Dr. Appelbaum, who is editor of this column, is the Elizabeth K. Dollard Professor of Psychiatry, Medicine and Law, Department of Psychiatry, Columbia University. Send correspondence to him at New York State Psychiatric Institute, 1051 Riverside Dr., Unit 122, New York, NY 10032 (e-mail: email@example.com).
Sex offenders are perhaps the most feared and reviled criminals in our society. Because their behavior is generally viewed as pathologic—and many suffer from paraphilias—the mental health professions are often expected to assess and treat sex offenders and to predict the likelihood of their recidivism. Civil commitment statutes that target a subgroup of offenders designated as "sexually violent predators" have increased in recent years. The statutes have received a great deal of criticism, both on the grounds that they are largely punitive in intent and because of concern about the role of mental health professionals (1).
Most identified sex offenders, however, reside in the community, having completed their prison terms or having been placed directly on probation without incarceration. An even larger body of legislation focuses on this group. Registration requirements and community notification statutes exist in every state, and a growing number of jurisdictions are enacting restrictions on where sex offenders can live and work. Although each of these policy approaches could be useful when targeted appropriately, they are now frequently designed in ways that are wasteful of resources at best and are often frankly counterproductive.
Registration requirements were the first statutes aimed directly at sex offenders to be put into place. Such laws were based on the seemingly reasonable presumption that it can be useful for the police to know the whereabouts of offenders who have committed serious sexual crimes as they investigate newly reported incidents. The federal government stimulated the adoption of registration statutes in most states, with a series of laws, often named after child victims of sexual offenses, tying the development of sex offender registries to receipt of federal funds for law enforcement purposes (2). The Jacob Wetterling Crimes Against Children and Sexually Violent Offender Registration Act of 1994 required states to establish registries for sex offenders for at least ten years after their release. Most recently, the Adam Walsh Act of 2006 expanded the crimes covered by the registries, made failure to register a federal crime, and mandated state reporting to a national database by 2009. Many states have enacted even more stringent statutes, covering anyone convicted of a sexual offense and requiring lifetime registration.
It is estimated that more than 627,000 persons are included on sex offender registries, more than 111,000 in California alone (3). Maintaining databases of this scope is a time-consuming and expensive task for law enforcement authorities and depends heavily on the willingness of offenders in the community to comply with registration requirements. Several years ago, California authorities estimated that they had lost track of more than 33,000 offenders, and one survey of state officials suggested that states on average are missing data on 24% of potential registrants (4). No comprehensive study of the utility of such databases has been performed, but the fact that only 14% of sex offenses are committed by persons with prior sexual convictions indicates the realistic limits of their applicability (5).
Registration requirements would not be quite so onerous if every state did not pair registration with community notification. Following the lead of New Jersey, in 1996 Congress passed Megan's Law, requiring the release of sexual offender registration information to the public. In 2003 a more specific directive was enacted, mandating all states to place the information on a Web site available to the general public. Today, there is no state that does not maintain such a Web site with a search function allowing identification of offenders by name and geographic area (6). Most states supplement this on-line information with direct notification of neighbors of at least some classes of sex offenders. This can include posting flyers in the neighborhood, sending notices directly to area homes, having the police go door to door, putting notices in newspapers or on television, and holding community meetings (2).
Information disseminated typically includes an offender's name, address, criminal offense, and photograph. Given the consternation aroused by sex offenders, it can hardly be unexpected that the typical consequences of such disclosure are loss of housing, jobs, and friends. Yet these are just the kind of supports that can anchor a released offender in a community and reduce recidivism. Numerous reports have surfaced of offenders being threatened, harassed, and in rare cases killed after community notification (2). Suicide also has been reported (7). Perhaps most disturbing is the large number of states that fail to limit disclosures to predatory offenders, instead extending the process to everyone convicted of a sexually related offense. Swept up in this net are people who have committed noncontact crimes, such as exhibitionism or peeping, those whose only offense occurred as children, and persons who engaged in consensual sex with a somewhat younger girlfriend or boyfriend and were convicted of statutory rape (2).
Public notification requirements are flawed in another way as well. The assumptions underlying the statutes are that sex offenders are particularly likely to reoffend, which is why they are singled out from all other categories of criminals for registration and notification, and that informing the community will better enable potential victims to take precautions. According to the best available data, however, sex offenders are less likely to commit a sex crime in the future than almost all other categories of criminals are to recidivate—though the problem of accounting for unreported offenses always must be taken into account. A large meta-analysis showed sexual offense recidivism rates of 13.4%, ranging from 12.7% for child molesters to 18.8% for rapists, over an average follow-up period of four to five years (8). A major federal follow-up study of sex offenders in 15 states who were released in 1994 found reconviction rates for sexual offenses of 5.3% over the subsequent three years, with 40% of arrests coming in the first year (5). Moreover, most sex offenses are committed by family members and friends well known to victims, not by strangers about whom warnings might be helpful (4).
Legal challenges to registration and notification laws have been singularly unsuccessful. Not only have such statutes repeatedly been upheld by state courts, but in 2003 the U.S. Supreme Court also rejected a set of constitutional challenges. In Connecticut Department of Public Safety v. Doe, the Court turned aside an argument that registration and notification requirements violated procedural due process (9), and in Smith v. Doe, it rebuffed the litigants' contention that the Alaska statute violated the Constitution's ex post facto clause (10).
The most draconian approach to protecting the public from sex offenders is reflected in the growing number of statutes and local ordinances restricting where they can live (11). More than 20 states and hundreds of municipalities have such laws today (2), but there seems little question that Georgia's statute is the most extreme. The Georgia law prohibits released offenders from working within 1,000 feet of schools, churches, or daycare centers and living or loitering within 1,000 feet of places where children congregate, including all of the facilities in the work restrictions plus such other locations as playgrounds and school bus stops. The last of these provisions is the most onerous, because stops for school buses are ubiquitous in the state and bus routes change constantly. Thus full implementation of the law would render most of the residential areas in the state off limits to sex offenders, and almost every registered offender would be faced with the prospect of moving immediately (12).
A challenge to the Georgia statute led to one of the very rare instances in which a court has struck down a law regulating sex offenders. In late 2007 the Georgia Supreme Court considered Mann v. Georgia Department of Corrections, an action brought by a registered sex offender who after his release had gotten married, bought a home, and established a business (13). Not long thereafter, child care facilities located themselves within 1,000 feet of both his home and workplace, and under the new statute he was barred from both locations. Mann asked for a declaratory judgment that the statute would result in an unconstitutional taking of his property without due compensation. In its opinion, the court seemed particularly impressed that any third party could compel offenders to leave their homes simply by locating a child care center or church near their residences. Thus the court struck down the residence restrictions. However, the court was less persuaded about the hardship wrought by preventing Mann from working at the business he owned and upheld that part of the law (13). Other litigation is under way in the federal courts, and enforcement of the school bus stop provision is being held in abeyance (12).
Residence restrictions appear to be based on the unverified proposition that offenders are more likely to target victims, particularly children, in the immediate vicinities of their residences or workplaces. Even if that were true, the laws at best would be internally inconsistent, because they do not prevent sex offenders from living next to or even with children, though they cannot reside or work near schools and other facilities where children are likely to be more closely supervised. Surveys of sex offenders themselves indicate that they do not believe that residence restrictions will actually keep predators away from children if the predators are intent on reoffending (14).
Taken as a whole, current legislation aimed at people dwelling in the community who have committed sexual offenses seems largely counterproductive. By publicly labeling and shaming offenders, it all but ensures that they will have difficulty reintegrating into community life, supporting their families, and making friends. In banning them from living or working in large portions of many states, it further isolates them, inhibits their ability to put down roots and achieve stability, and often throws them together with other sex offenders in the few locations where residence is still possible (15). Through the indiscriminate process of including anyone ever convicted of a sexually related offense, many states blur the focus of police and the public on the most dangerous offenders. If restrictions were more narrowly targeted, they might have some role to play in assuaging public fears and preventing further crimes by the most serious sex offenders, but it is difficult to resist the conclusion that the current system is ill thought through and may ultimately be more harmful than no legislation at all.
However, apart from the Georgia Supreme Court decision in Mann, there are few indications that the tide is close to turning. States and localities continue to adopt residence restrictions, and even more aggressive initiatives surface regularly. Texas recently announced that it would screen all evacuees during hurricanes and other emergencies to prevent sex offenders from boarding buses with other citizens (16). In another move, New Jersey joined Florida and Nevada in barring sex offenders who used the Internet in their crimes from ever logging on again, except as part of work or looking for a job; offenders will have to install special equipment on their computers so that their usage can be monitored, another unwelcome task for law enforcement (17).
Reasonable and effective responses to the commission of sexual offenses require thoughtful consideration of the effects—both positive and negative—of restrictive laws and of their costs. Every dollar spent on tracking nonviolent offenders or enforcing restrictions on persons who are unlikely to commit new crimes is money taken from other law enforcement needs. At some point, policy makers need to gain perspective on these issues and develop strategies to calm public fears, rather than inflame them.