Sarah W. Craun
University of Tennessee, College of Social Work
Poco D. Kernsmith
Wayne State University, School of Social Work
Method
Results
Discussion
Conclusion
SEXUAL OFFENDERS WHO commit further crimes have been a cause of fear among the public for some time. In an attempt to lessen the numbers of them who recommit crime, in 1996 the federal government implemented a law that required sex offenders to register with a local law enforcement agency and also required that the registry be made public. The stated intent for offender registration was to deter sexual offending, punish and incapacitate sex offenders, provide law enforcement officials with an information tracking tool, and to increase the safety of the public (Farkas, 2002). The law, commonly known as Megan’s Law, is quite popular with the public (Proctor, Badzinski & Johnson, 2002; Phillips, 1998). One survey in Washington State found that more than 8 out of 10 respondents ranked the law as “very important” (Phillips, 1998). Even though the law is strongly backed by the general public, there is a debate about whether juveniles who commit sex offenses should be required to register and whether that information should be made public.
Data regarding adult sex offenders indicates that, among sex offenders who have been released from prison, 43 percent were rearrested for another crime and 5.3 percent of them were arrested for another sex crime within three years after their release (Langan, Schmitt & Durose, 2003). In a meta-analysis of research on sexual recidivism, also among adult offenders, Hanson and Bussiere discovered that 13 percent of convicted sexual perpetrators offended again within four to five years (1998). That percentage dropped only one percentage point when investigating recidivism of sexual assault offenders who targeted children (Hanson & Bussiere, 1998). Research by Lisak and Miller (2002) showed that 120 adult individuals were responsible for a total of 1,225 acts of interpersonal violence, defined as rape, battery, child physical abuse, and child sexual abuse. This translated into an average of 5.8 assaults per offender that were not reported to law enforcement officials (Lisak & Miller, 2002).
The characteristics of the crimes and the victims are varied for both adult and juvenile sex offenders. Half of child molesters were more than 20 years older than their victims and most of their victims were under the age of 13 (Langan, Schmitt & Durose, 2003). The National Crime Victimization Survey (NCVS) illustrated that, of those people who were raped by a single offender, 10.8 percent of victims said the offender was under the age of 18 (Greenfeld, 1997). In a review of studies about juvenile sex offenders, researchers summarized that juvenile sex offenders are more likely to target victims much younger than themselves as compared to peers, and their victims tend to be female (Rightland & Welch, 2001). Moreover, 40 percent of those juvenile perpetrators abused victims under the age of six (Snyder, 2000). In a comparison between male and female juvenile offenders, it was found that, compared to males, females were more likely to target victims of the same sex (Vandiver & Teske, 2006). It should be noted, however, that sex crimes committed by juveniles are small in number. Stahl (2001) estimates that less than one percent of the cases in juvenile court are for forcible rape or other violent sex offenses. Additionally, the FBI reported that those under age 18 accounted for 16.2 percent of the forcible rape arrests in the United States in 2004. One should remember that these statistics are only applicable to those who were caught sexually offending.
In the debate over how to handle juvenile sex offender registration, some scholars have invoked the history of the juvenile justice system. In a summary of its foundation, Greenwood (2002) wrote that the juvenile justice system was intended to reflect the best interests of the child. Unlike criminal courts for adults, the juvenile court system was designed to give treatment and guidance for the young offender, rather than punishment (Greenwood, 2002). What are the implications of this regarding the registration of juveniles as sex offenders under Megan’s Law? Zimring argues in his book American Travesty: Legal Responses to Juvenile Sex Offending (2004) that the juvenile court system has typically been set up to protect juvenile offenders and do what is best for them; however, having juveniles register as sex offenders contradicts that goal. Trivits and Reppucci (2002) share Zimring’s concern, arguing that states requiring juveniles to register as sex offenders lose sight of the original intention of the juvenile justice system, because the goal of the registry is not to rehabilitate the offender, but rather to provide a sense of safety to the community. Furthermore, Letourneau and Miner (2005) hypothesize that applying sex offender policies such as registration and community notification to juveniles may cause adverse consequences for them. Registries may increase the likelihood of future offending by increasing the social isolation of offenders, a factor highly correlated with re-offending (Gutierrez-Lobos et al., 2001). Zimring (2004) does not deny that serious juvenile offenders need to be adjudicated through the juvenile court system, but he appears doubtful that registries are an effective deterrent, even among serious juvenile offenders. Zimring (2004) calls for informed research on this topic when he wrote, “there have been no extensive efforts to compare the characteristics and motivations of adolescent sex offenders with different types of adult offenders” (p. 55).
While this study does not measure motivation, it does purport to compare juvenile and adult sex offenders who are required to register as sex offenders under Texas state law. This descriptive study is intended to discover if there are significant differences between juvenile and adult offenders on a sex offender registry. It is hypothesized that there should be no difference between registered juvenile and adult offenders in terms of offender race and sex, crime committed, or risk level assigned by the state. One may expect to find differences with victim sex, as previous research has shown that juvenile offenders have a higher proportion of male victims as compared to adults (Aljazireh, 1993; Davis & Letienberg, 1987) and the average age of the victim, with juveniles hypothesized to be more likely to have younger victims. Moreover, it is expected that the age differential will be smaller between juvenile offenders and their victims, due to the juvenile offenders’ smaller age range.
This study compared the characteristics of adult and juvenile offenders and their respective victims on the State of Texas sex offender registry. Texas provided the dataset in February 2004 for secondary analysis. After removing observations that had missing values on the above areas of interest, the sample was comprised of 41,979 sexual offenses committed by 36,347 offenders.
Variables
To establish a dividing line between adult and juvenile offenders, the authors used the definition from the State of Texas: a juvenile, for criminal purposes, is age 17 or younger (Texas Department of Public Safety, 2004). This dataset did not include the dates of the sexual crimes, and consequently, the age of the offender was determined by using the disposition date and birth date, thus determining the age of the offender at the time
of disposition.
Risk level, as determined by the State of Texas, was included as a variable of interest. As dictated by law, offenders were rated as being low, moderate or high risk by a screening tool decided upon by a “risk assessment review committee” (Texas Code of Criminal Procedures, 2006). Approximately half of the sex offenders are given risk levels by the State of Texas. Offenders who were not assigned a risk level by the state are also included as a separate category.
Another variable of interest was crime committed by the offender. For ease of analysis, the crimes of which the offenders were convicted were grouped to narrow the categories from the twenty original categories to eight categories. Sexual assault, sexual assault with a child, indecency with a child —exposure, indecency with a child—sexual contact, and aggravated sexual assault—child victim remained as unique categories. The categories “kidnapping (victim under 17 years of age),” “aggravated kidnapping (with intent to violate),” and “aggravated kidnapping (with intent to violate victim under 17 years of age)” were combined to form “kidnapping.” “Aggravated sexual assault” and “aggravated sexual assault – Victim 65 years or older” became aggravated sexual assault/adult victim. Finally, the “other” category is comprised of all other sexual offenses with very small sample sizes, such as compelling prostitution, court/board-ordered registration, indecent exposure (second conviction), and prohibited sexual conduct (incest).
According to Texas law, “indecency with a child” is defined as sexual contact with a minor under the age of 17 by an offender who is at least three years older than the victim (Zimring, 2004). This theoretically should eliminate most of those offenders who had consensual sex, but where the victims were below the age of consent (17). The number of sexual offenses committed by each offender was tallied to produce a count variable.
Finally, demographic, offender, and victim characteristics from the sex offender registry were included: race, sex, and age of the offender. Victim characteristics, including sex and age, are included, except for victims under one year old or when the age was missing from the dataset. A variable defining the age difference between a victim and offender was established by subtracting the victim’s age from the offender’s age.
Analysis
T-tests for unequal variances were used to compare the mean ages of victims and the age gap between victims and offenders. Chi-squared tests of statistical significance were used on the bivariate tables between the independent and dependent categorical variables. Fisher’s exact test was used when the observations in a cell were less than five. Adjusted residuals were used to determine which specific cells in a table deviated significantly from independence. Agresti and Finlay (1999) suggest that an adjusted residual greater than 2.0 indicates that the variables are not truly independent. Results
Differences between the Characteristics of Juvenile and Adult Offenders
Of the 36,347 offenders in the study, 91.45 percent were adults at the time of disposition and 8.55 percent were juveniles. Nearly all of the offenders on the registry were male, whether they were adults (98 percent) or juveniles (98 percent). The mean age for juveniles at disposition was almost 15 years of age (SD = 1.4), as compared to a mean age of 33 years (SD = 12) for adult offenders. The majority of both adult and juvenile offenders were classified as White. There were a significantly higher percentage of Black juvenile offenders (25 percent) on the registry as compared to Black adult offenders (21 percent) and a lower percentage of White offenders listed as juveniles (75 percent) as compared to adults (79 percent) (p < .001).
Finally, juveniles were also classified as higher risk than adult offenders (p < .001). Adjusted residuals show that juveniles are more likely to be considered moderate risk than adult offenders. Comparing only low- and moderate-risk offenders, the odds of a juvenile being classified as a moderate-risk offender were 3.88 greater than those of an adult being classified as a moderate risk.
When examining the number of offenses per offender listed on the registry, most (88 percent of adults and 84 percent of juveniles) had committed only one offense. Among those with multiple offenses, juveniles had committed significantly more offenses (p < .001). This is unexpected, since older offenders would have had more years to offend. However, it is possible that youth were more likely not to be adjudicated until they had committed multiple offenses.
Differences between the Offenses
Committed by Juvenile and Adult
Registered Offenders
Table 1: Texas Registered Sex Offenders by Adult/Juvenile Offenders (Adjusted Residual)
Table 2: Sexual Offenses by Registered Sex Offenders for Adult/Juvenile Offenders (Adjusted Residual)
The crimes committed by adults and youth were significantly different at the p < .001 level. Adults were most likely to have committed indecency with a child (36 percent). Juveniles were most likely to have been convicted of aggravated sexual assault against a child (40 percent). Juveniles were significantly more likely to have committed aggravated sexual assault against either a child or an adult as compared to an adult offender. Adults were more likely to commit sexual assault against an adult or child than juveniles.
The age of the victims ranged from one to over ninety years for adults and one to eighty-five years for juveniles. As expected, the mean age of the victim was higher for adult offenders (13.6 years, SD = 7.9) than for juvenile offenders (8.3 years, SD = 4.8) (p < .001). The age difference between offender and victim was much larger for adult offenders than for juveniles. (p < .001). On average, adults were 20 years older (SD = 15) than the victim and juveniles were 6 years older (SD = 4.7).
The gender of the victim was also significantly different for adult and juvenile offenders. While female victims were most common for both adult (89 percent) and juvenile (66 percent) offenders, males were more likely to be victimized by juveniles than by adult offenders (p < .001). Using a cross-product odds ratio equation, we calculate that the odds of being a male victim of a juvenile offender were almost four times higher than those of being a male victim of an adult offender. Discussion
The presented descriptive analysis demonstrates that the adults and juveniles listed on the sex offender registry in the state of Texas differed in a variety of ways. Juveniles were more likely to commit offenses against male victims than were adult offenders. This finding resembles other studies with smaller sample sizes where the majority of juvenile sex offenders have female victims; however, as compared to adult offenders their proportion of male victims is higher (Aljazireh, 1993; Davis & Letienberg, 1987). Additionally, it appears that, based on risk classifications, juveniles who were required to register were considered higher risks to the community than those adults who are required to register. Research has shown that juvenile sex offenders use more force (Miranda & Corcoran, 2000), which may explain the higher proportion of juveniles listed as “moderate” risks. It may also explain why a higher proportion of juvenile sex offenders are identified as having committed aggravated sexual assaults against both children and adults. Additionally, juveniles were found to have committed a higher number of sex crimes per offender as compared to adults, which may also impact the determination of risk level. However, it is feasible that youths may be less likely to be adjudicated until they have committed multiple offenses.
The finding of racial differences between adult and juvenile offenders is perplexing. Namely, it was surprising that African-American juveniles comprised nearly 25 percent of the juvenile sex offender registry while African-American adult sex offenders comprised 20 percent of the adults on the registry. This decrease in proportion was unexpected, as was the increased proportion of white offenders when comparing juveniles and adults. Sampson and Lauritsen (1997) suggest utilizing multilevel analyses that include both environmental and individual level factors in examining racial differences in juvenile justice. Most likely there were environmental and individual explanatory variables that were not available through the sex offender registry, which was used for this exploratory analysis.
Limitations of the Study
One weakness of this study is the calculation that had to be used to determine the age of the offender, using the age at date of disposition. This may have resulted in some offenders being mis-classified because they fit the juvenile category at the time of the assault, but became adults by the date of disposition. The Texas Juvenile Probation Commission (2006) states that an offender who is between 17 and 18, but who committed the crime while under the age of 17, is considered a juvenile. Therefore, since no date of assault was provided, some of the sex offenders who may have committed the crime as a juvenile were coded as adults. Another limitation may be the lack of generalizability to Latinos. There are high percentages of Latinos in Texas, yet because the registry did not include the ethnicity of those on the registry, it excluded some useful and relevant information. Finally, knowledge of additional details of the assault would have been helpful to provide a clearer picture of the differences and similarities between juvenile and adult registered offenders. For example, was the victim an acquaintance or a stranger? What was the location and what were the circumstances surrounding the attack? All of these details would have helped provide a clearer picture of the characteristics of those on registries.
Yet, this study is only a basic building block for the future. More sophisticated studies, which might take a more comprehensive look at the impact of registering on juvenile sex offenders, should be conducted. There are many questions that still need to be answered, such as: What are the underlying mechanisms that show a difference in proportion between the juvenile and adult African-American sex offenders? Are those juveniles who reside in states where their registration is made public more or less likely to commit future crimes than those who live where juvenile data are not published? Does registration prove to reduce recidivism for some types of juvenile offenders, but not for others? For example, are those juveniles who commit sex crimes against children more likely to benefit from this type of intervention/prevention strategy than juveniles who commit crimes against adults? Additionally, little is known about the public perception of juveniles registering for Megan’s Law. Is the public supportive of putting the names and addresses of offenders in their youth on the Internet? Conclusion
The traditional legal system for adult offenders had been identified as using disintegrative shaming with particular approaches like the offender registry, as the registry can involve labeling, stigmatization, and ostracism of offenders (McAlinden, 2005). The dilemma regarding registration of juvenile offenders involves valuing public safety and the protection of vulnerable populations over rehabilitation, and possibly individual rights (Baranoski & Buchanan, 2003; Scott & Gerbasi, 2003). It was our intention that this research add to the informed discourse on sex offender registries and juvenile offenders by illustrating the specific differences between registered youths and adults.
Further examination can help clarify for both the public and public administrators the proper use of this law in its application to juveniles. Empirical policy analysis can help build, or possibly restructure, sex offender registration policy that can deter future sexual assaults and work for the benefit of the public, sexual offenders, and victims.
