenders on the assumption that some offense types, such
as burglary and robbery, serve as “gateway” offenses to sex crimes.
The purpose of this research is to highlight two of the common percep-
tions underlying sex offender laws, and the extension thereoJ and
examine them in light of current empirical evidence. We employ analy-
sis of variance techniques on Illinois arrest data from 1990 to 1997 to
examine the degree to which sex offenders have higher proportions of
repeat offending than other criminal categories and if some offense
types serve as “gateway or predicate offenses to sex crimes.
I’
Policy Implications:
Our results suggest that the extension of sex offender laws to non-
sexual offenders will likely have little effect on sexual victimization
rates. More importantly, our results illustrate that policies can be
founded on misconceptions, and these misconceptions not only have
financial consequences, but also can affect the likelihood that the poli-
cies enacted will achieve their goals. If nothing else, this research sug-
gests that policy makers need to become better informed on the issues
they subject to far-reaching and costly legislation.
*The authors wish to thank the Harry Frank Guggenheim Foundation and the
University of Missouri-St. Louis Graduate School for the funding to conduct this
research. The authors also wish to thank Richard Rosenfeld and the anonymous
reviewers for their comments and suggestions, all of which helped to make this a better
manuscript. The views expressed herein reflect only those of the authors.
sex crimes.’
The purpose of this paper is to highlight two common perceptions
underlying sex offender laws, and the extension thereof, and to review
them in light of current empirical evidence; the first being that sex offend-
ers, more than any other group of offenders, are highly likely to repeat
their crimes, and second, that some nonsexual crime types serve as predi-
cate offenses to sexual offending. We acknowledge that sex offender laws,
and their expansion, encompass a range of beliefs and assumptions about
sex offenders beyond those we examine here. We also recognize the
desires and intentions of public officials that lay behind these policies-
primarily that of protecting the public from sexual harm and easing public
fear. To this end, this commentary is not meant to dismiss the threat of sex
offending, make light of the harm that victims endure, demean policy mak-
ers’ judgments of the sex offender problem, or declare the current policies
ineffective. Rather, based on our review, we will simply make the case
that sex offenders may not be as dangerous as the laws would dictate or as
we have been led to believe.
Our discussion begins with a review of the movement to enact the cur-
rent sex offender laws and the public’s support for these reforms. We then
examine current evidence of sex offenders’ levels of recidivism and non-
sexual offenders’ propensities to commit future sex crimes. Lastly, to fur-
ther strengthen our argument, we include criminal history data from a
single state to examine the probabilities of sexual and nonsexual offenders
committing future sex crimes.
BACKGROUND
In the late 1980s and early 1990s, three specific incidents of sexual homi-
cides against children were catalysts for much of the sex offender legisla-
tion we have today. In October 1989, Jacob Wetterling, 11, was abducted
near his home in Minnesota by an armed masked stranger (National Crim-
inal Justice Association, 1997). To date, he has still not been found. His
case resembled that of a boy in a neighboring town who was abducted and
sexually attacked earlier in the year. Both incidents are believed to have
been committed by the same man, thus leading police to conclude they
were searching for a repeat sex offender. Although the Wetterling abduc-
tion drew attention to the repetitiveness of sex offenders’ behaviors, it was
the homicides of Polly Klaas and Megan Kanka that brought this issue to
the forefront of the policy agenda (Jenkins, 1998).
In 1993, the media widely disseminated the story of Polly Klaas, a 12-
~
year-old girl who was abducted from her bedroom in California, sexually
assaulted, and subsequently killed. Only one year later, the media
reported that 7-year-old Megan Kanka was missing from her New Jersey
home; she was later found sexually assaulted and murdered. Both Polly
Klaas and Megan Kanka had been murdered by previously convicted sex
offenders who had been released from prison. The parents of these mur-
dered children actively lobbied state and federal legislators for remedies to
address the repeat behavior of sex offenders. The results of their efforts
have been witnessed nationwide.
In 1994, the Jacob Wetterling Crimes Against Children and Sexually
Violent Offender Registration Act mandated that 10% of a state’s funding
under the Edward Bryne Memorial State and Local Law Enforcement
Assistance grant program be used for establishing a state-wide system for
registering and tracking convicted sex offenders (National Criminal Justice
Association, 1997). The act also “strongly encouraged” states to collect
DNA samples from registered sex offenders to be typed and stored in
databases and used to clear crimes (p. 7). The Wetterling Act was soon
amended by the passage of “Megan’s Law” in 1996, which requires states
to make sex offender registry information available to the public.
Although states began complying with the Wetterling act and its amend-
ment by requiring blood samples and registry information from only those
sex offenders convicted of violent sex acts against children, to date, all 50
states have expanded their registry, notification, and DNA laws to include
persons convicted of a violent or nonviolent sex crime2 against any person
regardless of their age.
The tragic deaths of these children who were murdered by previously
convicted and released sex offenders not only initiated legal reform but
also undoubtedly influenced our perceptions of sex offenders. These sto-
ries helped create an image of the sex offender as a compulsive recidivist
who continues to present a danger to society despite any efforts at rehabil-
itation or reform. It is most likely this image that drives the public’s sup-
port for the current sex offender laws.
Few scientific polls have been conducted to determine the public’s opin-
ions of sex offender laws, but those that have suggest that there is over-
whelming support for sex offender registration, community notification,
and DNA database laws. In 2001, the polling firm of MORI in Great Brit-
ain conducted a survey to determine if there was public support for
“Sarah’s Law,” which is similar to our “Megan’s Law” by giving the public
other types of offenders not subject to these laws, this type of comparison
is needed.
We found few studies that directly examine sex offenders’ recidivism
rates in comparison to that found for other groups. Most of these, how-
ever, suggest that sex offenders exhibit lower rates of reoffending than
nonsexual offender groups (Langan and Levin, 2002; Hanson et al., 1995;
Sapsford, 1998; Sipe et al., 1998). For example, Sipe et al. (1998) examined
the adult arrest records of a group of adjudicated sexual and nonsexual
juvenile offenders. They found that juvenile sex offenders were signifi-
cantly more likely than nonsexual offenders to be arrested for sex crimes
as adults. However, only a small percentage of both groups had an adult
arrest for a sex crime (9.7% for sexual and 3.0% for nonsexual offenders).
These authors also found that approximately 12% of juvenile nonsexual
offenders were arrested as adults for other violent offenses and 32.6% for
property crimes. These percentages were twice those found for juvenile
sex offenders arrested as adults for other violent and property offenses
(5.6% and l6.l%, respectively). In their investigation of recidivism rates
among child molesters and nonsexual criminals, Hanson et al. (1995)
found that 83.2% of nonsexual criminals were reconvicted over 15 to 30
years as opposed to approximately 62% of child molesters. More recently,
Langan and Levin (2002) examined rearrest rates over three years for
released prisoners in 15 states (N = 272,111). They found that those con-
victed of sexual assault (41.4%) and rape (46%) were among those with
the lowest re-arrest rates when compared to other offender groups such as
burglars (74%), robbers (70.2%), and thieves (74.4%). When examining
rates of sexual re-offending, these authors found only 2.5% of released
rapists were re-arrested for another rape within three years. Some sex
offenders obviously recidivate. However, the empirical evidence to date
does not seem to suggest that sex offenders have higher rates of recidivism
than other groups of offenders and are thus in need of greater levels of
surveillance and control.
In regard to the expansion of sex offender laws to non-sexual offenders,
few researchers have directly examined the degree to which groups such as
robbers or burglars go on to become sexual criminals. Blumstein et al.
(1988) used transition matrices to examine specialization and escalation in
offender patterns among 33,000 adults 17 years or older. They observed
an increase in the seriousness of offending following arrests for aggravated
assault. However, weaker levels of escalation were noted, especially for
white offenders, following arrests for robbery, auto theft, and drugs. Also
they noted, “white offenders exhibited high levels of de-escalation in seri-
ousness following arrests for rape (342)”. No increases in the seriousness
of offending was observed following arrests for burglary or larceny. In
contrast, using quasi-symmetry modeling, Britt (1996) noted significant
66 SAMPLE & BRAY
with child victims recidivate at different rates than those with adult vic-
tims, and that those viewing or manufacturing child pornography may
reoffend at different rates than do those who commit a violent sexual
assault (Lieb et al., 1998; Marques et al., 1994; Quinsey et al., 1995, 1995,
1998). To date, however, all sex offenders in all 50 states are now required
to provide blood samples and register with law enforcement agencies,
despite their classification as pedophiles, rapists, or molesters, thus infer-
ring that all sex offenders, regardless of their type, have a higher likeli-
hood of reoffending than do their nonsexual offender counterparts. To
assess this assumption, we then must look at sex offenders as a group,
regardless of their type of offense or clinical classification.
To conduct these analyses, we first reduced the 10,688 Illinois statutes
used to enact arrest charges into 24 general offense categories resembling
those found in Part I and I1 of the UCR. These categories were created
broadly and include attempts and all levels of aggravation, in order to
examine the between-group differences in rearrests.
Once arrest charges were classified into general categories, we charac-
terized arrestees as particular types of offenders. This could have been
accomplished in several ways. Using 1990 as the base year, we considered
classifying arrestees as a specific “type” based on their first arrest charge
of the year. However, if an arrestee was charged with a burglary in Janu-
ary and a homicide in March, this categorization would label the arrestee
as a burglar and under-represent serious violent felonies in the analysis.
We also considered classifying arrestees by their most frequent charge in
1990, but again, if an arrestee’s larceny charges outnumbered sex offense
charges for the year, violent or sexual offenses would be lost. Ultimately,
we characterized arrestees as a particular “type” based on their most seri-
ous charge for 1990, so that violent and sexual felonies were not under-
represented in the analysis.
To determine the most serious charge, we created a hierarchy of offense
types that resembles the offense seriousness scale recommended by Sellin
and Wolfgang (1964). The hierarchy is based on the seriousness of the
overall crime category, not necessarily the seriousness of the individual
offense. Offense types were classified in terms of Class I offenses based
on bodily injury, property loss, or property damage, with bodily injury
being the most serious. Offense categories that did not include the ele-
ments needed for a Class I offense were then characterized as Class I1
offenses in order of seriousness based on the factors of intimidation; threat
of property loss; primary, secondary, tertiary, and mutual victimization;
and no victimization. For crimes such as stalking that did not exist in 1964,
we used our own discretion when applying Sellin’s and Wolfgang’s (1964)
characterization of offenses. Table 1 includes the 24 offense categories
70 SAMPLE & BRAY
ranked from most to least serious, the number of charges in the dataset for
each type, and their percent of total arrest charges.
RESULTS
Figure 1depicts patterns of rearrests among the listed offense categories
72 SAMPLE & BRAY
g 70.0%
f 60.0%
50.0%
5 40.0%
0
E: 30.0%
0)
2 20.0%
a
o. 10.0%
0.0%
Similar to the findings at the national level, those arrestees whose most
serious offense charge in 1990 was a robbery had the highest probability of
rearrest (74.9%) within five years followed by arrestees charged with bur-
glary (66%), nonsexual assault (%%), and larceny (52.9%). Persons in
the sex offense category had rearrest percentages of 21.3%, 37.470, and
45.1% for any offense within one, three, and five years, respectively.
These rates are lower than those found for all other crime categories in the
analysis except homicide (44.2% within five years) and property damage
(38.8%).
Sex offenders in Illinois do not appear to commit future offenses, in
general, at a higher rate than do other offenders. However, they may have
higher levels of recidivism for their crimes than other types of offenders
exhibit for their particular offenses. Figure 2 shows the percentage of 1990
arrestees rearrested for the same offense within one, three, and five years.
Again, all group differences are statistically significant at 0.05.
ARE SEX OFFENDERS DANGEROUS? 73
5 40.0%
E
;i 30.0%
c
," 20.0%
C
g 10.0%
p" 0.0%
Persons whose most serious charge in 1990 was property damage had
the highest percentage of rearrests for their crimes in five years (38.8%)
followed by those in the nonsexual assault category rearrested for assault
(37.2% within five years) and persons in the larceny category rearrested
for larceny (30%). The sex offender category had a lower offense-specific
rearrest rate in five years (6.5%) than did arrestees in most other catego-
ries. Robbers were rearrested for robbery (17.9%), burglars were rear-
rested for burglary (23.1Yo), and those arrested for public-order offenses
(21.4%) were rearrested for public-order crimes in greater proportions
than sex offenders were rearrested for sex crimes. Homicide (5.7%), kid-
napping (2.8%), and stalking ( 5 % ) were the only categories with lower
offense-specific rearrest rates within five years than sex offending, and
those differences are very small.
Only a small percentage of sex offenders in 1990 in Illinois were rear-
rested for committing another sex crime within five years. However, it is
then possible that sex offenders were not the only group of offenders com-
mitting sex crimes, and some crime type may be identified as the "gate-
way" offense to sex offending. Figure 3 depicts the percent of 1990
arrestees rearrested for a sex offense, as defined by those crimes for which
offenders must register as a sex offender in Illinois. Again, the differences
in the percentages among offense categories for one, three, and five years
are statistically significant.
74 SAMPLE & BRAY
-8
cfl
p!
60%
50%
2
c
40%
0
r
30%
C
2.0%
!
n 1.0%
0.0%
Not surprisingly, arrestees whose most serious charge in 1990 was a sex
offense have the highest percentages of rearrest for a sex crime in one,
three, and five years (2.2%, 4.8%, and 6S%, respectively). Within five
years, the robbery, kidnapping, and stalking categories all had between
2% and 3% of rearrests for sex crimes. Arrestees in the nonsexual assault,
burglary, larceny, and public order categories had between 1% and 2%
rearrests for sex crimes, and those in the homicide and property damage
classifications had less than 1% rearrests for sex offenses. Note that no
crime categories had higher than 6.5% of rearrests for sex crimes, which
suggests that the overwhelming majority of offenders in all listed crime
categories were not rearrested for a sex crime, including those persons
classified as sex offenders.
The percentages of rearrest for listed offenses for sex crimes and for the
same charges reported here are relatively low in comparison to other
recidivism studies conducted on sex offenders, which consistently find
that, on average, between 10% and 5.5% of sex offenders will be rear-
rested or reconvicted for new crimes (for reviews, see Furby et al., 1989
and Hanson and Bussiere, 1998). There are several possible explanations
for this. First, many sex offender recidivism studies examine samples of
sex offenders who have been incarcerated for their crimes. Under these
circumstances, these types of sex offenders typically represent the worst-
of-the-worst of the sex offending population. Their crimes were of such an
egregious nature, or their offending histories were so long, that probation
was not a viable option. One would expect higher rates of reoffending
ARE SEX OFFENDERS DANGEROUS? 75
among samples of incarcerated sex offenders than we have found here for
arrestees. Secondly, due to the potential mobility of the arrestees we
examined here, they may have reoffended in another state. Our analysis
would then underestimate recidivism because we have not captured these
offenses in the Illinois criminal history database. Lastly, the low rates of
reoffending found here may be a function of diminished opportunity.
Arrestees may have been incarcerated and thus are no longer able to com-
mit a new offense on the street. To investigate this possibility, we look to
IDOC data.
Corrections data indicate that except for persons arrested for robbery
(58.8%), the majority of all other types of arrestees spent no time in a
state correctional facility from 1990 to 1997. Only 30.8% of persons
arrested for a sex crime in 1990 were incarcerated in a state prison for
some time during the period of the analysis, leaving almost 70% of offend-
ers on the street with the opportunity to reoffend. In addition, approxi-
mately 85% of those arrested in 1990 for nonsexual assault, 87% of
persons arrested for kidnapping, and 82% of persons arrested for larceny
had no state prison sentence. Surprisingly, compared with sex offenders, a
greater proportion of people arrested for burglary were incarcerated in
state correctional facilities during the period. However, this may reflect
the greater likelihood of persons arrested for sex offenses to “plead down”
their charges.
Even if most sex offenders were not incarcerated over the seven-year
period, those who were may have spent more time in prison than persons
sentenced for other crimes. Illinois Department of Corrections data indi-
cate that incarcerated sex offenders did spend more time, on average, in
correctional facilities than other types of arrestees (27.7 months). How-
ever, this mean time served is less than three years and remains well within
the window of the analysis. Those in the homicide and robbery categories,
on average, spent two months fewer in custody than sex offenders (25.7
and 25.4, respectively), but recall that these are broad categorizations of
offense types that include attempts and all levels of aggravation. With
respect to homicide, the inclusion of attempts may be responsible for the
appearance of shorter mean time served than found for the sex offense
category. On average, persons in the other categories spent between one
and two years in prison. The average time spent in custody for all those
arrested and incarcerated for these offenses indicates that most had ample
time to reoffend during the analysis period. Even so, it is possible that the
difference in rearrest rates between sex offenders and others is due to the
somewhat longer time spent in prison by the 31% of them who were incar-
cerated. It is unlikely, however, that all of the differences in rearrest per-
centages between sexual and nonsexual offenders can be attributed to
time served.
76 SAMPLE & BRAY
these agents and possibly affect their ability to perform any of their duties
effectively.
With every classification of offender that policy makers add to DNA
collection or sex offender registries, they substantially increase the number
of people law enforcement, probation, and parole agencies must monitor.
For example, under a 1986 Illinois statute 3,609 sex offenders with child
victims were potentially subject to DNA collection and registration, based
on the 1990 arrest data. After the 1995 extension of this statute to include
all sex offenders with child or adult victims, the number grew to 5,483, an
increase of 52%. How effectively can criminal justice agents respond to
this increase in workload? How effectively can they monitor a rapidly
growing population of offenders? It seems reasonable to assume that the
effectiveness and efficiency of sex offender laws may be diminished with
their expansion to more people and behaviors. Agencies will be forced to
spread their time and scant resources across an ever-increasing number of
offenders and behaviors. As criminal justice agents attempt to manage
this larger population of offenders, they would likely be less effective at
monitoring the behavior and tracking the whereabouts of sex offenders,
which is the group that policy makers originally hoped to control.
Criminal justice agents will not be the only group to be affected by the
expansion of sex offender laws to nonsexual offenders. Current sex
offender policies require a substantial financial commitment from the tax-
payer. In 1999, the Illinois State Police received approximately $500,000
from the legislature to maintain the sex offender registry and place this
information on the World Wide Web. Although this may sound like a
large sum, it does not fully cover the costs of the computer mainframe,
software, servers and printers, personnel salaries, training, and research
needed to support registration and community notification laws. In addi-
tion, the Illinois Department of Corrections requested $115,000 to supple-
ment the costs of the photography and computer equipment, development
fees, and personnel salaries needed to take the photographs of sex offend-
ers that are posted on the registration website. The maintenance of the
sex-offender DNA databank also requires a sizeable financial commit-
ment. In 1999, $950,000 of public funds were allocated to the ISP Forensic
Division to maintain the DNA databank, take blood samples from both
sexual and nonsexual offenders, and perform the scientific tests necessary
to match evidence left at a crime scene to samples they possess. Despite
the sum received to carry out these duties, the Forensic Division continues
to run six months behind in matching crime scene evidence to database
informations. The addition of nonsexual offenders to registration and
REFERENCES
Bachman, Ronet
1998 The factors related to rape reporting behavior and arrest: New evidence
from the National Crime Victimization Survey. Criminal Justice and
Behavior 25:8-29.
Becker, Judith V. and John A. Hunter Jr.
1992 Evaluation of treatment outcomes for adult perpetrators of child sexual
abuse. Criminal Justice and Behavior 19:74-92.
Blumstein, Alfred, Jacqueline Cohen, Somnath Das, and Soumyo D. Moitra
1988 Specialization and seriousness during adult criminal careers. Journal of
Quantitative Criminology. 4:303-345.
Britt, Chester L.
1996 The Measurement of Specialization and Escalation in the Criminal
Career: An alternative modeling strategy. Journal of Quantitative
Criminology. 12:193-222.
Bureau of Justice Statistics
1996 Sourcebook. Washington D.C.: US. Department of Justice.
Cattabiani, Mario F.
1999 Dent bill to expand DNA database. Allentown Morning Call, (March
26):Bl.
Dundes, Lauren
2001 Is the American public ready to embrace DNA as a crime fighting tool?
A survey assessing support for DNA databases. Bulletin of Science,
Technology & Society 21:369-375.
Furby, Lita, Mark R. Weinrott, and Lyn Blackshaw
1989 Sexual offender recidivism: A review. Psychological Bulletin 105:3-30.
Hanson, R. Karl and Monique T. Bussiere
1998 Predicting relapse: A meta-analysis of sexual offender recidivism studies.
Journal of Consulting and Clinical Psychology. 60348-362.
Hanson, R. Karl, Heather Scott, and Richard A. Steffy
1995 A comparison of child molesters and nonsexual criminals: Risk predictors
and long-term recidivism. Journal of Research in Crime and Delinquency
32~325-337.
Jenkins, Philip
1998 Moral Panic, Changing Concepts of the Child Molester in Modern
America. London: Yale University Press.
Koss, M.
1996 The measurement of rape victimization in crime surveys. Criminal Justice
and Behavior 23:55-69.
Langan, Patrick A. and David J. Levin
2002 Recidivism of Prisoners Released in 1994. Washington D.C.: U.S.
Department of Justice. Bureau of Justice Statistics.
Lieb, Roxanne, Vernon Quinsey, and Lucy Berliner
1998 Sexual predators and social policy. Crime and Justice: A Review of
Research 23:43-114.
Marques, Janice K., David M. Day, Craig Nelson, and Mary Ann West
1994 Effects of cognitive-behavioral treatment on sex offender recidivism
preliminary results of a longitudinal study. Criminal Justice and Behavior
21:28-54.
National Crime Victimization Survey
1995 Criminal Victimization in the United States. Washington D.C.: U S .
Department of Justice. Bureau of Justice Statistics.
National Criminal Justice Association
1997 Sex Offender Community Notification. Washington D.C.: National Crimi-
nal Justice Association.
ARE SEX OFFENDERS DANGEROUS?
Phillips, Dretha M.
1998 Community Notification as Viewed by Washington’s Citizens. Washing-
ton: Washington State Institute for Public Policy.
Prentky, Robert A., Raymond A. Knight, and Austin F. S. Lee
1997 Risk factors associated with recidivism Among extrafamilial child molest-
ers. Journal of Consulting and Clinical Psychology 65:141-149.
Prentky, Robert A., Raymond A. Knight, Austin F. S . Lee, and David D. Cerce
1995 Predictive Validity of Lifestyle Impulsivity For Rapists. Criminal Justice
and Behavior 22:106-128.
Quinsey, Vernon L.
1984 Sexual aggression: Studies of offenders against women. In David N.
Weisstub (ed.), Law and Mental Health: International Perspectives, Vol.
1. New York: Pergamon.
Quinsey, Vernon L., Arunima Khanna, and P. Bruce Malcolm
1998 A retrospective evaluation of the regional centre sex offender treatment
program. Journal of Interpersonal Violence, 13:621--644
Quinsey, Vernon L., Marnie E. Rice, and Grant T. Harris
1995 Actuarial prediction of sexual recidivism. Journal of Interpersonal
Violence 10:85-105.
Rice, Marnie E., Vernon L Quinsey, and Grant T. Harris
1991 Sexual recidivism among child molesters released from a maximum
security psychiatric institution. Journal of Consulting and Clinical Psy-
chology 59~381-386.
Sample, Lisa L.
2001 The social construction of a sexual offender. Ph.D. dissertation. Unpub-
lished.
Sapsford, Roger J.
1998 Further research applications of the “Parole Prediction Index.” Interna-
tional Journal of Criminology and Penology. 6247-254.
Sellin, Thorsten and Marvin E. Wolfgang
1964 The Measurement of Delinquency. New York: Wiley.
Sipe, Ron, Eric L. Jensen, and Ronald S. Everett
1998 Adolescent sexual offenders grow up recidivism in young adulthood.
Criminal Justice and Behavior 25:109-124.
Spaulding, Tom
1999 State wants DNA samples from convicted burglars. Sarasota Herald
Tribune (Nov. 16):lB.
Stevens, Aaron P.
2001 Arresting crime: Expanding the scope of DNA databases in America.
Texas Law Review. 79:921-961.
Uniform Crime Report
1996 Uniform Crime Report. Washington D.C.: Federal Bureau of Investiga-
tion. Available online http://www.fbi.gov/ucr/ucreports
Willing, Richard
2000 Many rapists were thieves first: Results may lead to taking of DNA for
lesser crimes. USA Today (July 10):3A.
SAMPLE & BRAY