Abstract
Sexual violence is a significant public health problem in the United States. In an effort to decrease the incidence of sexual assault, legislators have passed regulatory laws aimed at reducing recidivism among convicted sexual offenders. As a result, sex offenders living in the United States are bound by multiple policies, including registration, community notification, monitoring via a global positioning system, civil commitment, and residency, loitering, and Internet restrictions.
These policies have led to multiple collateral consequences, creating an ominous environment that inhibits successful reintegration and may contribute to an increasing risk for recidivism. In fact, evidence on the effectiveness of these laws suggests that they may not prevent recidivism or sexual violence and result in more harm than good.
THE INCIDENCE OF SEXUAL assault in the United States is alarming. Every year, an estimated 300 000 women are raped1,2 and 3.7 million are confronted with unwanted sexual activity.3 In addition, of the approximately 900 000 children who are maltreated each year, 9% are sexually abused.4
The physical and mental health problems experienced by survivors make sexual assault more than a criminal justice concern but a public health concern as well.3 As such, a continued focus on prevention-based policy is needed. Over the past 14 years, legislation has evolved to ensure this focus, but the effectiveness of these policies in curbing the incidence of sexual violence is questionable. I review the current status of laws related to registered sexual offenders (RSOs) and discuss why they may be ineffective in preventing sexually violent crimes.
Since the early 1990s, in an effort to reduce the prevalence of sexual crimes, a significant public health problem, federal, state, and local legislation related to sexual offenders has proliferated. Initially, the Jacob Wetterling Crimes Against Children Act and Sexually Violent Offender Registration Act5 was created to help law enforcement officials track sex offenders and thus, theoretically, reduce the likelihood that they would recidivate. Under this regulatory law, convicted sex offenders were obligated to register and verify their current names and addresses with local police.
In May 1996, Megan's Law6 was passed, amending the Wetterling Act by requiring states to establish systems for making registry information available to the public through methods of community notification. In doing so, Megan's Law made photographs, names, and addresses of registered sex offenders available to the public via the Internet and other forms of community notification.
On July 27, 2007, President George W. Bush signed the Adam Walsh Protection Act (AWA), a federal law mandating state registration requirements “in response to the vicious attacks by violent predators against children.”7 AWA, which has 6 titles, streamlines the tracking of sex offenders and notification requirements at the federal level. Under AWA, all states were required to have all or a large portion of AWA implemented by 2009 or risk losing 10% of their Byrne grant funds (money used by law enforcement agencies for crime prevention initiatives). Interestingly, a recent analysis indicated that the cost of implementing AWA far outweighs the loss states would incur by the reduction in Byrne grant funds.8 Since most states were unable to implement AWA by 2009, the deadline was extended to July 27, 2010.
The first title of AWA, the Sex Offender Registration and Notification Act, provides states with minimum registration and community notification procedures in their management of registered sex offenders. For some registered sexual offenders, the AWA mandates will result in increased or changed registration duties. Some of the requirements under AWA include but are not limited to the following:
Classifying registered sex offenders via a 3-level tier system based on the crime committed versus their risk for reoffense (registered sex offenders who committed lower level offenses would be classified as tier 1; those who committed moderate-level offenses as tier 2; and those who committed higher level offenses as tier 3);
Making kidnapping and false imprisonment of a child, regardless of sexual intent, a registerable sexual offense;
Requiring registered sex offenders to register in any jurisdiction, not only where they live, but also where they work or attend school;
Requiring sex offenders to verify their addresses once per year for those at tier 1, twice a year for those at tier 2, and 4 times per year for those at tier 3;
Making failure to register a felony offense punishable by a maximum of 10 years in prison;
Giving a registered sex offender 3 days to report a change of address to law enforcement agencies;
Requiring that registered sex offenders' entire criminal history, not just the sexual offense, finger prints, palm prints, and a DNA sample be reported to law enforcement agencies;
Mandating that the registry be made available on the Internet and that all tier levels, including tier 1, be subject to such community notification;
Mandating the length of time a registered sexual offender would be required to register; offenders at tier 1 would register for 15 years, offenders at tier 2 for 25 years, and offenders at tier 3 for life; and
Providing additional funds to support offices, software, training, and additional personnel to enforce registered sex offender laws.
In addition, Title III of AWA mandates civil commitment for certain dangerous offenders, and Title IV makes it illegal for a registered sex offender to sponsor a family member, such as a wife, husband, or child for permanent residency in the United States. As with Megan's Law, AWA was developed by the government in an effort to protect the public from registered sex offenders who are likely to repeat their offenses.
These laws are indicative of the government's effort to be “tough on crime.” On the surface, they have intuitive appeal. Who would not want to take the toughest measures to protect children and families from someone who has committed a sexually based offense? However, intuition is not science, and a closer look at the laws' purpose, intent, and outcomes reveals that problems do exist.
The philosophy of the Wetterling Act, Megan's Law, and AWA is that society must be protected against a group of dangerous criminals who pose a high risk for reoffense. According to the Centers for Disease Control and Prevention (CDC), “sexual violence perpetrators are … at increased risk of perpetrating again.”1 Not mentioned by the CDC, however, is that the risk is still quite low,9–14 with most criminal justice and community-based studies showing that registered sex offenders are rarely returned to prison for another sexual crime.15–18 A survey I conducted of 89 registered sex offenders participating in online support and education forums revealed a 0% self-reported recidivism rate (Bonnar-Kidd, unpublished data, 2009).
Furthermore, in New York, of the 11 898 registered sex offenders released from prison between 1985 and 2001, 251 (2.1%) were returned to prison for another sex crime.17 The Arizona Department of Corrections19 reported that between 1984 and 1998, the recidivism rate for sex offenders was 5.5%, and Ohio20 reported that sex offenders released from prison in 1989 had a 10-year recidivism rate of 8%. According to the US Department of Justice,18 registered sex offenders are the least likely class of criminals to reoffend, with 3.5% of registered sex offenders released from prison in 1994 being reconvicted for another sexual offense within 3 years of their release. Finally, Harris and Hanson16 found that the risk for recidivating decreases significantly over time, with most reoffenses occurring within 5 years of the original conviction.
Given the limitations in the recidivism data, it is difficult to make definitive conclusions about the rate at which sex offenders repeat their crimes. The definition or interpretation of recidivism varies in many studies, with some using arrest or conviction data for any crime, such as a parole violation, and others using arrest or conviction data for a sexually based crime. Furthermore, even the definition of sexual assault varies. The Uniform Crime Reports collected by the Department of Justice have methodological problems and typically only include data on violent rape, when sexual assault includes many other acts such as statutory rape, possession of child pornography, and indecent exposure. These inconsistencies can result in either overestimates or underestimates of recidivism rates.
That most sexual crimes go unreported is another significant limitation of the recidivism data.21 For example, in the United States in 2007, an estimated 42% of rapes or sexual assaults were never brought to the attention of authorities.22 Unfortunately, the number of unreported crimes committed by registered sex offenders relative to the number among individuals without a record of sexual offenses is unknown. We do know, however, that most new sexually based crimes are committed by someone not on the registry. In Ohio in 1999, 92% of those convicted of a sex offense against a child and 93% of those convicted of a sex offense against a teenager were first-time offenders.23 Most recently, Sandler et al.,24 in their analysis of the effectiveness of Megan's Law in New York, reported that 96% of all new arrests for sexual crimes occurred among those without previous sexual crime convictions.
As such, although it is reasonable to conclude that many sexually based crimes go unreported, it may be incorrect to assume that the bulk of these unreported crimes are being committed by the 650 000 registered sex offenders in the United States today. With that in mind, the underreporting of sexual crimes could mean that recidivism data are underestimated.
These laws may also be limited by their focus on the prevention of sexual victimizations by strangers,25 a concept known as “stranger danger.” Levenson et al.26 surveyed members of the public to assess their perceptions of registered sex offenders and found that respondents believed approximately 42% of sexual crimes were committed by a stranger. However, findings from the Second National Incidence Studies of Missing, Abducted, Runaway, and Thrownaway Children revealed that 115 stranger abductions occurred in the United States in 1999; in 56 of those cases, the victim was sexually assaulted, indicating that only a small percentage of sexually based crimes against children occur via stereotypical abductions. It is further estimated that 93% of sexually driven crimes are committed by a family member or someone known to the victim.4,27 The fear of “stranger danger” is misplaced and should not be used to justify the proliferation of registered sex offender laws.
Another common theme driving the proliferation of laws is the perception that rates of sexual crimes, pervasive in our society, are higher now than ever (L. Sample, PhD, unpublished data, 2001). This statement is difficult to operationalize because no formal procedures exist for uniformly reporting sexually based crimes with the exception of forcible rape, leaving national trends in other registerable offenses such as statutory rape, voyeurism, and indecent exposure essentially immeasurable. With this in mind, forcible rapes declined significantly in the early 1970s, when violent crime in general was decreasing. The Bureau of Justice Statistics reported a decreasing trend in rapes from 2.5 per 1000 people in 1973 to 0.5 per 1000 people in 200528 It seems the risk for being raped is lower today than in the past and is thus not pervasive in our society. However, after decades of decreasing rape trends, recent data indicates that, since the passage of community notification and other laws, they may be increasing.
Initial data on sexual assault trends since the inception of Megan's Law in the late 1990's suggest that rates have not significantly decreased29,30 and that, in many states such as Virginia, Florida, Pennsylvania, New York, Louisiana, Vermont, Indiana, and Illinois, rates may even be increasing.31–38 Similar trends have been seen in studies examining the effectiveness of Megan's Law, indicating these laws may not be protecting the public or serving their purpose of protecting against sexual assault.
Empirical research examining Megan's Law has generally indicated that community notification is not effective in preventing sexually based crimes24,29,39–46 and may actually create a context wherein the risk of recidivism increases.47 Barnowski,48 in his study of recidivism among 8000 sex offenders in Washington, found that the 5-year recidivism rate among those released from prison in 1990 was 5% higher than the rate among those released in 1997, the year community notification laws were implemented. That study's results must be interpreted with caution because the reduction in recidivism for registered sex offenders mirrored a statewide trend in reduced recidivism for other types of crimes. Recidivism rates were declining prior to community notification laws, and after an 11-year downward trend, the recidivism rates of registered sex offenders in Washington began to increase in 1997, the year Megan's Law was implemented. More recently, effectiveness studies from New Jersey44 and New York24 concluded that Megan's Law has had no significant impact on rates of recidivism or sexual violence, suggesting that the costs of implementing such laws may outweigh the benefits.
Issues surrounding the effectiveness of Megan's Law would be unimportant if the consequences had little impact. However, community notification has many collateral consequences for both community members and registered sex offenders. According to Levenson et al.,26 it is likely that parents experience fear after notification of a neighborhood sex offender. This fear can lead to community-wide hysteria, which has occurred in many towns.
In June 2007, for example, members of a small, rural community in New York were notified that 2 registered sex offenders were residing in their neighborhood and reacted by posting signs warning that “Monsters Live Here”; they also initiated a protest using the media and contacted the registered sex offenders' landlord asking that they be evicted. One resident noted that
since he explained to his daughters about the ‘monsters’ across the street, he has seen a look of fear and terror in their faces, and they sometimes refuse to sleep alone.49(pB1)
Another resident published a letter to the editor online in North Country This Week (St. Lawrence County, New York) in response to her new registered sex offender neighbor. One part of the letter read
I have slept approximately a combined 5 hours in the past 3 nights because I wake up in a panic and I need to get up and make sure that my children are okay.50
This hysteria and panic is a collateral consequence of community notification. These kinds of reactions have led to a proliferation of registered sex offender laws above and beyond community notification. Since 1996, some states have created “safety zones,” or places where a registered sex offender cannot be; others are mandating lifetime global positioning system (GPS) monitoring. Still others have successfully passed laws banning registered sex offenders from wearing Halloween costumes or mandating them to be indoors with outdoor lights off on Halloween night. In Louisiana in 2008, Governor Bobby Jindal signed SB 144, a bill making chemical castration through administration of the hormone medroxyprogesterone mandatory for certain offenders. In other states, registered sex offenders are subject to civil commitment, have been banned from using the Internet, have “sex offender” imprinted on their driver's licenses or license plates, and have lost all parental rights. Many registered sex offenders now also face restrictions related to employment and loitering and, most widespread, restrictions in where they can live.
Because community members often react with fear to the presence of a registered sex offender, the mentality labeled “NIMBY” (not in my backyard) has led to the evolution of laws restricting where registered sex offenders can live. As of August 2006, more than 20 states and hundreds of localities nationwide had passed residency restrictions, and many more were considering them.51,52 Usually passed retroactively, residency laws restrict registered sex offenders from living within a certain number of feet from schools, day-care centers, and churches, and their passage has produced much controversy. In California, Proposition 83 (also known as Jessica's Law) was passed in 2007 to limit registered sex offenders from living within 2000 feet of a school or a park. As a result of this ordinance, approximately 2700 registered sex offenders were told to move, with many ending up homeless.53
In some cases, residency restrictions are so severe that they essentially banish a registered sex offender from living anywhere in the city. In Miami, Florida, for example, residency restrictions were so strict—2500 feet from schools, playgrounds, licensed day-care centers, and parks—that the only location registered sex offenders' probation officers would approve for housing was underneath the Julia Tuttle Causeway, a bridge connecting Miami Beach to Miami.54
These laws have intuitive appeal. Who would want a registered sex offender living near children? When communities successfully get them to move, community members' fear subsides, thus making them feel safe. But there are logical problems with residency restrictions that could result in communities having a false sense of security.
Residency restrictions could violate registered sex offenders' fundamental human and constitutional rights, for example. In most cases, the laws are being applied retroactively to those who have served their time, which is a likely violation of ex post facto application of new laws as well as rights against double jeopardy. A case in Tippecanoe County, Indiana, serves as an example. A nonrecidivating 56-year-old man convicted in 1988 of a sex crime petitioned the court to relieve him of his duty to abide by residency restrictions. He had lived in his home for 7 years, a home that he shared with his wife and children; in July 2007, he was forced to move because his home was within 1000 feet of a church, a violation of the new residency restrictions.55 Although the case is being appealed, a judge ruled that because the registered sex offender's wife owns the house, the residency restriction did not violate his property rights56 and that residency restrictions, being a regulatory measure, do not violate registered sex offenders' constitutional rights.
Residency restrictions were developed on the basis of the assumptions that (1) registered sex offenders are at a high risk for recidivism, (2) sexual crimes are committed by strangers who lurk in areas where children congregate in an attempt to stereotypically abduct them, (3) all registered sex offenders have committed crimes against children, and (4) children and families are protected from sexual crimes if a registered sex offender does not live in their neighborhood. I addressed the first 2 assumptions earlier; recidivism rates among registered sex offenders are generally low, and most sexually based crimes are committed by someone known to the victim. As reported by Levenson and Cotter57(p170) in their study of registered sex offenders' perceptions of residency restrictions, “Most abuse happens in homes or with family or close friends, not at bus stops or schools.” The third assumption is also a public perception or myth created by the media.
Individuals who are required to register as sex offenders are those who have committed crimes against not only children but adults as well. Although the list of registerable sex offenses varies by state, registered sex offenders are also classified as those who have, for instance, possessed child pornography, solicited prostitution, participated in exhibitionism, or engaged in indecent exposure (including urinating in public), voyeurism, or oral or anal sex. Juveniles who have had consensual sexual relations with another juvenile also frequently fall under this category. In the United States in 2001, for example, children and adolescents younger than 18 years were arrested at a higher rate than any other age group (e.g., 140.7 per 100 000 arrests among those aged 13 to 14 years versus 74.6 per 100 000 arrests among those aged 40 to 44 years).58 Although in the media we hear most often about sexual crimes committed against children, these are not the only crimes that lead to designation as a registered sex offender.
Furthermore, as mentioned, residency restrictions may provide a false sense of security in communities where registered sex offenders do not live.46 People residing in registered sex offender-free areas are not automatically protected against sexual abuse.26 In fact, evidence suggests that residency restrictions do not affect rates of sexual assault and therefore are ineffective.
The Minnesota Department of Corrections examined the “sexual reoffense patterns of 224 recidivists released between 1990 and 2002 who were reincarcerated for a sex crime prior to 2006”59 and concluded that not one of the new offenses would have been prevented if residency restrictions had been in place. This makes intuitive sense. Residency restrictions simply mandate where a registered sex offender can and cannot live. It is possible for those few who seek to reoffend to drive or walk to a location if their intent is to commit another sex offense. In fact, the Minnesota Department of Corrections report acknowledged that “when direct contact offenders look for a victim, they are more likely to go to an area relatively close to home (i.e., within 20 miles of their residence), but still far enough away (i.e., more than 1 mile) to decrease the chances of being recognized.”59
Many states have attempted to manage fear or hysteria by establishing procedures for determining and notifying the community about the recidivism risk posed by registered sex offenders. Typically, there are 3 risk levels, high, medium, and low, differentiated according to a registered sex offender's “dangerousness” or risk for committing another sex offense. Those who are classified as high risk have a high potential for recidivating, whereas those who are classified as low risk are not likely to recidivate. The outcome of a high-risk classification is severe and includes but is not limited to additional registration requirements, community notification, residency restrictions, GPS monitoring, and potential civil confinement. Determinations of risk level are usually made on the basis of the outcome of an actuarial assessment.
Many risk-level determination assessments are in use in the United States, including the STATIC-99 and the Rapid Risk Assessment for Sexual Offense Recidivism.60 Unfortunately, rarely can one ever predict with 100% accuracy the future behavior of a registered sex offender, even when using these actuarial assessments. However, the chances of correctly predicting future behavior increase when the instruments are valid.
Although a thorough discussion of the psychometric properties of these tools is beyond the scope of this review, research indicates that the validity of the tools is questionable.61–63 The 3 best tools in use today have only moderate levels of predictive validity (i.e., area under the curve values of 0.64–0.70),64 which in the field of public health is generally acceptable. However, moderate validity (a 30%–36% likelihood of error) should not be acceptable for instruments designed to limit freedoms and impose additional regulatory restrictions against registered sex offenders. In light of this, one must question whether moderate validity is acceptable in these cases. For example, Barnowski, in his study examining the relationship between risk levels and recidivism among registered sex offenders in Washington, concluded that “[t]he notification levels determined by the [End of Sentence Review Committee] do not classify sex offenders into groups that accurately reflect their risk for reoffending.”62
Registered sex offenders face many integration barriers as a result of their designation; for example, limited access to housing, education, and employment as well as community segregation and harassment, which could increase their risk for committing additional crimes, including subsequent sex crimes. In 2000, the Campus Sex Crimes Prevention Act65 was implemented, requiring all institutions of higher education to monitor registered sex offenders who register at their schools. Many institutions place additional admissions restrictions on registered sex offenders. At Eastern Oregon University, registered sex offenders can be denied admission if their coursework requires them to have close contact with an individual in a private setting.66
In addition, at the State University of New York, registered sex offenders who apply for admission must go before a committee to explain their crimes and describe the measures they have taken to ensure rehabilitation. If the school perceives that the risk to the campus community is high, then admission can be denied. If the registered sex offender is granted acceptance, the campus police must take measures to notify the campus community, including posting the information on a Web site, contacting classmates and professors, and posting flyers around campus.67
Also, whereas it is challenging in general for a person with a criminal history to find employment, it may be even more difficult for a registered sex offender to do so.68–70 There has been much qualitative research on the employment effects of a registered sex offender designation. For example, the Human Rights Watch conducted a 2-year study on the collateral consequences of the designation and reported the effects that it has on registered sex offenders' ability to secure employment and thus income. Results showed, more often than not, that registered sex offenders have a difficult time holding a job. The report cited cases of individuals who were fired from long-term employment when their status as a registered sex offender became public.45 “The Predator Next Door,” 71 an NBC News documentary, tells the story of a registered sexual offender whose employment was terminated because of community notification and exemplifies this collateral consequence. In some states, laws mandate that employer information be included as part of any community notification. This could also serve as a deterrent to employment.
A final consequence for registered sex offenders is vigilantism, ostracism, and community segregation. Explicit within Megan's Law is the disclaimer that the registry cannot be used to threaten or harass registered sex offenders. However, many such cases have occurred. Sample and Kadleck72 interviewed 35 Illinois legislators to examine their perceptions of sex offenders and how those perceptions might influence policy. According to one legislator:
I wouldn't say this in some crowds, but we have documented cases of vigilantism with people going to the wrong house or beating up the wrong guy. That's what I was afraid of. We are supposed to be stopping violence, not promoting it, and what does it promote when you tell everyone where these guys live?72(p56)
A quick search for stories about the reactions of communities to the public registry reveals many examples of such unintended consequences. For example, in Michigan a registered sex offender was beheaded and his body burned by a group of teenagers.73 In Helenwood, Tennessee, in September 2007, the wife of a man died after 2 neighbors set their house on fire, an action prompted by the man's recent arrest for possession of child pornography.74 In April 2006, a man traveled from Canada to Maine intending to kill 34 registered sex offenders whose names he had collected from the Maine Sex Offender Registry. He murdered 2 of them before he was caught.75 In other, less extreme cases, registered sex offenders are protested against, with communities posting signs warning of the offender in the neighborhood. In their study of the effects of Megan's Law on reintegration, Levenson and Cotter found that one third of their participants had experienced physical threats, and most reported negative effects of community notification including “stress, isolation, loss of relationships, fear, shame, embarrassment, and hopelessness.”76(p49)
The collateral consequences do not end with registered sex offenders, but continue with their families. Registered sex offenders' family members also suffer the consequences of registration, community notification, and residency restrictions. When registered sex offenders are released from prison, rarely are they completely alone. Most if not all have a family—mothers, fathers, brothers, sisters, aunts, uncles—and many even have children of their own. Research on the collateral consequences of Megan's Law for family members is limited, but Levenson and Cotter found that 67% of registered sex offenders in their study reported that their families suffered emotional distress as a result of community notification.76 This finding indicates that, whereas some may advocate the need for “shame and blame” effects of community notification, rarely do we consider the impact, by proxy, on the registered sex offender's family.
Although it is difficult to find compassion for individuals who have committed sexually based offenses, the history of the United States proves that segregating a class of citizens on the basis of emotionally driven laws is risky considering that the empirical evidence supporting their effectiveness is sparse. There is also evidence that these laws could be doing more harm than good.51 Study findings suggest that those with histories of any kind of criminal offense reintegrate more successfully when they are offered social support and opportunities to reintegrate into society through housing and employment.47,76,77 Although this has yet to be empirically tested, because registered sex offenders are often denied these basic essentials of life, the increasing limits on their rights and freedoms could increase instead of decrease their risk for recidivism.
Therefore, those with a passionate interest in preventing sexual crimes should work to provide ex-offenders with stable community support that can assist in their success. Because of the proliferation of laws directly related to registered sex offenders, however, this may prove a difficult task. For example, according to the Center for Sex Offender Management:
[M]yths about sex offenders and victims, inflated recidivism rates, claims that sex offender treatment is ineffective, and highly publicized cases involving predatory offenders fuel negative public sentiment and exacerbate concerns by policymakers and the public alike about the return of sex offenders to local communities. Furthermore, the proliferation of legislation that specifically targets the sex offender population—including longer minimum mandatory sentences for certain sex crimes, expanded registration and community notification policies, and the creation of “sex offender free” zones that restrict residency, employment, or travel within prescribed areas in many communities—can inadvertently but significantly hamper reintegration efforts.78
It is clear from this review of recent policies enacted to protect communities from sexual violence that the proliferation of well-intentioned political efforts to curb sexual violence has led to the creation of laws lacking a solid evidence base. Although additional community-based studies are needed, research to date indicates that after 15 years the laws have had little impact on recidivism rates and the incidence of sexually based crimes. The most significant impact of these laws seems only to be numerous collateral consequences for communities, registered sex offenders (including a potential increased risk for recidivism), and their family members.
Such findings have led many to question the benefits of this legislation in light of the costs of managing registered sex offenders in our communities. Tertiary resources to manage registered sex offenders after their release may be better spent on measures designed to offer additional assistance to victims of sexual violence, identify first-time offenders through community- and school-based educational programs, provide counseling to young people with risk factors or tendencies for sexual violence, and fund efforts to research and advocate rational and rehabilitative, evidence-based laws.
Human Participant Protection
No protocol approval was necessary.
