Examining Specialization Among Sex Offenders Released From Prison

Abstract

A prevailing cultural stereotype about sex offenders is that they tend to specialize in sexual offending. Many recent policy developments—mainly aimed to restrict the liberties of sex offenders—are rooted in this idea. We examined the correctional and arrest records of a sample of 312 sex offenders released on parole in Colorado to determine the prevalence of sexual specialization among these offenders, and to compare the legal and social characteristics of specialists and versatile sex offenders. Overall we found that very few participants officially classified as sex offenders fit the specialist stereotype. Study participants generally displayed versatile histories of criminal offending. We also found that specialists were distinguishable from versatile offenders on certain indices of social integration and mental health, and they were more likely to have had a history of offending against children.

Keywords

sexual offender sexual recidivism criminal recidivism specialization

A pervasive cultural stereotype about sex offenders is that they “specialize” in sexual predation, especially toward children (Farkas & Stichman, 2002; Garland, 2001; Harris, Knight, Smallbone, & Dennison, 2011; Jenkins, 1998; Lieb, Quinsey, & Berliner, 1998; Quinn, Forsyth, & Mullen-Quinn, 2004; Sample & Bray, 2003; Vasquez, Mannan, & Walker, 2008; Zimring, 2004). This logic has provided the rationale for the adoption of increasingly restrictive policies designed to monitor, deter, and punish sexual crime in the United States (Lieb et al., 1998; Terry & Ackerman, 2009). These policies include registration, residency restrictions, community notification, polygraph testing—and in some cases, Global Positioning System (GPS) monitoring, indefinite confinement to psychiatric institutions, and chemical castration (Terry & Ackerman, 2009).

Some sex offenders may indeed fit the stereotype of a sexual offense specialist but others do not. A number of studies have attempted to determine the degree to which sex offenders focus on sexual offending. These studies consistently find that although certain sex offenders display some amount of specialization in sex crimes, specialization is overestimated and criminal versatility is common (Harris et al., 2011; Magers, Jennings, Tewksbury, & Miller, 2009; Miethe, Olson, & Mitchell, 2006; Simon, 1997, 2000; Weinrott & Saylor, 1991). Those who commit sex offenses also commonly commit property, violent, drug, and other types of crimes. Lussier (2005) posits that versatility and specialization may not be entirely contradictory dynamics: “The fact that most sexual offenders do not restrict themselves to sexual crime does not preclude some form of specificity in their offending behavior” (p. 283).

Given the expansive scope of sexual offending laws, however, the legal status of “sex offender” has been applied to a variety of offender types with varying degrees of sexual specialization. A number of researchers have found child molesters (particularly extrafamilial child molesters) to be more specialized than other types of sex offenders (Harris et al., 2011; Lussier, 2005; Lussier, LeBlanc, & Proulx, 2005; Miethe et al., 2006). Relatedly, Parent, Guay, and Knight (2011) found that clinical risk assessment instruments yielded marginal to modest predictive accuracy for sexual recidivism overall, but they were better at predicting sexual reoffending among child molesters than rapists. Lussier (2005) speculated that sex offenders who offended against children might be more specialized than those who offended against adult women because offenders tend to specialize as they get older (Paternoster, Brame, Piquero, Mazerolle, & Dean, 1998), and child molesters tend to persist longer. Soothill, Francis, Sanderson, and Ackerley’s (2000) longitudinal study of sex offenders in England and Wales found that those convicted of indecent assault on males were more specialized than other sex offender subgroups.

Although the sexual specialization stereotype has been called into question, few studies have addressed characteristic differences between specialist and versatile sex offenders within the same sample. Harris, Mazerolle, and Knight (2009) compared specialist and versatile male sex offenders who had been referred for civil commitment in Massachusetts between 1959 and 1984, finding that specialists had fewer problems in school and were less likely to have presented adolescent antisocial behaviors (see also Harris et al., 2011). Specialists were less likely to exhibit traits associated with psychopathy but more likely to show sexual preoccupation and emotional congruence with children. This finding fits with those studies mentioned above that find greater specialization among child molesters than other sex offender types. Other studies have also found that sexually relevant variables such as deviant sexual preferences and prior sexual offending are predictive of sexual recidivism (Hanson, 2002; Hanson & Bussiere, 1998; Hanson & Morton-Bourgon, 2005). Very little is known, however, about differences in social characteristics between specialists and versatile sex offenders in the modern “get tough” era of sex offending policy. (Harris et al.’s sample is about three decades old.)

Studies of sexual recidivism suggest that factors associated with low self- and social control such as lifestyle instability, antisocial personality, and relatedly, problems forming positive bonds with friends and romantic partners (“intimacy deficits”) are empirically associated with sexual reoffending (Hanson, 2002; Hanson & Morton-Bourgon, 2005; Lussier et al., 2005). Broader criminological scholarship has shown that social supports related to marriage, parenthood, familial closeness, educational attainment, and vocational engagement correlate negatively with the propensity to commit crime in general (e.g., Gottfredson & Hirschi, 1990; Hirschi, 1969). Sex offender-specific policies are designed to monitor a group of offenders who presumably lack such supports and are thus at an elevated likelihood of repeat sexual offending; items related to social support are commonly found in actuarial risk assessment instruments used to predict sexual reoffending risk. Thus, we find it reasonable to ask whether sex offenders with higher degrees of offense specialization exhibit weaker supports than those who specialize less. Harris et al. (2009) tested some of these characteristic differences, finding no significant differences between specialists and versatile offenders on measures of relationship difficulties, marital histories, and employment problems.

In this article, we examine data drawn from correctional and arrest reports in Colorado to explore the degree that sex offense specialists can be identified among those officially categorized as sex offenders, and to describe the distinct social characteristics of this subgroup. In other words, what proportion of those statutorily classified as sex offenders are specialists? Do specialists exhibit weaker social supports than versatile sex offenders?

Method

Research Setting

In this study, we examined data about a representative sample of sex offenders released from prison to the Denver, Colorado area, in 2009 and 2010. The participants of this study were on parole at the time of data collection, making them distinct from “less serious” populations of sex offenders, such as those on probation or those who had only served jail time without an attached period of parole. Virtually all participants were subject to registration requirements and residency restrictions. All sex offenders in the community in Colorado are subject to specialized conditions of supervision, as determined by a “community supervision team” that works in partnership with other criminal justice personnel (Colorado Sex Offender Management Board, 2000). These conditions include prohibitions on child and/or victim contact, prohibitions on changing residences without prior approval, requirements that sex offenders disclose their legal statuses to those with whom they live, and registration requirements. Sex offenders are also responsible for paying for the evaluations and treatments they receive.

Participants

Our dataset was constructed by drawing samples of sex offenders on parole in Colorado on two separate dates in 2009 and 2010. Some offenders were active on both dates. When this was the case, we coded available files from the first date only, ignoring any informational updates received by the second date. Thus, the second round of data collection added only those offenders who had been paroled between the first and second dates. We drew two samples to increase the number of cases available for analysis. The final sample contained 312 participants. Participants’ average age was 37.6 years and 97% were male. Forty-three percent of participants were White, 22% were Latino, 18% were Black, and 17% were of other or unknown race/ethnicity. Over two thirds (67%) had a high school diploma or higher education. Thirty-nine percent of participants were married or in a long-term romantic partnership, and 58% had children.

Measures

We coded participants’ legal and social information from archived reports provided by the Colorado Department of Corrections (DOC). Two reports served as principal sources of information:

Admission Data Summary (ADS): This document reports participants’ demographics, criminal histories, and other personal characteristics.

Diagnostic Needs Summary (DNS): This document reports, in detail, participants’ social characteristics, mental health issues, and patterns of drug and alcohol use.

These reports provided rich information about participants’ legal and social characteristics, described below.¹ Data coding was performed by a team of undergraduate and graduate students and supervised by the authors. Student coders were given extensive training by the authors and final coding guidelines were developed after a series of training sessions—which included practice coding—about the interpretation of data in the aforementioned DOC reports. During the actual coding process, the first author recoded a selection of cases (i.e., every 10th case) to check for agreement in coding decisions. An inter-rater reliability analysis was performed using the kappa statistic (for nominal and ordinal measures) and Pearson’s r (for continuous measures). Upon completion of coding, two undergraduate students entered the data into a database. Both data enterers checked each record; one would enter the data and the other would recheck the entire electronic record for accuracy. Inconsistencies throughout the coding and data entry process were resolved in dialogue with the lead author.²

DOC reports provided rich information about participants’ legal and social characteristics. Below, we summarize the measures we used for the available data to construct:

Demographics: age at release from prison; gender; race/ethnicity (White, Black, Hispanic, Asian/Other/no answer).

Criminal history: number of prior felony arrests, prior misdemeanor arrests, prior probation terms, prior probation revocations, prior parole terms, prior parole absconds, prior parole revocations.

Educational attainment/achievement and IQ: highest degree achieved (less than high school, high school diploma/General Education Diploma (GED), any college or more), Tests of Adult Basic Education (TABE) scores for reading, language, math, and overall, IQ (below average, average, above average—as recorded by DOC staff).

Employment: months worked at last job, whether last job was skilled/white collar, whether the subject had ever worked in a skilled/white collar job.

Family history: marital status, whether subject has children, “family network score” indicating the number of immediate family member types (spouse, children, mother, father, siblings) living in state.³

Psychopathology: drug use per week (days), age at first drug use, alcohol use per week (days), age at first alcohol use, reported psychiatric hospitalization (ever), reported mental health treatment or counseling (ever), reported psychotropic medication (ever).

Inter-rater reliability was high for all collected measures. The inter-rater agreement averaged 97.4% across all measures.⁴ Kappa statistics ranged from 0.59 to 1.00. Correlation coefficients (Pearson’s r) ranged from 0.85 to 1.00. All kappa statistics and correlation coefficients were statistically significant (p ≤ .001).

Participants’ complete criminal records, including their histories of sexual offending, were constructed from arrest data provided by the Colorado Division of Criminal Justice and the Colorado Bureau of Investigation. These arrest data included all charges for all arrest incidents in the state of Colorado including traffic-related arrests. Thus, we were able to determine whether any given arrest was for a sex crime, even if the sex crime charge “sat underneath” a more serious charge. We did not count arrests for failure to register as a sex offender, but all other sexual crimes were included, including non-contact offenses such as possession of child pornography and indecent exposure. We were also able to identify incidents in which offenders were arrested for sexual crimes against children. These data therefore provided a comprehensive, accurate record of each participant’s official offending history in Colorado, including sexual offending history (as measured by arrests). We note here that official offending histories are not necessarily accurate with regard to actual offending histories, as many crimes are undetected. This is a limitation of our study, just as it is a limitation of all such studies.

Procedures

To indicate sex offense specialization, we applied a standard specialization threshold used in other studies: Participants with more sex offense arrests than non-sex offense arrests (i.e., arrests in which there was no indication of any sexual malfeasance in any charges) were considered specialists (Harris et al., 2011; Harris et al., 2009; Miethe et al., 2006). We did not consider charging and conviction in this definition. Twenty-two out of the original 312 participants in the data (7.1%) were first-time offenders; they had only one sex arrest on record and no others of any type. Because measuring offense specialization in cases involving only one arrest would be inappropriate, these first-time offenders were dropped from our bivariate comparisons. Thus, our final sample for analysis contained 290 participants.

Using arrest history to measure specialization was not ideal. Interpersonal and institutional dynamics may have skewed the nature of data. Because many arrests did not lead to a conviction, a limitation of these data may be the overestimation of criminal activities because of unfounded or unproven arrests. Zilney and Zilney (2009) further point out that sex offenders who victimized strangers were more likely to come to the attention of authorities than those who victimized people they knew. Racial and class dynamics in arrest decisions may also have affected the characteristics of the study sample. We are sensitive to these issues, but the fact remains that study participants represent the population of sex offenders being released from prison to the community in Colorado regardless of any institutional dynamics that may influence their chances of having been arrested and prosecuted.

We compared specialists against versatile offenders across legal and social characteristics to determine whether specialists were distinguishable from other sex offenders. To do this, we conducted bivariate comparisons and applied standard tests of statistical significance with a significance threshold (alpha) of p ≤ .05. To compare nominal and ordinal variables across specialists and versatile offenders, we used a chi-square test. To compare continuous variables across specialists and versatile offenders, we used a t test.

Results

Most sex offenders in the present sample did not fit the specialist stereotype. When the 50% specialization threshold was applied (excluding those with only one arrest on record), 59 out of 290 participants (20.3%) met the criteria for specialization (see Table 1). Thus, most sex offenders released on parole in Colorado were more aptly described as versatile offenders who had committed sex offenses, rather than specialists predominantly focused on sexual offending. Only 13 (4.5%) participants had been arrested exclusively for sex offenses (and had more than one arrest on record).

Table 1.

Descriptive Statistics on Specialization.

	n	%
Full sample	312	100.0
First-time offenders (one sex offense and no others)—EXCLUDED	22	7.1
Sample for analysis	290	100.0
Specialists (more than 50% of arrests were for sex offenses)	59	20.3
Exclusively arrested for sex offenses (more than one arrest)	13	4.5

To protect against family-wise error, we first grouped all measured variables except age, race, and child-victim ever into the five theoretically defined domains presented in Table 2: criminal history, educational attainment/achievement and IQ, employment, family history, and psychopathology. Each domain was analyzed separately, using multivariate analysis of variance (MANOVA) with the variables in each domain as the independent variables and offender type (versatile vs. specialist) as the dependent variable. Follow-up analyses (t tests for continuous measures and χ² for bivariate nominal scales) were conducted only for domains that yielded significant overall MANOVAs (p < .05; Wilks’s λ). Age, race, and child-victim ever were analyzed separately. Table 2 presents the overall results for the five MANOVAs. Only the employment domain failed to reach the overall significance criterion.

Table 2.

MANOVA Clusters and Significance.

Cluster	Included variables	Wilks’s λ	F	p
Criminal history	Number of prior felony arrests; prior misdemeanor arrests; prior probation terms; prior probation revocations; prior parole terms; prior parole absconds; prior parole revocations	0.781	9.014	.000
Educational attainment/achievement and IQ	Highest degree achieved; TABE scores; IQ	0.940	2.776	.012
Employment	Months worked at last job; whether last job was skilled/white collar; whether subject had ever worked in a skilled/white collar job	0.998	0.213	.887
Family history	Marital status; whether subject has children; “family network score”	0.959	3.323	.011
Psychopathology	Drug use per week; age of first drug use; alcohol use per week; age of first alcohol use; reported psychiatric hospitalization (ever); reported mental health treatment or counseling (ever); reported psychotropic medication (ever)	0.920	2.540	.016

Note. TABE = Tests of Adult Basic Education.

Tables 3 and 4 present the analyses for age, race, and child-victim ever and the post hoc analyses for the variables in the four significant domains identified through our MANOVAs. Table 3 reports comparisons of nominal and ordinal variables between specialists and versatile offenders. Table 4 reports mean comparisons of continuous variables. When data were missing, cases were excluded in these specific analyses.

Table 3.

Nominal and Ordinal Variable Comparisons, Versatile Offenders Versus Specialists.

Domain	Measure	Percentage of versatile offenders	Percentage of specialists	p	χ²	Cramer’s v	n
Demographics	White	36.8	59.3	.006	12.298	0.206	290
	Black	21.6	6.8	.006
	Hispanic	24.2	22.0	.006
	Asian/other/no answer	17.3	11.9	.006
Criminal history	Ever arrested for a sex offense against a child (including non-contact)	35.5	66.1	.000	18.103	0.250	290
Education/IQ	Less than high school	32.3	32.6	.362	2.031	0.085	282
	High school diploma/GED	57.0	49.2
	Any college or more	10.8	16.9
Education/IQ	Below average IQ	9.0	13.0	.368	2.000	0.085	275
	Average IQ	72.9	75.9
	Above average IQ	18.1	11.1
Family history	Married/long-term partnership	39.8	33.9	.404	0.697	0.049	290
	Single/divorced/other	60.2	66.1
Family history	Have children	62.8	44.1	.009	6.794	0.153	290
Family history	No family members living in Colorado	12.1	23.7	.024	3.804	0.115	290
Psychopathology	Report psychiatric hospitalization ever	8.7	23.7	.001	10.314	0.189	290
Psychopathology	Report mental health treatment/counseling ever	51.9	37.3	.044	4.042	0.118	290
Psychopathology	Report taking psychotropic medication ever	26.8	40.7	.038	4.314	0.122	290

Table 4.

Comparisons of Means, Versatile Offenders Versus Specialists.

Domain	Measure	Mean among versatile offenders (SD)	Mean among specialists (SD)	p	N
Demographics	Age in years	38.0 (11.3)	36.0 (13.6)	.314	254
Criminal history	Felony arrests	5.9 (3.7)	3.1 (2.3)	.000	251
Criminal history	Misdemeanor arrests	7.9 (6.3)	1.9 (1.8)	.000	251
Criminal history	Prior probations	1.2 (1.2)	0.8 (0.7)	.006	251
Criminal history	Prior probation revocations	0.9 (1.1)	0.6 (0.7)	.044	251
Criminal history	Prior paroles	0.9 (1.1)	0.3 (0.5)	.000	251
Criminal history	Prior parole absconds	0.3 (0.7)	0.1 (0.3)	.021	250
Criminal history	Prior parole revocations	0.7 (0.9)	0.2 (0.5)	.001	248
Education/IQ	Reading TABE (mean grade level)	8.5 (3.4)	9.4 (3.3)	.072	270
Education/IQ	Language TABE (mean grade level)	7.2 (4.2)	8.1 (3.8)	.156	269
Education/IQ	Math TABE (mean grade level)	7.2 (3.0)	8.1 (3.3)	.053	269
Education/IQ	Total TABE (mean grade level)	7.6 (3.6)	8.5 (3.7)	.132	262
Family history	Family network score: +1 for every family member type in state	2.2 (1.4)	2.1 (1.6)	.566	290
Psychopathology	Drug usage per week (mean days)	2.8 (3.0)	2.0 (2.6)	.061	290
Psychopathology	Age at first drug use	15.1 (3.6)	15.2 (3.4)	.828	220
Psychopathology	Alcohol use per week (mean days)	1.9 (2.3)	1.8 (2.1)	.757	290
Psychopathology	Age at first alcohol use	15.3 (4.2)	15.9 (3.8)	.356	221

Note. TABE = Tests of Adult Basic Education.

As can be seen in Table 3, specialists and versatile offenders were equivalent in age, but they differed in race. Specialists were significantly more likely to be White (59.3% vs. 36.8%) and significantly less likely to be Black (6.8% vs. 21.6%). Specialists were significantly more likely than versatile offenders to have been arrested for a sexual offense against a child (66.1% vs. 35.5%).

Specialists had shorter criminal records than versatile offenders (see Table 4). They had significantly fewer felony and misdemeanor arrests on record (3.1 vs. 5.9, and 1.9 vs. 7.9, respectively). Specialists also had fewer prior probation terms, probation revocations, parole terms, parole absconds, and parole revocations.

Neither specialists nor versatile offenders could be described as well educated. They were similar in terms of their highest degrees achieved, with very few members of either group having attended college. Only 16.9% of specialists and 10.8% of versatile offenders reported educational attainment of “any college or more” (difference not statistically significant). Low educational achievement was also reflected in participants’ TABE scores—which indicate the grade levels of participants’ reading, language, and math skills. Overall, both groups’ academic functioning rated around middle school levels. Specialists and versatile offenders were no different on IQ either. The vast majority of each group scored “average” on IQ (determined by tests administered by correctional staff).

Specialists and versatile offenders differed across certain measures of social integration (or social support). Specialists were far less likely to have children; only 44.1% of specialists reported having children, compared with 62.8% of versatile offenders. Specialists were also more likely to report having no immediate family members (i.e., spouse, child, parent, sibling) living in Colorado; 23.7% of specialists did not have any immediate family members in state, compared with only 12.1% of versatile offenders. We identified no difference, however, across our “family network score”—a measure that captures the number of immediate family member types living in state.

We identified few differences across the two groups on drug and alcohol use patterns. Drug usage per week, age at first drug use, weekly alcohol consumption, and age at first alcohol use were all statistically similar. Specialists and versatile offenders differed in their histories of psychiatric treatment. Specialists were less likely than versatile offenders to have received psychiatric or mental health counseling during their lives (37.3% vs. 51.9%). However, specialists were more likely than versatile offenders to have taken psychotropic medication in their lives (40.7% vs. 26.8%), and specialists were far more likely to report psychiatric hospitalization at some point in their lives (23.7% vs. 8.7%). Specialists’ elevated rates of psychotropic medication and psychiatric hospitalization may actually be explained by their repeated sexual offending, which frequently results in such treatment approaches because they are mandated by the criminal justice system. Our data do not address this issue directly, so we cannot know for sure. Future research in this area should explore these issues more explicitly.

Discussion

The results of our analyses highlight disconnection between the presumed characteristics of offenders supervised under sexual offending policies and their actual characteristics. The prevailing cultural stereotype—that all sex offenders are specialists—found no support in a representative sample of Colorado sex offenders released on parole. This is not to say that study participants have not committed serious offenses, or that they do not present any danger to public safety. Rather, findings simply illustrated a gap between the stereotypical image of sex offenders that these policies promote—and are justified by—and the real characteristics of the sex offenders that the policies are applied to.

Most fundamentally, sex offense specialists were difficult to locate in our study sample. After excluding first-time offenders, only 20.3% of the sample could be described as specialists, meaning more than half of their arrests involved sex offenses and they had more than one arrest on record. Only 4.2% of the sample had been arrested exclusively for sex offenses. The criminal versatility of the sample suggested that very few participants could accurately be described as sexual specialists. Regardless of their psychological, cognitive, or social orientations toward sexual offending, sex offenders released on parole in Colorado generally appeared to be offenders first, and sex offenders second. Most specialists engaged in other types of crime.

The data suggest that specialists have reduced levels of social support. Our findings around social support measures show that specialists were distinguishable from versatile offenders on certain indices of social integration. They were less likely to have children. They were more likely to have no immediate family members living in state. (They were also less likely to report being married, although this difference did not reach statistical significance.) Specialists, therefore, appeared to exhibit fewer social supports that may inhibit offending. Observed differences between specialists and versatile offenders in our sample provided conditional support for this idea. That is, specialization appeared more likely among those with weaker social supports. Policies that inhibit these controls, such as residency and employment restrictions, and prohibitions on family contact, may have the unintended effect of aggravating issues that promote sexual offending (Kruttschnitt, Uggen, & Shelton, 2000; Levenson, D’Amora, & Hern, 2007; Zevitz, Crim, & Farkas, 2000). The Colorado Department of Public Safety (2004) has itself found that sex offenders with more positive social supports—such as family, friends, and employers—commit significantly fewer probation violations and new offenses than others.

However, this finding may also reflect a “chicken and egg” dilemma involving sexual interest in children. A critical area of difference between specialists and versatile offenders in our sample had to do with their histories of sexual offending against children. Almost two-thirds of specialists in our sample had been arrested for a sexual offense against a child, but only 36% of versatile offenders had been arrested for such a crime. Do more specialized sex offenders exhibit fewer social supports because such supports inhibit their criminal interests, or do their criminal interests result in the diminishment of these supports? Our findings show that diminished support and against-child offending are both more prevalent among specialists, but our analyses cannot establish a causal or directional relationship. We can only observe that these characteristics co-vary. Our findings are, however, in line with other studies of specialization, that have found that sex offenders who victimize children are more likely to specialize than those who do not (see, for example, Harris et al., 2011; Lussier, 2005; Lussier et al., 2005; Miethe et al., 2006). Lussier (2005) has posited that child molesters persist longer than other sex offender types, which coincides with specialization because offenders generally specialize more as they get older. Indeed, in our sample, study participants who had been arrested for victimizing children were significantly older than those who had not (39.1 vs. 36.5; not shown in tables).

Sex offender policies are commonly justified by the logic of risk management. If officials and the public know where these offenders are, what they are doing, and what they are thinking, future sexual offending can be deterred. This may be true to an extent; we cannot entirely know how much sexual offending would take place in the absence of these policies. However, the balance of evaluation research shows few benefits, and the implementation of these policies is very expensive (Adkins, Huff, & Stageberg, 2000; Petrosino & Petrosino, 1999; Schram & Darling Milloy, 1995; Welchans, 2005). Colorado has struggled with the decision to adopt federal sex offender standards because they are so difficult and expensive to implement—even though the failure to adopt those standards would cost the state half a million dollars in federal grant money (Fender, 2011). Most offenders caught in their purview are more aptly described as versatile offenders who have committed sex offenses, rather than sex offense specialists. Thus, registration, community notification, and other high-intensity monitoring policies are actually managing a group of offenders with a diversity of social and clinical issues that transcend the specifics of sexual offending. Sex offense–specific policies aim to monitor and control a group of offenders who are believed to need such close supervision because of their propensity toward repeated sexual offending, but our examination of this sample of sex offenders belies this stereotype. Some sex offenders may fit this profile, but many do not.

Despite the concerns raised by this study, our findings point to areas of potential improvement in the application of these policies. We call for increased granularity in sex offending policies, moving away from “one size fits all” approaches and toward strategies that acknowledge the complexity and diversity of sexual crime. If persistent sex offense specialists can be socially or clinically distinguished from within the “widened net” of sex offender labeling, appropriate interventions can be applied, and those who do not need such interventions could be transferred to more traditional supervision programs. We find specialists to exhibit fewer of social supports than versatile offenders. If diminished social support is encouraging sexual offending among specialists, policymakers and other officials can tailor policy and practice to strengthen these controls or focus monitoring efforts on those who appear to have the lowest levels of social support. We also find specialists to be more likely to have victimized children. Ironically, modern U.S. sex offending policies have politically and rhetorically revolved around the protection of children, but as our study and others show, many different types of offenders now fall under their purview, spreading state resources thin. To be clear, we are not recommending that sex offender–monitoring policies be limited to child molesters, only that states examine the diversity of offending included in these policies.

Clinicians have already developed actuarial risk assessment instruments for sex offenders, and these are widely used to make classification and programmatic decisions, but the broad label of “sex offender” is an offense-based classification (Levenson et al., 2007). Sex offenders are so labeled because of the crimes they have committed, rather than their specific risk profiles and clinical treatment needs. As we suggest in our analyses, many of those officially classified as sex offenders exhibit fundamentally dissimilar criminogenic needs than those who present high risk of repeat sexual offending—including those who persistently target children.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Adkins

Huff

Stageberg

(2000). The Iowa sex offender registry and recidivism. Des Moines: Iowa Department of Human Rights.

Colorado Department of Public Safety. (2004). Report on safety issues raised by living arrangements for and location of sex offenders in the community. Denver: Colorado Department of Public Safety.

Colorado Sex Offender Management Board. (2000). Standards and guidelines for the assessment, evaluation, treatment and behavioral monitoring of adult sex offenders. Denver: Colorado Department of Public Safety, Division of Criminal Justice, Office and Research and Statistics.

Farkas

M. A.

Stichman

(2002). Sex offender laws: Can treatment, punishment, incapacitation, and public safety be reconciled? Criminal Justice Review, 27, 256-283.

Fender

(2011, February 27). Colorado may skip offer of federal grant and stick with own sex-offender standards. Denver Post. Retrieved from http://www.denverpost.com/ci_17493514

Garland

(2001). The culture of control: Crime and social order in contemporary society. Chicago, IL: University of Chicago Press.

Gottfredson

M. R.

Hirschi

(1990). A general theory of crime. Stanford, CA: Stanford University Press.

Hanson

R. K.

(2002). Recidivism and age: Follow-up data from 4,673 sexual offenders. Journal of Interpersonal Violence, 17, 1046-1063.

Hanson

R. K.

Bussiere

(1998). Predicting relapse: A meta-analysis of sexual offender recidivism studies. Journal of Consulting and Clinical Psychology, 66, 348-362.

10.

Hanson

R. K.

Morton-Bourgon

K. E.

(2005). The characteristics of persistent sex offenders: A meta-analysis of recidivism studies. Journal of Consulting and Clinical Psychology, 73, 1154-1163.

11.

Harris

D. A.

Knight

R. A.

Smallbone

Dennison

(2011). Postrelease specialization and versatility in sexual offenders referred for civil commitment. Sexual Abuse: A Journal of Research and Treatment, 23, 243-259.

12.

Harris

D. A.

Mazerolle

Knight

R. A.

(2009). Understanding male sexual offending: A comparison of general and specialist theories. Criminal Justice and Behavior, 36, 1051-1069.

13.

Hirschi

(1969). Causes of delinquency. Berkeley: University of California Press.

14.

Jenkins

(1998). Moral panic: Changing concepts of the child molester in modern America. New Haven, CT: Yale University Press.

15.

Kruttschnitt

Uggen

Shelton

(2000). Predictors of desistence among sex offenders: The interaction of formal and informal social controls. Justice Quarterly, 17, 61-88.

16.

Levenson

J. S.

D’Amora

D. A.

Hern

A. L.

(2007). Megan’s Law and its impact on community re-entry for sex offenders. Behavioral Sciences & the Law, 25, 587-602.

17.

Lieb

Quinsey

Berliner

(1998). Sexual predators and social policy. Crime and Justice, 23, 43-114.

18.

Lussier

(2005). The criminal activity of sexual offenders in adulthood: Revisiting the specialization debate. Sexual Abuse: A Journal of Research and Treatment, 17, 269-292.

19.

Lussier

LeBlanc

Proulx

(2005). The generality of criminal behavior: A confirmatory factor analysis of the criminal activity of sex offenders in adulthood. Journal of Criminal Justice, 33, 177-189.

20.

Magers

Jennings

Tewksbury

Miller

J. M.

(2009). An exploration of the sex offender specialization and violence nexus. Southwest Journal of Criminal Justice, 6, 133-144.

21.

Miethe

T. C.

Olson

Mitchell

(2006). Specialization and persistence in the arrest histories of sex offenders: A comparative analysis of alternative measures and offense types. Journal of Research in Crime & Delinquency, 43, 204-229.

22.

Parent

Guay

J. P.

Knight

R. A.

(2011). An assessment of long-term risk of recidivism by adult sex offenders: One size doesn’t fit all. Criminal Justice and Behavior, 38, 188-209.

23.

Paternoster

Brame

Piquero

Mazerolle

Dean

C. W.

(1998). The forward specialization coefficient: Distributional properties and subgroup differences. Journal of Quantitative Criminology, 14, 133-154.

24.

Petrosino

A. J.

Petrosino

(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.

25.

Quinn

J. F.

Forsyth

C. J.

Mullen-Quinn

(2004). Societal reaction to sex offenders: A review of the origins and results of the myths surrounding their crimes and treatment amenability. Deviant Behavior, 25, 215-232.

26.

Sample

L. L.

Bray

T. M.

(2003). Are sex offenders dangerous? Criminology & Public Policy, 3, 59-82.

27.

Schram

D. D.

Darling Milloy

(1995). Community notification: A study of sex offender characteristics and recidivism. Olympia: Washington State Institute for Public Policy.

28.

Simon

L. M.

(1997). The myth of sex offender specialization: An empirical analysis. New England Journal of Criminal and Civil Confinement, 23, 387-403.

29.

Simon

L. M.

(2000). An examination of the assumptions of specialization, mental disorder, and dangerousness in sex offenders. Behavioral Sciences & the Law, 18, 275-308.

30.

Soothill

Francis

Sanderson

Ackerley

(2000). Sex offenders: Specialists, generalists—Or both? A 32-year criminological study. British Journal of Criminology, 32, 325-337.

31.

Terry

K. J.

Ackerman

A. A.

(2009). A brief history of major sex offender laws. In Wright

R. G.

(Ed.), Sex offender laws: Failed policies, new directions (pp. 65-98). New York, NY: Springer.

32.

Vasquez

B. E.

Mannan

Walker

J. T.

(2008). The influence of sex offender registration and notification laws in the United States. Crime & Delinquency, 54, 175-192.

33.

Weinrott

Saylor

(1991). Self-report of crimes committed by sex offenders. Journal of Interpersonal Violence, 6, 286-300.

34.

Welchans

(2005). Megan’s Law: Evaluations of sexual offender registries. Criminal Justice Policy Review, 16, 123-140.

35.

Zevitz

R. G.

Crim

Farkas

(2000). Sex offender community notification: Managing high risk criminals or exacting further vengeance? Behavioral Sciences & the Law, 18, 375-391.

36.

Zilney

L. J.

Zilney

L. A.

(2009). Perverts and predators: The making of sexual offending laws. Lanham, MD: Rowman & Littlefield.

37.

Zimring

F. E.

(2004). An American travesty: Legal responses to adolescent sex offending. Chicago, IL: University of Chicago Press.