Offenders: Characteristics and Treatment

8 downloads 78924 Views 398KB Size Report
adult victims, child victims, or both. Child molesters are a specific and signifi- cant category of sex offenders. The term incest refers to sexual acts, which may or.
Offenders: Characteristics and Treatment Judith V. Becker Abstract Judith V. Becker, Ph.D., is professor of psychology and psychiatry at the University of Arizona.

People who sexually abuse children are diverse in terms of age, occupation, income level, marital status, and ethnic group. At one time it was believed that sex offenders could be easily categorized along three dimensions: offending against either adults or children; offending against either members of their families or against acquaintances and strangers; offending in noncontact ways (for example, exhibitionism) or through bodily contact. There is growing evidence that a number of offenders offend across these categories. Many child sexual abusers are themselves adolescents, and many adult offenders first offended when they were adolescents. A review of the literature reveals a paucity of controlled therapy outcome studies on the effectiveness of treatment, and existing uncontrolled studies are marked by methodological problems. This has led some to conclude that there is insufficient evidence to prove the effectiveness of treatment for child molesters. However, there have been major changes in treatment programs over the years, and some more recent studies provide reason for optimism about the effectiveness of current treatment methods for some offenders.

A

s other articles in this journal issue have noted, far too many children in our society fall prey to individuals who use them to fulfill their own sexual needs. The offenders can be relatives of the child victim, or may be acquaintances or strangers. Their activities may come to the attention of the child protection system (if the child’s parents are involved) or of the criminal justice system, or they may escape detection. Offenders may be adults or may themselves be minors.

This article uses the term child molester to refer to those who choose only or primarily child victims, and sex offender to refer broadly to those who offend against adult victims, child victims, or both. Child molesters are a specific and significant category of sex offenders. The term incest refers to sexual acts, which may or may not include intercourse, between members of a family other than a husband and wife. Incest may occur between The Future of Children

two adults, but all references to incest in this article refer to sexual acts involving a child. The population of incarcerated sex offenders has grown rapidly. Between 1988 and 1990, the population of sex offenders in American prisons grew by 48%.1 In some states, as much as one-third of the prison population is made up of sexual offenders.2 It is unknown to what extent

SEXUAL ABUSE OF CHILDREN

Vol. 4 • No. 2 – Summer/Fall 1994

177

this is the result of longer sentences for incarcerated offenders. This article reviews what is known about offenders and treatment of them. First, the role of paraphilia (consistent deviant sexual interests) in child molestation is discussed. Second, the article looks at what is known about offenders, including data on juvenile offenders and incest offenders. Third, the article discusses recidivism and the difficulty of determining recidivism rates, with a summary of what is known about recidivism of untreated offenders. Fourth, the article looks at treatment, including mechanisms for getting offenders into treatment, treatment providers, goals and types of treatment, the efficacy of treatment, and the need for postincarceration monitoring and longterm treatment. Finally, recommendations are made.

The Role of Paraphilia in Child Molestation The universe of child molesters is broad. A distinction could be made between (a) those individuals with a normal pattern of sexuality who may impulsively or opportunistically perform a single deviant act, and (b) those individuals who have a consistent deviant sexual interest, known as a paraphilism. There is some link between the two categories: opportunistic deviance may develop into full-blown paraphilia over time, as described below. For a person to be diagnosed as having a paraphilia, it is required that a sexual fantasy or pattern of behavior involving a nonconsenting person, an animal, or nonhuman objects have existed for at least six months, that the sexually arousing fantasies or urges are recurrent and intense, and that the person either acted on the fantasies or suffered serious distress because of them.3 An individual who occasionally has paraphilic fantasies but is not markedly distressed by them and does not act on them is not diagnosed as suffering from a paraphilia. In this author’s clinical experience, most,

but not all, of the offenders who come to the attention of the criminal justice system or child protective system have one or more paraphilia. In some cases, isolated acts may occur without paraphilia, usually in the context of alcohol or substance abuse, or during stressful situations such as psychiatric illness or marital problems. Paraphilic fantasies or actions involving prepubescent children as sexual partners are termed pedophilia. 3 Pedophiles may be attracted to youth of the same sex, opposite sex, or both. While both adult males and females have been diagnosed with pedophilia, clearly the majority of cases involve adult men molesting female children. (See the article by Finkelhor in this journal issue.) Although at one time it was assumed that most offenders could be simply classified as suffering from a single paraphilia, a growing body of evidence indicates that many or most offenders have more than 4 one category of deviant sexual behavior. For example, it used to be assumed that incest offenders could be clearly separated from other child molesters, but current evidence indicates that a substantial percentage of child molesters offend in both spheres.4 There is also some evidence that many exhibitionists engage in additional deviant sexual acts, rather than exhibitionism being an isolated category of behavior, as was once assumed.5,6 As will be discussed below, a substantial percentage (one study suggests the number may be as high as 30% to 50%)7 of child molestations are perpetrated by juveniles, and a substantial percentage of adult offenders may begin their molesting behavior as juveniles.8 While juvenile offenders may resist treatment as stubbornly as adult offenders, their younger age means that a higher percentage of juvenile offenders will not yet have developed full blown paraphilias and entrenched patterns of behavior, and may therefore be more amenable to treatment.

THE FUTURE OF CHILDREN – SUMMER/FALL 1994

178

Offenders Until recently, most of what was known about perpetrators of child sexual abuse was based on incarcerated offenders. While studies on incarcerated offenders have added to our knowledge about this population, these studies do not necessarily represent accurately the population of nonincarcerated offenders, which may be substantially larger. Also, incarcerated sex offenders are likely to conceal the exact nature and scope of their offenses because of fear of further prosecution or extended incarceration. This section discusses a major study of nonincarcerated offenders, and summarizes information from other studies, with discussion of knowledge specific to juvenile offenders and incest offenders. A typology defining a range of distinct types of nonincestuous offenders is described, and current knowledge of the etiology of child molestation is summarized. Abel Study of Nonincarcerated Offenders From 1977 to 1985, Abel and colleagues4 interviewed a large sample (n = 561) of nonincarcerated male sex offenders under a federal certificate of confidentiality, which protected information provided to the researchers by those who participated in the study.9 The sample was made up of 270 child molesters, as well as 152 sex offenders who offended against both children and adults, an additional 92 who sexually offended exclusively against adults, and 47 other paraphiliacs who participated in sexual behavior that was not classifiable by victim age (for example, fetishism).

Child molesters may engage in incestuous, as well as nonincestuous, abuse and may target children of both genders. The subjects were recruited through informal discussions with health care professionals, formal presentations at mental health, parole, probation, forensic, and criminal justice meetings, and through ads in the local media. Although subjects were seen at two sites (Memphis, Tennessee, and New York City), some of them resided in other areas of the country. They ranged

in age from 13 to 79. The average age of these sex offenders was 31.5 years. Eighty percent were between the ages of 20 and 49 years of age, and 7% between the age of 50 and 79. Because this was a study of adult male offenders, juvenile offenders were underrepresented and female offenders were not represented at all. These were people from every walk of life. Forty-seven percent of the offenders were and always had been single. Twentynine percent were married, and the remainder were formerly married or had, at some point, formed a significant “living with” relationship of some type with an adult partner. All educational levels were represented, with 40% having at least one year of college. The majority were employed, with only 20% being either temporarily unemployed or unemployed longer than one month. All subjects volunteered to participate in the study, seeking assessment and/or treatment. Consequently, this pool of subjects would have to be characterized as self-selected. Although the subjects included representatives of each of the categories of paraphiliacs, we cannot assume that this group represents the normal distribution of categories of paraphiliacs in the general population. (However, it should be remembered that any group of sex offenders being studied will show some selection bias, since they will either be self-selected or will be in the study by virtue of having come to the attention of the legal system.) A major concern was that offenders were expected to underreport their sex crimes. Therefore, each subject first watched a one-hour videotape describing the study and the protection afforded by the certificate of confidentiality. Each subject then underwent a structured clinical interview, lasting from one to five hours, depending upon the subject’s ability to recall and describe his deviant history and the variety and complexity of his paraphiliac interests. In a few instances where the subject was suspected of overreporting his behavior, the interview was repeated until data were consistent. If repeated interviews failed to yield consistency, the data were not included in the study. The Abel study yielded some important information. Child molesters may engage in incestuous, as well as nonincestuous, abuse and may target children of both genders. Of the total population in-

Offenders: Characteristics and Treatment terviewed in the Abel study, 67% targeted only females, 12% targeted only males, and 21% targeted both. While 56% engaged only in nonincestuous behavior (molesting or assaulting nonrelatives) and 12% engaged only in incestuous behavior, 23% engaged in both incestuous and nonincestuous behavior. Again, 47 subjects (8%) engaged in activities not classifiable in this subcategory (for example, fetishism). Data from Other Studies of Offenders In addition to the Abel study, a number of other studies provide information about characteristics of offenders. For example, one of the most publicized characteristics of sex offenders is a past history as a victim of abuse, and indeed both physical and sexual abuse histories have been noted in many studies of offenders. Johnson and 10 Shrier reported that 66% of a sample of male juvenile sex offenders had been victimized either physically (19%) or sexually (49%). In this author’s clinical practice, which includes primarily less serious, nonincarcerated juvenile offenders, about 19% report a history as sexual abuse victims.11 Longo reported that 47% of the adolescent sex offenders in his treatment program had been sexually abused.12 Some have assumed that child molesters are typically passive, unassertive individuals who interact with children by mutual consent. However, data indicate that child molesters are frequently aggressive. Of 250 child victims studied by DeFrancis,13 50% experienced physical force, such as being held down, struck, or shaken violently. In an unpublished study of 38 child molesters, using selfreports of offenders and interviews with medical examiners, Christie, Marshall, and Lanthier14 found that 59% of the child molesters used force, and in 42% of cases the force was sufficient for the medical examiners to find “noticeable injury” to the victim. Characteristics of Juvenile Offenders Until somewhat recently, the major focus of research and treatment has been on adult child molesters. However, the importance of the juvenile offender is becoming increasingly clear. In a research project on 411 adult offenders seen voluntarily at an outpatient clinic, 58% of the adult offenders reported the onset of deviant sexual interests in adolescence.15 Groth reported that 60% to 80% of adult offenders admitted that they had begun their deviant sexual behaviors as adoles-

179 cents.16 In a study of 54 incarcerated child molesters in Florida and Connecticut, inquiries about the offender’s age at first offense produced a bimodal curve, with one group whose first child molestation offense occurred between the ages of 13 and 16, and a second group whose first offense occurred between the ages of 31 and 35.16 In a national sample of 863 adolescent males, the rate of sexual assaults committed per 100,000 adolescent males ranged from 5,000 to 16,000,17 depending upon the definition of sexual assault and whether arrest rates or self-report data were used. In a 1984 study of 401 child sex abuse cases, 56% of the male child victims and 28% of the female child victims reported being abused by a juvenile offender.7 Two other 1980s studies had similar results, with 56% of male child sexual abuse victims reporting a juvenile offender18 and 15% to 25% of female child victims reporting a juvenile offender.19

Juveniles may offend against their peers, against younger children, or against adults. Many of their victims may be children simply because the juvenile offender lacks the means to offend against an adult. Juveniles may offend against their peers, against younger children, or against adults. Many of their victims may be children simply because the juvenile offender lacks the means to offend against an adult or has more opportunity to offend against children. Gail Ryan at the C. Henry Kempe Center for Prevention and Treatment of Child Abuse and Neglect in Denver, Colorado, has established a national data base on 1,600 youth who have been referred to 90 specialized treatment programs.20 That data base indicates that juvenile sexual offenders range in age from 5 to 19 with a median age between 14 and 15. Juvenile sex offenders are representative of all ethnic, racial, and socioeconomic classes. Ninety percent of the juvenile offenders were male, and more than 60% of the sexual offenses involved penetration. Over 90% of the juvenile offenders se-

180

THE FUTURE OF CHILDREN – SUMMER/FALL 1994 lected as a victim a person known to them (for example, a relative or an acquaintance).20 The most common scenario involves a 7- or 8-year-old victim.20

Extensive information on juvenile offenders is included in a 1993 book, The Juvenile Sex Offender, edited by Barbaree, Marshall, and Hudson.32

Individual characteristics prevalent among juvenile sex offenders include a lack of assertive and social skills.21-23 Low academic performance,22 learning problems, and learning disabilities have also been diagnosed in juvenile sex offenders. 24,25 in a study of Smith and colleagues, 262 male adolescent offenders, found lack of impulse control to be present in many of their patients. Fehrenbach and col26 leagues have reported that, in a sample of 293 male adolescent offenders, 44% committed nonsexual offenses prior to their first sexual offense.

Characteristics of Incest Offenders

Depression may also be a major characteristic in juvenile sex offenders. In a study of 246 adolescent sex offenders, 27 Becker and colleagues found that 42% of their clients experienced major symp-

Unfortunately, controlled studies that measure the effectiveness of different types of treatment against different types of offenders are lacking.

tomatology as measured by the Beck Depression Inventory. The mean score of the sex offenders on this inventory was two times higher than a random sample of adolescents as reported by Kaplan and colleagues.28 Most of the above-mentioned characteristics are not unique to juvenile sex offenders. In two studies that compared juvenile sex offenders to other violent and nonviolent juvenile offender populations, no significant differences were found regarding neuropsychological, intellectual, or psychological capacities.29,30 Several characteristics of family environment have also been described in the research literature on juvenile offenders. These include unstable home environments, a sexual pathology within a parent, and the child viewing sexual interactions between parents or parent surrogates.24,25 Viewing or experiencing family violence has also been suggested. 31 Parental loss or separation may also play a role for incest offenders.

Traditionally, it was thought that incest offenders were exclusively incestuous, and that their recidivism risk was extremely low. With newer and more complete information, these assumptions are being dropped. Of the 159 incest offenders against female children in the Abel study, 49% had histories of being involved in nonincest female pedophilia, 12% in male pedophilia, 19% in rape, 20% in exhibitionism, 7% in voyeurism, 6% in frottage (public sexual rubbing against nonconsenting persons), and 6% in sadism.4 Recidivism among incestuous molesters, especially among those whose sexual offenses are exclusively incestuous, is still generally lower than for nonincestuous offenders (see discussion of recidivism, later in this article). However, freedom from recidivism should never be assumed, and family reunification in incest cases should be approached with cau33 tion. Recidivism may not occur until many years later, especially where the offender clearly has a paraphilic sexual interest in children but has better impulse control than most sex offenders. In this author’s clinical experience, such offenders may go years without offending, but may reoffend when placed in close, frequent contact with children. Some have put forth the theory that incestuous men tend to relate sexually to children in response to some family disruption.34 However, the majority (59%) of the subjects in the Abel study had the onset of their deviant sexual interest pattern during adolescence, which indicates a disposition to offend, predating and independent of any family disruption. A more plausible explanation may be related to the presence of a child who resides in the house of the age and sex to which the perpetrator is attracted. Typology of Child Molesters The population of child molesters is heterogeneous as to the offenders’ motivations and the type of their sexual behaviors with children. Efforts are ongoing to differentiate types of molesters in order to provide guidance for improved understanding of the etiology of sexual offend-

181

Offenders: Characteristics and Treatment ing and to assign offenders to appropriate treatment. The first step is to develop and validate a typology of sexual offenders. Knight, Carter, and Prentky35 developed and tested a typology using the cases of 177 nonincestuous child molesters who had been civilly committed to the Massachusetts Treatment Center for Sexually Dangerous Persons in Bridgewater, Massachusetts, between 1959 and 1981. One goal was to create a typology which was consistently valid, so that individual offenders would be assigned to the same type by different reviewers working separately, and so that individual offenders would be consistently assigned to the same type over time. Another goal of the typology was to separate the mass of offenders into distinct categories, based upon their behavior and thought patterns and upon the differing developmental antecedents leading to their paraphilia. This typology is based upon a selected population of child molesters who had been apprehended and convicted. Offenders were excluded from the typology if they had no physical contact with their victims (for example, if they were exhibitionists). Also excluded from the typology were incest-only offenders and offenders who chose as victims both adults and children. Knight found that nonincestuous child molesters could be classified by both the offender’s degree of fixation on children and the offender’s behavior during the molestation. The behavior history allowed Knight to break offenders into six types representing a range of behavior. At one extreme are interpersonal offenders who have frequent nonsexual contact with children and have shown a sustained interest in an individual child in a relationship that extends beyond sexual involvement, where sexual contact is typically nongenital and nonorgasmic. At the opposite extreme are sadistic offenders who rarely have nonsexual contact with children and who are more likely to choose strangers as victims. Sadistic offenders by definition have a history of violent acts that are sexually arousing to the offender and that produce physical injury to the victim. Knight and colleagues have recently published updated data on the validity of the typology.36 Development and validation of a typology are important steps in differentiating offenders. Unfortunately, controlled studies that measure the effectiveness of different types of treatment against different

types of offenders are lacking. Even so, many clinical programs make use of Knight’s typology, often in a modified form, in assessing offenders and assigning them to treatment. Essential next steps for the field are to (1) validate a typology

In all likelihood, there is not one causative factor, but rather multiple pathways by which a person develops a sexual attraction to minors. that applies to incest-only offenders, female offenders, offenders who chose both adults and children as victims, and noncontact offenders, (2) conduct controlled studies that assess which types of treatment are most effective with different types of offenders, and (3) use current knowledge of typology in developing an improved understanding of the etiology of child molestation. Etiology While a number of theories have been proposed in an attempt to explain why some individuals are attracted to minors, we are lacking a theory that has been empirically derived and universally agreed upon, and that addresses the heterogeneity of the offender universe. In all likelihood, there is not one causative factor, but rather multiple pathways by which a person develops a sexual attraction to minors.37 Biological factors have been postulated as perhaps being causative in the development of deviant sexual interest patterns. It has also been suggested that abnormal levels of androgens may contribute to inappropriate sexual behavior. It is important to note, however, that the majority of studies of androgen treatment have dealt only with violent sex offenders and have yielded inconclusive results.38 Psychoanalytic and psychodynamic theories relating the cause of deviant sexual behavior to conflicts or trauma experienced in early childhood have been of limited utility in fostering effective treatment for offenders (see the discussion of treatment options that appears below). Learning theorists propose that sexual arousal develops when an individual engages in a sexual behavior that is sub-

182

THE FUTURE OF CHILDREN – SUMMER/FALL 1994 sequently reinforced through sexual fantasies and masturbation.39,40 It is thought that there are certain vulnerable periods (such as puberty) when this can occur. For example, if an adolescent engages in sexual activity with a small child and there are no negative consequences to the adoles-

An additional problem with conviction records as a measure of recidivism is that many sexual offenses are plea bargained down to nonsexual offenses. cent, the adolescent may fantasize about having sex with young children and masturbate to those fantasies, thereby developing arousal to younger children. Given our largely invalidated and limited understanding of the etiology of sexual abuse, some observers recommend that analysis and research focus on the conditions that must exist in order for individuals to move beyond fantasy to carry out their sexual impulses toward 41 children. Finkelhor theorizes that four factors are necessary before pedophilic fantasies will turn into action: (1) the adult finds it emotionally satisfying to relate to children; (2) the adult experiences a physiological response to a child, for example, erection; (3) the adult is blocked in his or her ability to get needs met by an adult; and (4) the individual may have poor impulse control or may utilize substances such as alcohol or drugs that lower inhibitions. It is important to note, however, that alcohol, in and of itself, does not cause a person to commit a sexual offense. In some instances, individuals consume alcohol or drugs to give them the “courage” to do what they wanted to do in the first place. While there are numerous theories that address the etiology of sexual offending behavior, what is needed is a comprehensive, integrated theory, that is empirically derived. Marshall and Barbaree42 describe such an integrated theory, which brings together psychological, biological, and sociological factors. Currently, a review of the literature reveals only two models to describe the etiology of such behavior in juveniles. Ryan, Lane, Davis, and Isaac43 identify a

“sexual assault cycle.” This model incorporates six steps, beginning with a juvenile having a negative self-image with an increased probability of maladaptive coping strategies; this leads to the second step when the youth anticipates negative reaction from others. To protect against this rejection, the youth isolates and withdraws. To compensate for feelings of powerlessness, the youth fantasizes situations in which he or she is in control and acts on those fantasies (victimizes someone). This leads to more negative self-imaging, and the vicious cycle is repeated. Becker and Kaplan’s44 model proposes that sexual offending and abusive behavior by juveniles result from a combination of individual characteristics (for example, poor impulse control, history of victimization, lack of social interactional skills); family variables (including poor parent-child bonding, domestic violence, child maltreatment, lack of boundaries); and socioeconomic factors (cultural acceptance of violence, violent role models, objectification of people). While both of these models of the etiology of sex offending in juveniles make intuitive sense, neither has been empirically validated. Screening for Pedophilia in Volunteer Organizations Current knowledge about offenders is not sufficient to be used to assess the likely guilt or innocence of an accused person. However, efforts are ongoing to develop nonintrusive testing instruments that could aid youth-serving volunteer organizations in screening volunteers to avoid giving pedophiles unsupervised access to children.45

Recidivism Among Untreated Offenders Recidivism (repeated offenses after conviction) for most serious crimes is high. According to a U.S. Department of Justice report,46 of the more than 100,000 persons released from prisons in 11 states in 1983, an estimated 62.5% were rearrested for a felony or serious misdemeanor within three years. A RAND report which looked at prison populations generally concluded that “released inmates, as a group, pose a very serious threat to public safety, but we cannot predict with useful accuracy which inmates will recidivate.”47 In California, the state Department of Justice concluded

183

Offenders: Characteristics and Treatment that “sex offenders do not differ significantly in terms of overall recidivism from most other types of offenders.”48 In understanding recidivism among child molesters, an important question is what level of recidivism can be expected in the absence of treatment, and whether any characteristics of offenders have been identified that predict an above- or below-average risk of recidivism. To address these topics, we must first consider the problem of measuring recidivism. Measuring Recidivism The key question in measuring recidivism is whether the once-convicted offender commits a second sexual offense when returned to the community. One source of such information would be official arrest or conviction records, although these would be imperfect, given that so many offenses will never come to the attention of the system.49 But official statistics are extremely limited. In discussing the shortcomings of official data sources, Furby and colleagues50 note the following: “The actual number of sex offenses is grossly underestimated by official reports. Probably less than 10% of rape assaults are ever reported to the police (citation omitted). Furthermore, of those sex offenses reported to law enforcement authorities, some are not recorded, and records of others are lost. There is little indication of how extensive this problem is at the site of original recording. However, users of Uniform Crime Rate statistics or Federal Bureau of Investigation (FBI) rap sheets should be aware that some local law enforcement agencies do not provide the FBI with arrest reports. Conviction rates for sex offenses are even lower than arrest rates and dramatically so (e.g. of 315 rapes reported to police in Seattle, Washington in 1974, only 15 resulted in conviction [citation omitted]). Furthermore, data on dispositions are estimated at only 50% complete.” An additional problem with conviction records as a measure of recidivism is that many sexual offenses are plea bargained down to nonsexual offenses.50 To more accurately reflect recidivism, some researchers include arrests or convictions for violent nonsexual offenses. Furby also notes that “one other measure of recidivism that has been used by some (e.g. United States Department of Justice, 1985) is the percentage of sex offenders admitted to prison in a given year

who are recidivists. . . . [This information] is conceptually flawed as a recidivism measure: It tells us what percentage of a group of offenders has committed previous offenses, which is not necessarily the same as the percentage of a group of offenders who will go (or went) on to reoffend.”50 Given these serious limitations on official crime statistics, most studies of recidivism rely on data from multiple sources, which generally start with arrest and conviction records, and are supplemented with interviews with the offender, his parole officer, local police departments, and/or the offender’s family. 51 Marques and colleagues found that simply reviewing parole office records, in addition to law enforcement records, produced a 33% increase in their estimates of serious crimes committed by released sex offenders. Recidivism rates are most meaningful if they cover at least a five-year period, postincarceration. Twenty-seven percent of recidivists in one study did not begin to recidivate until four years or more after their release.52 However, it can take several years to identify a treatment popula-

Given these serious limitations on official crime statistics, most studies of recidivism rely on data from multiple sources. tion and control group of sufficient size to study, which must then be followed for several more years, resulting in studies often taking ten years or more to complete. Studies of Recidivism Among Untreated Offenders Prentky and Burgess,53 in the context of an article comparing the costs of treating child molesters versus not providing treatment, stated that determining a recidivism rate for untreated molesters was “the weakest component” of their economic analysis. Nonetheless, they concluded that the most defensible figure from recent research came from a control group of 53 untreated molesters in a 1988 study of treatment impact by Marshall and Barbaree.54 This control group showed a recidivism rate of 32% over four years. If incest offenders were dropped from this group, the recidivism rate rose to 40%.

184

THE FUTURE OF CHILDREN – SUMMER/FALL 1994 Because it is extremely difficult to measure recidivism and numerous offenders may go several years or longer before recidivating, no statistics on recidivism for untreated offenders should be taken as definitive. However, as discussed below, there is support from other sources for the reasonableness of Marshall and Barbaree’s 40% recidivism rate for untreated nonincest offenders.

A 1988 study showed a recidivism rate of 32% over four years. If incest offenders were dropped from this group, the recidivism rate rose to 40%. Studies of untreated molesters alone have primarily taken place in Europe. Christiansen and colleagues,55 in a 1965 Scandinavian study of 2,934 sex offenders followed for 12 to 14 years, found 24% were sentenced for a new offense during the observation period. Presumably a higher percentage committed offenses for which they were not sentenced. In England in 1978, Soothill and Gibbens56 estimated that 48% of their sample of sex offenders would be reconvicted by the end of 22 years at risk. In Canada, Hanson and colleagues57 used two control groups of untreated convicted child molesters in a treatment effectiveness study. Group 1, made up of 31 untreated offenders who were released from prison between 1958 and 1965, had a 48% rate of reconvictions for sexual or violent offenses. Group 2, made up of 60 untreated offenders released between 1965 and 1973, had a 33% reconviction rate. The length of the follow-up period varied according to when offenders were released, but 93% were followed for 15 years or more. The authors noted that the greatest risk period appeared to be the first 5 to 10 years after release from incarceration, but that 23% of the recidivists in their sample were reconvicted more than 10 years after they were released. Factors Other Than Treatment That Affect Recidivism McGrath58 in a 1991 literature review of recidivism risks for sex offenders (looking broadly at rapists, child molesters, and ex-

hibitionists) analyzed 18 variables that predict recidivism to varying degrees. He found that incest offenders have the lowest untreated recidivism rates, often less than 10%.59 Extrafamilial offenders who molest boys generally have higher reoffense rates (13% to 40%) than those who molest girls (10% to 29%).58 Multiple paraphilias can also be a powerful indicator.60 In a study of 98 treated child molesters who were followed for one year,61 those offenders who targeted both males and females and both pre- and postpubertal victims showed a 75% recidivism rate (n = 9). This variable alone correctly classified 84% of both recidivists and nonrecidivists in the study. A prior criminal record is also a strong predictor of recidivism.62 Romero and Williams,52 in a 10-year follow-up study of 231 sex offenders, found the offenders’ number of prior sex offense arrests to be the single best predictor of recidivism. Twentyseven percent of the recidivists in their study did not begin to recidivate until the fourth year after their release or later. Other factors identified by McGrath as increasing the risk of recidivism among child molesters are sexual arousal patterns, alcohol abuse, use of force, absence of social supports, and employment status.

Treatment for Offenders This section discusses mechanisms for getting offenders into treatment, who provides treatment, common goals and types of treatment, and what is known about the efficacy of treatment. Briefly, there are serious problems with the application of scientific standards of research to treatment of sex offenders. For example, many in the field (including this author) feel it is ethically questionable to randomly assign sex offenders who are being treated in the community to treatment or to nontreatment control groups. These and other methodological problems are discussed later in this section. A review of the literature reveals a paucity of controlled therapy outcome studies on the effectiveness of treatment. Existing uncontrolled studies are marked by methodological problems. This has led some to conclude there is insufficient evidence to prove the effectiveness of treatment for child molesters.50 The most widely cited analysis concluding that treatment has not been proven effective has recently been challenged63 for focusing on outdated programs. There have been major

185

Offenders: Characteristics and Treatment changes in treatment programs over the years, and more recent studies51,64 provide reason for optimism about the effectiveness of current treatment methods for some offenders. Many sexual offenders, whether or not incarcerated, receive no treatment. Where in-depth treatment is available to incarcerated offenders, it is typically available to only a very small portion of the incarcerated offender population.65-67 Any person who has a sexual interest in children should receive professional help in gaining control of this behavior. This author believes strongly that modem treatments, which have changed and improved significantly in the past 15 years, have a significant impact on the recidivism rates for many offenders. However, because of the need to follow treated offenders for several years in order to establish recidivism rates, it will be some time before this perception can be reliably demonstrated. Mechanisms for Getting Child Molesters into Treatment The discussion below describes the most common mechanisms by which child molesters enter treatment. Unfortunately, no data are available on the frequency of use of these mechanisms. A small number of offenders voluntarily seek treatment. Among offenders who have not yet had contact with the legal system, the majority are probably not motivated to seek treatment, and reporting laws that require therapists to report these individuals to authorities probably deter most of those who might consider seeking treatment voluntarily. Some patients enter therapy when informal agreements are made between the offender, the victim’s family, and/or the offender’s employer (for example, when the employer is a church). However, this means of getting molesters into treatment is probably shrinking in significance, as victims are increasingly likely to turn to the legal system. If an offender is arrested, he may be directed to a pretrial diversion program that includes treatment. Nonparticipation in treatment would result in criminal charges being filed against the offender, but participation in treatment would allow the offender to avoid having a criminal record. In other cases, the offender may be found guilty at trial or through a plea

bargain, and be sentenced to probation, with treatment required as a condition of probation. (This is a common outcome. In a 1987 survey of counties in three states, 80% of convicted child molesters were sentenced to probation, with no prison time served.68) If treatment is a condition of probation, nonparticipation in treatment could result in incarceration. Some offenders who are found guilty at criminal trial or plead guilty are incarcerated where they may receive treatment in prison. However, this is likely only in special facilities dedicated to sex offenders. Upon release from incarceration, some sex offenders are required to participate in treatment as a condition of parole. Noncompliance is a violation of parole and can result in reincarceration. Finally, some offenders may be civilly committed to residential mental health programs if judged to be a danger to the community. However, standards for such involuntary commitment are strict. Between 1937 and 1972, 25 states and the District of Columbia passed sexual psychopath laws, most of which were sub-

Where in-depth treatment is available to incarcerated offenders, it is typically available to only a very small portion of the incarcerated offender population.

sequently repealed or modified when they were found to serve only the unconstitutional purpose of preventive detention.69 The state of Washington recently passed a law specifically allowing the civil commitment of “sexual predators,”70 which has been upheld as constitutional by the Washington Supreme Court.71 Even if commitment procedures meet constitutional standards, it is a problem in some jurisdictions that standards for release from civil commitment are vague and subjective. Some incest offenders whose cases do not enter the criminal system may be required by a family court judge to take part in treatment as a condition of family reunification.

186

THE FUTURE OF CHILDREN – SUMMER/FALL 1994 Treatment Providers One thousand five hundred programs that provide sex-offender assessment and treatment programs responded to a 1992 survey for the Safer Society, a private, nonprofit national resource and referral center.72 This is an increase of 133% from the 643 providers identified by the Safer Society in 1986. The 1992 list is a fairly complete national listing; only 194 programs nationally were identified but failed to respond to the survey. The survey makes no distinction between programs for child molesters and programs for sex offenders who offend against adult victims.

percent of programs serving juvenile sex offenders include treatment to preadolescent children who molest. Sixty-four percent of juvenile treatment facilities offer specialized treatment for juvenile female sex offenders; but according to the Safer Society, the demand for services for this population is escalating dramatically, and the availability of treatment is not keeping pace with the need. Similarly, services to lower-functioning/developmentally disabled sex offenders are increasing but are still inadequate. Forty-nine percent of juvenile providers and 51% of adult providers offer services to the developmentally disabled.

Of the 1,500 identified programs, about half (755) serve juvenile offenders and half (745) serve adult offenders. The 1,500 listed programs include only professional mental health providers specializing in treating sex offenders; other potential resources not included are self-help programs for offenders, nonspecialized treatment programs that include sex offenders, and the several hundred groups nationally that serve only incest offenders in the context of the family.

Counselors employed in sex offender treatment programs may come from a variety of disciplines. The Society’s 1994 survey, due later this year, will include information on the academic qualifications and training of staff members in the surveyed programs. Currently only the state of Washington73 requires special certification for counselors treating sex offenders, although other states, including Texas and Wisconsin, are considering such a requirement.74

About 25% of sex offender treatment programs are in residential settings, which includes both programs for incarcerated offenders and residential mental health facilities. The other 75% are communitybased outpatient programs. The private sector accounts for the majority of all outpatient services for juveniles (60%) and adults (66%), but provides a smaller share

Goals of Treatment

Currently only the state of Washington requires special certification for counselors treating sex offenders, although other states, including Texas and Wisconsin, are considering such a requirement.

of all residential services for juveniles (45%) and adults (18%). Services are unevenly distributed geographically. The Pacific states offer the greatest number of services, and Texas alone accounts for 9% of the total (131 programs). Services to specialized populations are expanding, but are still insufficient relative to the number of offenders. Sixty-five

Successful treatment is most frequently defined as a lack of recidivism. Other treatment goals include decreasing deviant arousal, increasing nondeviant arousal, developing functional interactional skills, anger management, stress control, and eliminating alcohol and substance abuse if they exist. Therapy might also target empathy training, sexual values clarification, and cognitive distortions. There is controversy over whether family reunification should normally be a goal in incest cases.75,76 Clinical interviews, psychological testing, and paper-and-pencil tests are often used within treatment programs to assess whether treatment goals have been attained. However, these measures have serious limitations because they may not predict actual behavior. Measures such as the penile plethysmograph (a physiological assessment technique used to assess erectile responses in males) may be used to determine if the offender has experienced a decrease in deviant arousal and/or an increase in nondeviant arousal. Plethysmograph results are often used in therapy to confront the offender who denies or minimizes his deviant sexual interests. Although penile plethysmography is

187

Offenders: Characteristics and Treatment the most valid measure of sexual arousal and has been validated against an adult non-sex-offender population,77 both false positives and false negatives occur, and it cannot be used either as the sole predictor of recidivism or to determine if a person has committed a sex offense. The Association for the Treatment of Sexual Abusers has formulated a series of guidelines for therapists in regulating the ethical use of the plethysmograph.78 Amenability to Treatment McGrath’s 1991 literature review,58 discussed earlier, noted that a sex offender can be considered amenable to treatment only if he acknowledges that he has committed a sexual offense, he considers his sexual offending a problem behavior that he wants to stop, and he is willing to participate fully in treatment. These requirements alone indicate that many known offenders are not good candidates for treatment, though McGrath notes that “amenability to treatment is not a static variable” because offenders who resist treatment at one point in time may be amenable to treatment at a later date. Types of Treatment As discussed earlier in this paper, the offender population is extremely diverse in demographics, paraphilic interests, and behaviors. A great deal remains to be done in developing and validating typologies that cover the full range of offenders and in understanding which types of treatment are most effective with which types of offenders. In general, treatment can be grouped into five categories: biological therapies, traditional psychoanalysis and psychodynamic therapies, family therapy, behavioral therapies, and relapse prevention programs. Any of these models may be used either for adult or juvenile offenders, and some of these treatments may be used in combination. Biological Therapies

Biological therapies have been utilized for some pedophiles with long-established offense histories. At one time, surgical castration was widely used in Europe. Heim and Hursch provide an extensive review and critique of the literature on the surgical castration of sex offenders.79 Although low recidivism rates have been reported, these authors conclude that there is no scientific (based on methodological problems with studies conducted) or ethical

basis for the use of surgical castration in the treatment of sex offenders. In view of the important role androgens play in the maintenance of sexual arousal, treatments have predominantly

A sex offender can be considered amenable to treatment only if he acknowledges that he has committed a sexual offense, he considers his sexual offending a problem behavior that he wants to stop, and he is willing to participate fully in treatment. focused on blocking or decreasing the level of circulating androgens.38 Antiandrogenic medications do not appear to influence the direction of the sexual drive; however, they do act to decrease the sexual drive in general. While these medications have proven effective with some offenders, some antiandrogens have potential side effects, such as hypertension, weight gain, and mild lethargy, which at times provoke difficulty in compliance.80 Recent work81 has suggested the use of antidepressants with some sex offenders, where sex offending behavior is combined with poor impulse control or compulsive behavior. There is no expectation that this medical treatment alone will eliminate sexually offending behavior, but there are indications it may aid in relapse prevention. Psychoanalysis and Psychodynamic Therapies

Psychoanalysis and psychodynamic therapies have been used in treating sexual offenders. Often involving individual counseling, these therapies focus on identifying and resolving early life conflicts and traumas. Evaluating results of these interventions has been somewhat complicated because there are no common standards of measurement. Furthermore, several researchers have reported disappointing results with this form of intervention with sexual offenders.82 According to the Safer Society’s 1992 survey,72 98% of juvenile and adult sex offender treatment programs prefer to offer counseling in peer groups (see discussion of cognitive-behavioral therapies, below). Only 2% of juvenile and 2% of adult programs indicate that they use in-

188

THE FUTURE OF CHILDREN – SUMMER/FALL 1994 dividual treatment by choice, rather than because of inadequate numbers of clients by gender or age. Family Therapy

A community-based treatment program for families in which incest has occurred has been described by Giarretto, Giarretto, and Sgroi.83 This treatment program combines individual therapy with group therapy in self-help groups. While Giarretto reported no recidivism in more than 600 families receiving treatment,84 unfortunately no data regarding length of followup have been reported. Giarretto noted that the authority of the criminal justice system was essential in treating incest cases; program dropouts were largely those men who were not under criminal justice system supervision.85 Cognitive-Behavioral Therapies

Perhaps the most widely available and most widely researched forms of therapy for sex offenders are the cognitivebehavioral therapies. The goal of these therapies is to teach individuals how to recognize and change their inaccurate beliefs (for example, that a victim enjoys being victimized) and to teach specific means to control their inappropriate impulses and behaviors. These treatments are usually multicomponent and utilize various behavioral techniques, especially altering maladaptive thinking, learning control of inappropriate fantasies and behaviors, skills building, and victim empathy. Low recidivism rates have been reported with these forms of therapy (see discussion below regarding outcome studies).

The authority of the criminal justice system was essential in treating incest cases; program dropouts were largely those men who were not under criminal justice system supervision. In a 1994 literature review, Marshall and Pithers63 note the prevalence of these multidimensional approaches, saying, “Implementation of a single therapeutic intervention, even by the most high skilled practitioners, cannot be considered sufficient treatment for most sex offenders, and we doubt that anyone today would deem such an approach to be satisfactory.”

Relapse Prevention Programs

A model gaining in popularity is the “relapse prevention” model.86,87 Under this model, therapists assist the child molester in identifying the molester’s cognitive and behavioral patterns that are precursors to sexual abuse. This model enhances selfmanagement techniques and provides supervision. Techniques used to increase self-management are based on cognitive behavioral intervention, combined with community-based supervision, which can include probation or parole officers, family members, or other designated people in the community. Efficacy of Treatment Numerous serious methodological issues limit the scientific validity of existing studies and preclude firm statements about the efficacy of treatment of sex offenders in general. In addition, the diversity of the offender population and the rapid and substantial change in treatment methods over the past 15 years complicate our understanding of this complex topic. While no treatment has been shown to be 100% effective, this author believes the research literature provides definite grounds for optimism about the responsiveness of some segments of the offender population to existing treatment modalities. Methodological Concerns

To assess recidivism rates after treatment, it is essential to compare the behavior of treated offenders with the behavior of comparable untreated offenders; to create comparable groups, ideally offenders should be randomly assigned to treatment and control groups. However, this author and others have argued that it is ethically questionable to randomly assign sex offenders to treatment or control groups. Rather, determinations regarding community safety and offender motivation generally dictate which offenders are placed in treatment. On the other hand, it has been argued that the greater good would be served by conducting efficacy studies with maximum methodological rigor, including random assignment, so that effective methods could be more quickly and clearly identified and disseminated. In addition to the lack of random assignment in most studies, other methodological concerns are discussed at length by Furby and colleagues in an exhaustive literature review of treatment efficacy studies.50 They note that sample selection

189

Offenders: Characteristics and Treatment generally represents a specialized group of offenders: some treatment programs handle the most dangerous cases (judged most likely to reoffend), while other programs restrict themselves to the easiest cases (the offenders most motivated to reform), but almost no programs treat a representative sample of all offenders, or even of incarcerated offenders. In addition, if clinicians allow offenders into treatment according to such factors as motivation for treatment or because they presumably pose a greater threat to the community, then substantial differences may exist between treatment and control groups. According to Furby, other common methodological limitations in the literature assessing treatment effectiveness include the problem of attrition among treatment participants (often as high as 30%) and the often relatively short followup period. Studies of Adult Offender Treatment Programs

In 1989, Furby50 reviewed 42 published and unpublished sex-offender recidivism studies (30 treatment outcome studies and 12 studies of untreated offenders) with sample sizes over 10. The review excluded studies in which outcomes were assessed solely by means of self-report or physiological or psychological measures. Within those criteria, the review was intended to be exhaustive. After noting the numerous methodological challenges and varying study results, the authors conclude “despite the relatively large number of studies of sex offender recidivism, we know very little about it. . . . There is as yet no evidence that clinical treatment reduces rates of sex reoffenses in general and no appropriate data for assessing whether it may be differentially effective for different types of offenders.” More recently, Marshall and Pithers63 have pointed out problems with the Furby analysis. More than half of the Furby data come from subjects at programs which are now closed and whose treatment methods are now considered outdated. Marshall and Pithers note that “treatment approaches of that time (for example, milieu therapy, nondirective group therapy) bear little resemblance to the present-day multifaceted cognitive-behavioral programs that employ cognitive restructuring, masturbatory reconditioning, role playing, skills training, desensitization, stress management, and other techniques” to target a range of behavior, attitude, and cognitive changes.

Furby’s critique of treatment efficacy was also discussed in a 1991 literature review by Marshall, Jones, and colleagues,62 who reframed the question by stating, “At this stage in the development of our understanding of treatment and its effectiveness, we believe that the best approach is to ask not the categorical question ‘Can sex offenders be treated?’ but rather a

Numerous serious methodological issues limit the scientific validity of existing studies and preclude firm statements about the efficacy of treatment of sex offenders in general. more modest question such as ‘Can we discern grounds for optimism in the treatment outcome literature?’” While acknowledging that, in general, studies of nonbehavioral psychological treatment showed no impact on recidivism, Marshall noted that other studies did show treatment effectiveness in reducing recidivism. With the exception of the Ontario program cited in the next sentence, none of these studies were included in Furby’s review, because they were either in press at the time or were published after Furby’s review. An Ontario penitentiary program88,89 found recidivism rates of approximately 11% (though that rate changed to 18% to 20% in later years when the program was revised).90 In Vermont, a prison-based treatment program achieved a 3% recidivism rate over a six-year followup period.91 These results, plus the results of Barbaree and Marshall92 and Abel,61 discussed below, led Marshall to conclude that, at the least, treatment and supervision can delay recidivism in some and prevent it in others. Barbaree and Marshall,92 in a study of 68 treated and 58 untreated child molesters followed for 1 to 11 years (average 4-year follow-up), found a recidivism rate of 13% for the treatment group and 34% for the untreated group. When the untreated group was further subdivided, recidivism was 22% for incest offenders and 43% for nonincest offenders. The untreated group was composed of men who admitted their guilt and requested treatment but did not receive it. Treatment consisted of electrical aversion, masturbatory reconditioning, self-administration of smelling salts contingent on deviant

190

THE FUTURE OF CHILDREN – SUMMER/FALL 1994 thoughts or urges, and skills training. Conflict resolution and constructive use of leisure time were also taught. 93

Lang, Pugh, and Langevin described a multimodal treatment for 29 incest offenders and 27 heterosexual pedophiles. By the end of a three-year follow-up, 7% of the incest offenders and 18% of the pedophiles had reoffended. Rice and colleagues94 evaluated 153 nonincest child molesters, including 50 who participated in their treatment program. Of the 50 participants, 46 received electrical aversion therapy, and 32 were given biofeedback training, with both procedures aimed at reducing deviant arousal. Progress was measured by changes in arousal patterns, but only half of the subjects reached the goal criterion. It should not be surprising that this brief intervention, involving no aftercare or

clinical follow-up for serious offenders, had no measurable impact on recidivism. Similarly, Hanson57 provided shortterm treatment in the form of aversion therapy, individual and group counseling, and other general treatments to 106 incestuous and nonincestuous child molesters. Compared with two control groups of 31 and 60 untreated offenders, there was no significant difference in recidivism rates. However, there are indications that members of the treatment group were substantially more “hard core” than members of the control groups. Sixty-three percent of the treated sample had previous sexual convictions, versus 32% and 35% for the control groups. A more extensive treatment program was provided by Abel to 192 nonincarcerated pedophiles.61 This population was a subgroup of the 561 nonincarcerated sex offenders studied by Abel, discussed at the beginning of this article. Treatment consisted of thirty 90-minute weekly group sessions, focusing on decreasing deviant arousal, sex education and cognitive restructuring, and social and assertiveness skills training. Of the 192 pedophiles entering this voluntary treatment program, 35% dropped out. Of the 98 pedophiles evaluated one year after treatment ended, 12 had recidivated. Recidivists were likely to have a history of more varied offending behavior and more varied targets than nonrecidivists. A methodologically superior study by Marques and colleagues51 (currently in progress) compares three groups of sex offenders: matched volunteers who were randomly assigned to treatment or no treatment and a matched group of untreated nonvolunteers. Few sex-offender treatment studies have ever been able to create such closely matched control groups. Treatment refusers and treatment dropouts are also being studied. The treatment program is extensive. Participants attend group meetings 4.5 hours per week and receive 3 hours of individual counseling weekly. All participants are also provided a series of specialty groups based on individual need, such as social skills training, sex education, stress and anger management, substance abuse, and preparation for release. To preserve consistency of interventions, all treatment groups are conducted from manuals specifying goals and methods for each session.

191

Offenders: Characteristics and Treatment The Marques study is still in process; preliminary data (based on 7 years of a planned 15-year study) do not yet yield conclusive results regarding the program’s effectiveness. Two statistically significant findings in the preliminary data are that (1) those who dropped out of treatment were at higher risk for new sex crimes than were those completing a year or more of the program, and (2) treated child molesters were less likely to commit violent nonsexual crimes after release than were molesters in the volunteer control group. Effectiveness of Juvenile Offender Treatment Programs

While there are more than 800 treatment providers in the United States for youthful sexual offenders, relatively few providers are systematically evaluating their treatment and reporting their results in the research literature. There is, however, an emphasis in the literature on two approaches to treatment of juveniles: cognitive-behavioral and multisystemic. Becker, Kaplan, and Kavoussi95 describe a cognitive-behavioral treatment program for adolescent sexual offenders. The components of treatment include methods to (1) eliminate or decrease deviant sexual thoughts; (2) alter maladaptive belief systems, for example, that a young child who complies out of fear actually “enjoys” the sexual contact; (3) appreciate the consequences such behaviors have to victim and offender, and learn how to control inappropriate urges and behaviors; (4) increase social interactional skills; and (5) increase sex knowledge and clarify sex values, for example, that one should not engage in sexual relations outside a caring, responsible, consensual relationship. Becker96 reports an 8% recidivism rate for a sample of 80 youths followed in some cases for up to two years with this treatment format. Recidivism was determined by interviewing the youths, their families, and the referral source. Multisystemic therapy helps the offender improve his functioning in a variety of milieus, with emphasis on cognitive processes (changing maladaptive beliefs), family relations, peer relations, and school performance. Borduin and colleagues64 compared individual therapy to multisystemic therapy in a small sample. At completion of therapy, each youth was followed, on average, for 37 months. Only 12.5% of those youths receiving the mul-

tisystemic therapy reoffended, as compared with 75% sexual recidivism for youths receiving individual therapy. More recently, Schram, Milloy, and Row,97 in an unpublished report, evaluated the reoffending behaviors of 197 male adolescent sex offenders who had received various types of treatment at a number of treatment centers in their state.

States and localities vary widely in the amount of supervision, if any, they provide to sex offenders who are released from incarceration.

The average length of follow-up was 6.8 years. Ten percent of the youth were reconvicted for new sex offenses; however, 48% were convicted for nonsex offenses. Factors associated with recidivism included (1) having deviant arousal patterns, (2) having a history of truancy, (3) experiencing cognitive distortions, and (4) having one prior conviction for a sex offense. In this author’s view, parent support and educational groups should ideally be run concurrently with juvenile offender treatment. According to the Safer Society 1992 survey, 90% of juvenile sex offender programs include family therapy in their range of services, but no information is available on the rate or quality of parent participation. Unfortunately, there does not exist in the literature a description of a controlled treatment outcome study for parents of youthful sexual offenders. Need for Postincarceration and Posttreatment Services States and localities vary widely in the amount of supervision, if any, they provide to sex offenders who are released from incarceration. Some provide no supervision at all, some maintain registries of sex offenders, and some require probation or parole, either for all released sex offenders or at the discretion of the parole board or sentencing judge. Kim English, research director of the Colorado Division of Criminal Justice, is preparing a national survey of local practices in this regard, under a grant from the National Institute of Justice.98

THE FUTURE OF CHILDREN – SUMMER/FALL 1994

192

As described earlier, a model gaining in popularity is the “relapse prevention” model. This model emphasizes the need for a continuity of counseling and supportive services for nonincarcerated sex offenders, especially those recently released from custody, coupled with close supervision and immediate sanctions for inappropriate behavior, such as drug abuse or possession of child pornography. In Arizona, a 1987 law allows sex offenders to be placed on lifelong probation, allowing continuous supervision and sanctions for inappropriate activities that are likely to precede reoffending.99 Under the Arizona law, the trial court has discretion over whether to order lifetime probation for sex offenders. In practice, virtually all convicted child molesters receive this sentence, either in lieu of prison time or upon release from prison.100 Once on probation, offenders are required to take part in 45 hours of education over six weeks, then continue weekly 2-hour group treatment sessions so long as the probation office determines there is a need.101 In most cases, the offender pays for the therapy and for periodic physiologic testing. A special detail of the probation team makes unannounced visits on evenings and weekends to those probationers considered at highest risk of reoffending. Although the probation department considers their supervision of offenders to be strict, it appears that the sex offender unit does not return a higher percentage of probationers to prison than do the other units of the state’s adult probation office. 100

In Arizona, a 1987 law allows sex offenders to be placed on lifelong probation, allowing continuous supervision and sanctions for inappropriate activities. There are as yet no published data on the impact of this specialized probation program on recidivism, although evaluation of the program is ongoing. However, this model has support among professionals in the field. The American Bar Association, in a report based on a telephone survey of 100 probation departments and extensive site visits in four jurisdictions,

reported that most officials adamantly believed that specialized, intensively trained probation units with reduced case loads were needed to supervise child sexual abuse offenders on probation.68

Recommendations n Empirically derived typologies need to be developed for incest-only offenders, female offenders, offenders who chose both adults and children as victims, and noncontact offenders. n There is a lack of empirically derived theory on the etiology of sex-offending behavior. Current knowledge about typology should be used in developing an improved understanding of the etiology of child molestation. n We need controlled treatment outcome studies with long-term follow-up that assess which treatments are most effective with different categories of offenders. n While the mental health profession has progressed at developing treatment programs for juvenile offenders, the wider community and, in particular, the justice system have not made much progress toward better understanding and control of inappropriate juvenile sexual behavior. Community members—including parents, teachers, police, child protective workers, and juvenile court judges—are in need of education regarding what constitutes age-appropriate versus age-inappropriate sexual behaviors. They are greatly in need of information as to how to respond to youth who abuse sexuality, and the consequences of young offenders inadvertently receiving the message that their behavior is acceptable. The community also needs to be informed of the availability and the success of various treatment strategies. n A continuum of care should be made available so that individuals can receive treatment if they are incarcerated or placed in residential treatment, followed by coordinated services after release. We are in need of specialized group homes, therapeutic foster care, specialized day treatment, and community-based treatment for both individuals and groups. n Both civil commitment and incarceration programs that make use of indeterminate sentencing need to implement clear and objective criteria for releasing offenders.

Offenders: Characteristics and Treatment 1. Number of sex offenders in prisons increases 48% (survey). Corrections Compendium (1991) 16:9-16. 2. Ford, R. A need for help is going unmet: Sex offender treatment cut. Boston Globe. April 11, 1993, at Metro/Region p. 29. 3. American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 3rd ed. rev. Washington, DC: APA, 1987. 4. Abel, G., Becker, J., Cunningham-Rathner, J., et al. Multiple paraphilic diagnoses among sex offenders. Bulletin of the American Academy of Psychiatry and the Law(1988) 16,2:153-68. 5. Scott, L. Sex offenders: Prevalence, trends, model programs, and costs. In Critical issues in crime and justice. A. Roberts, ed. Thousand Oaks, CA: Sage, 1994. In a 1989-90 sample of 41 exhibitionists referred to probation officers in a Phoenix, Arizona, treatment program, 38 admitted to additional “hands-on” behavior involving minors. 6. Abel, G., and Osborn, C. The paraphilias: The extent and nature of sexually deviant and criminal behavior. In Psychiatric clinics of North America. J. Bradford, ed. Philadelphia: W.B. Saunders, 1992, pp. 675-87. Of 146 individuals who carried out hands-off deviant sexual acts, 64% previously or concomitantly had carried out inappropriate sexual behavior involving touching of the victim. 7. Rogers, C., and Tremain, T. Clinical interventions with boy victims of sexual abuse. In Victims of sexual aggression. I. Stuart and J. Greer, eds. New York: Van Nostrand Reinhold, 1984, pp. 91-104. 8. Sgroi, S. Handbook of clinical intervention in child sexual abuse. Lexington, MA: D.C. Heath, 1982. 9. This procedure was an important advance in research, since normally a researcher would be required to report to authorities any child molestations revealed by study participants; this author is unaware of any other study of sex offenders conducted utilizing a certificate of confidentiality. 10. Johnson, R., and Shrier, D. Sexual victimization of boys: Experience at an adolescent medicine clinic. Journal of Adolescent Health Care (1985) 6:372-76. 11. Becker, J. The effects of child sexual abuse on adolescent sexual offenders. In Lasting effects of child sexual abuse. G.E. Wyatt and G.J. Powell, eds. Beverly Hills, CA: Sage, 1988, pp. 193-207. 12. Longo, R. Sexual learning and experiences among adolescent sexual offenders. International Journal of Offender Therapy and Comparative Criminology (1982) 26:235-41. 13. DeFrancis, V. Protecting the child victim of sex crimes committed by adults. Denver, CO: American Humane Association, 1969. 14. Christie, M., Marshall, W., and Lanthier, R. A descriptive study of incarcerated rapists and pedophiles. Unpublished paper. 15. Abel, G., Mittelman, M., and Becker, J. Sex offenders: Results of assessment and recommendations for treatment. In Clinical criminology: Current concepts. J. Ben-Aron, S. Hucker, and C. Webster, eds. Toronto: M & M Graphics, 1985. 16. Groth, A., Longo, R., and McFadin, J. Undetected recidivism among rapists and child molesters. Crime and Delinquency (July 1982) 28:450-58. 17. Ageton, S. Sexual assault among adolescents. Lexington, MA: Lexington Books, 1983. 18. Showers, J., Farber, E., Joseph, J., et al. The sexual victimization of boys: A three-year survey. Health Values: Achieving High-Level Wellness (1983) 7:15-18. 19. Farber, E., Shower, J., Johnson, C., et al. The sexual abuse of children: A comparison of male and female victims. Journal of Clinical Child Psychology (1984) 13:294-97. 20. Ryan, G. Juvenile sex offenders: Defining the population. In Juvenile sexual offending. G. Ryan and S. Lane, eds. Lexington, MA: Lexington Books, 1991, pp. 3-8. 21. Awad, G., Saunders, E., and Levene, J. A clinical study of male adolescent sex offenders. International Journal of Offender Therapy and Comparative Criminology (1979) 28,2:105-16. 22. Fagan, J., and Wexler, S. Explanations of sexual assault among violent delinquents. Journal of Adolescent Research (1986) 3:363-85. 23. Katz, R. Psychological adjustment in adolescent child molesters. Child Abuse & Neglect (1990) 14:567-75. 24. Smith, W., Monastersky, C., and Deisher, R. MMPI-based personality types among juvenile sexual offenders. Journal of Clinical Psychology (1987) 43,4:422-30.

193

194

THE FUTURE OF CHILDREN – SUMMER/FALL 1994 25. Smith, H., and Israel, E. Sibling incest: A study of the dynamics of 25 cases. Child Abuse & Neglect (1987) 11:101-8. 26. Fehrenbach, P., Smith, W., Monastersky, C., and Deisher, R. Adolescent sexual offenders: Offender and offense characteristics. American Journal of Orthopsychiatry (1986) 56:225-33. 27. Becker, J., Kaplan, M., Tenke, C., and Tartaglini, A. The incidence of depressive symptomatology in juvenile sex offenders with a history of abuse. Child Abuse & Neglect (1991) 15:531-36. 28. Kaplan, M., Hong, G., and Weinhold, C. Epidemiology of depressive symptomatology in adolescents. Journal of the Academy of Child Psychiatry (1984) 23:91-98. 29. Tarter, R., Hegedus, A., Alterman, A., and Katz-Garris, I. Cognitive capacities of juvenile, non-violent and sexual offenders. Journal of Nervous and Mental Disease (1983) 171:564-67. 30. Lewis, A., Shankok, S., and Pincus, J. Juvenile male sexual assaulters. American Journal of Psychiatry (1979) 136:1194-96. 31. Davis, G., and Leitenberg, H. Adolescent sex offenders. Psychological Bulletin (1987) 101:417-27. 32. Barbaree, H., Marshall, W., and Hudson, S., eds. The juvenile sex offender. New York: Guilford, 1993. 33. O’Connell, M., Leberg, E., and Donaldson, C. Working with sex offenders: Guidelines for therapist selection. Newbury Park, CA: Sage, 1990. 34. Thorman, G. Incestuous families. Springfield, IL: Charles C. Thomas, 1983. 35. Knight, R., Carter, D., and Prentky, R. A system for classification of child molesters. Journal of Interpersonal Violence (1989) 4,1:3-23. 36. Knight, R., Prentky, R., and Cerce, D. The development, reliability, and validity of an inventory for the multidimensional assessment of sex and aggression. Criminal Justice and Behavior (1994) 21,1:72-94. 37. Quinsey, V. Men who have sex with children. In Law and mental health: International perspectives. D. Weisstub, ed. New York: Pergamon Press, 1989, pp. 140-72. 38. Bradford, J. The antiandrogen and hormonal treatment of sex offenders. In Handbook of sexual assault: Issues, theories and treatment of the offender. W. Marshall, D. Laws, and H. Barbaree, eds. New York: Plenum, 1990, pp. 297-327. 39. Knopp, F. Retraining adult sex offenders: Methods and models. Orwell, VT: Safer Society, 1984. 40. Marshall, W., and Eccles, A. Pavlovian conditioning processes in adolescent sex offenders. In The juvenile sex offender. H. Barbaree, W. Marshall, and S. Hudson, eds. New York: Guilford, 1993, pp. 118-42. 41. Finkelhor, D. Child sexual abuse: New theory and research. New York: Free Press, 1984. 42. Marshall, W., and Barbaree, H. An integrated theory of the etiology of sex offending. In Handbook of sexual assault: Issues, theories and treatment of the offender. W. Marshall, D. Laws, and H. Barbaree, eds. New York: Plenum Press, 1990, pp. 257-75. 43. Ryan, G., Lane, S., Davis J., and Isaac, C. Juvenile sex offenders: Development and correction. Child Abuse & Neglect (1987) 11:385-95. 44. Becker, J., and Kaplan, M. The assessment of adolescent sexual offenders. In Advances in behavior assessment of children and families. R. Prinz, ed. Greenwich, CT: JAI Press, 1988, pp. 97-118. 45. Abel, G., Lawry, S., Karlstrom, E., et al. Screening tests for pedophilia. Criminal Justice and Behavior (1994) 21:115-31. The authors found good preliminary results using the Abel Screen, a three-part test involving (1) a questionnaire, (2) a series of self-rated sexual responses to slides of partially clothed children and adults, and (3) a hand monitor that records client physiological responses to slides of unclothed children and adults. In tests, the Abel Screen correctly classified 98% of the normal, nonmolesting control group (n = 101) and 84% of the offenders against boys (n = 55). The instrument was less effective at discriminating between nonoffenders and offenders against girls, correctly classifying 77% of the normal, nonoffending control group (n = 101) and 88% of the offenders against girls (n = 130). 46. Beck, A., and Shipley, B. Recidivism of prisoners released in 1983. Bureau of Justice Statistics Special Report. Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics, 1989.

Offenders: Characteristics and Treatment 47. Klein, S., and Caggiano, M. The prevalence, predictability and policy implications of recidivism. Report prepared for the Bureau of Justice Statistics, U.S. Department of Justice. Santa Monica, CA: RAND Corporation, 1986. 48. Effectiveness of statutory requirements for the registration of sex offenders. Report from the California Department of Justice. Sacramento: California Department of Justice, July 1988. 49. Other sources agree that arrest records seriously underestimate child sexual abuse. In the Abel study (see note no. 4), offenders reported that their average ratio of arrests to commissions of rape and child molestation was approximately 1 to 30, and the ratio of arrests to commission of noncontact crimes, such as exhibitionism and voyeurism, was approximately 1 to 150. 50. Furby, L., Weinrott, M., and Blackshaw, L. Sex offender recidivism: A review. Psychological Bulletin (1989) 105,1:3-30. 51. Marques, J., Day, D., Nelson, C., and West, M. Effects of cognitive-behavioral treatment on sex offender recidivism: Preliminary results of a longitudinal study. Criminal Justice and Behavior (1994) 21:28-54. 52. Romero, J., and Williams, L. Recidivism among convicted sex offenders: A 10-year follow-up study. Federal Probation (1985) 49,1:58-64. 53. Prentky, R., and Burgess, A. Rehabilitation of child molesters: A cost-benefit analysis. In Child trauma 1: Issues and research. A. Burgess, ed. New York: Garland, 1992, pp. 417-42. 54. Marshall, W., and Barbaree, H. An outpatient treatment program for child molesters. Annals of the New York Academy of Sciences (August 1988) 528:205-14. 55. Christiansen, K., Elers-Nielsen, M., LeMaire, L., and Sturup, G. Recidivism among sexual offenders. In Scandinavian studies in criminology. Vol. 1. K. Christiansen, ed. London: Tavistock Press, 1965, pp. 55-85. 56. Soothill, K., and Gibbens, T. Recidivism of sexual offenders: A re-appraisal. British Journal of Criminology (1978) 18:267-76. 57. Hanson, R., Steffy, R., and Gauthier, R. Long-term followup of child molesters: Risk predictors and treatment outcome. Unpublished report prepared for the Corrections Branch, Ministry of the Solicitor General of Canada, 1992. 58. McGrath, R. Sex-offender risk assessment and disposition planning: A review of empirical and clinical findings. International Journal of Offender Therapy and Comparative Criminology (1991) 35:328-50. 59. Gibbens, T., Soothill, K., and Way, C. Sibling and parent-child incest offenders. British Journal of Criminology (1978) 18:40-52. 60. Maletsky, B. Treating the sexual offender. Newbury Park, CA: Sage, 1991. Offenders with histories of multiple paraphilias were more than five times more likely to be treatment failures or recidivate. 61. Abel, G., Mittelman, M., Becker, J., et al. Predicting child molesters’ response to treatment. Annals of the New York Academy of Sciences (August 1988) 528:223-34. 62. Marshall, W., Jones, R., Ward, T., et al. Treatment outcome with sex offenders. Clinical Psychology Review (1991) 11:465-85. On page 467, the authors cite five studies which concluded that the likelihood of reoffense among released offenders with more than one prior conviction was significantly greater than among first offenders. 63. Marshall, W., and Pithers, W. A reconsideration of treatment outcome with sex offenders. Criminal Justice and Behavior (1994) 21:10-27. 64. Borduin, C., Henggeler, S., Blaske, D., and Stein, R. Multi-systemic treatment of adolescent sexual offenders. International Journal of Offender Therapy and Comparative Criminology (1990) 34:105-13. 65. See note no. 51, Marques, Day, Nelson, and West, p. 37. The California full-time therapeutic program for incarcerated sex offenders is limited to 50 beds, with approximately 15,000 sex offenders in California prisons. 66. See note no. 5, Scott, p. 68. Arizona part-time program for incarcerated sex offenders is limited to 55 beds, with approximately 2,200 incarcerated sex offenders. 67.

Kessler, B. Sex offenders programs get start in state prisons: Texas’ effort stymied by money, perceptions. Dallas Morning News. October 17, 1993, at Al. “The number enrolled in the Texas program, just 200, is dwarfed by the 10,000 or so who could be.”

195

196

THE FUTURE OF CHILDREN – SUMMER/FALL 1994 68. Smith, B., Hillenbrand, S., and Goretsky, S. The probation response to child sexual abuse offenders: How is it working? Chicago: American Bar Association, 1990. 69. Prentky, R., and Burgess, A. Rehabilitation of child molesters: A cost-benefit analysis. American Journal of Orthopsychiatry (January 1990) 60:108-17. See pp. 115-16. 70. Wash. Rev. Code, Title 17, §71.09.020 et seq. 71. In re Young, 857 P.2d 989 (Wash. 1993). 72. Knopp, F., Freeman-Longo, R., and Stevenson, W. Nationwide survey of juvenile and adult sexoffender treatment programs and model. Orwell, VT: Safer Society, 1992. 73. Wash. Rev. Code, Title 18, §155. 74. Telephone call between Donna Terman, Center for the Future of Children, and Pat Freeman-Longo, co-director of the Safer Society Program, May 10, 1994. 75. Berliner, L. Commentary: Is family preservation in the best interest of children? Journal of Interpersonal Violence (1993) 8:556-57. 76. Gelles, R. Commentary: Family reunification/family preservation: Are children really being protected? Journal of Interpersonal Violence (1993) 8:557-62. 77. Lalumiere, M., and Quinsey, V. The discriminability of rapists from non-sex offenders using phallometric measures: A meta-analysis. Criminal Justice and Behavior (1994) 21:150-75. 78. Pithers, W., and Laws, D. The penile plethysmograph. In A practitioner’s guide to treating the incarcerated male sex offender. B. Schwartz, ed. Washington, DC: U.S. Government Printing Office, 1988, pp. 85-94. 79. Heim, N., and Hursch, C. Castration for sex offenders: Treatment or punishment? Archives of Sexual Behavior (1979) 8:281-304. 80. Berlin, F., and Meinecke, C. Treatment of sex offenders with antiandrogenic medication: Conceptualization, review of treatment modalities, and preliminary findings. American Journal of Psychiatry (1981) 138:601-7. Eleven of 20 patients discontinued taking MPA against medical advice, and 10 of those 11 reoffended. 81. Kafka, M. Successful antidepressant treatment of nonparaphilic sexual addictions and paraphilias in men. Journal of Clinical Psychiatry (1991) 52:60-65. 82. Crawford, D. Treatment approaches in pedophiles. In Adult sexual interest in children. M. Cook and K. Howells, eds. London: Academic Press, 1981, pp. 181-217. 83. Giarretto, H., Giarretto, A., and Sgroi, S. Coordinated community treatment of incest. In Sexual assault of children and adolescents. A. Burgess, A. Growth, L. Holstrom, and S. Sgroi, eds. Lexington, MA: Lexington Books, 1978. 84. Giarretto, H. A comprehensive child sexual abuse treatment program. Child Abuse & Neglect (1982) 6:263-78. 85. Giarretto, H. Humanistic treatment of father-daughter incest. Journal of Humanistic Psychology (1978) 32:20-25. 86. Pithers, W., and Kafka, M. Relapse prevention with sex aggressors: A method for maintaining therapeutic gain and enhancing external supervision. In The handbook of sexual assault: Issues, theories and treatment of the offender. W. Marshall, D. Laws, and H. Barbaree, eds. New York: Plenum Press, 1990, pp. 343-61. 87. Laws, D. Relapse prevention with sex offenders. New York: Guilford, 1989. 88. Davidson, P. Recidivism in sex offenders: Who are the bad risks? Paper presented at the Second National Conference on the Evaluation and Treatment of Sexual Aggressors. New York, May 1979. 89. Davidson, P. Outcome data for a penitentiary-based treatment program for sex offenders. Paper presented at the Conference on the Assessment and Treatment of the Sex Offender. Kingston, Ontario, March 1984. 90. Leger, G. Research on sex offenders: Regional treatment centre (Ontario). Forum on Corrections Research (1989) 1:21-28. 91. Pithers, W., and Cumming, G. Can relapses be prevented? Initial outcome data from the Vermont treatment program for sexual aggressors. In Relapse prevention with sex offenders. D. Laws, ed. New York: Guilford, 1989, pp. 313-25. 92. Barbaree, H., and Marshall, W. Deviant sexual arousal, offense history and demographic variables as predictors of reoffense among child molesters. Behavioral Sciences and the Law (1988) 6:267-80.

Offenders: Characteristics and Treatment 93. Lang, R., Pugh, G., and Langevin, R. Treatment of incest and pedophilic offenders: A pilot study. Behavioral Sciences and the Law (1988) 6:239-55. 94. Rice, M., Quinsey, V., and Harris, G. Sexual recidivism among child molesters released from a maximum security psychiatric institution. Journal of Consulting and Clinical Psychology (1991) 59:381-86. 95. Becker, J., Kaplan, M., and Kavoussi, R. Measuring the effectiveness of treatment for the aggressive adolescent sexual offender. Annals of the New York Academy of Sciences (August 1988) 528:215-22. 96. Becker, J. Treating adolescent sexual offenders. Professional Psychology: Research and Practice (1990) 21:362-65. 97. Schram, D., Milloy, C., and Rowe, W. Juvenile sex offenders: A follow-up study of reoffense behavior. Unpublished manuscript, 1991. 98. Telephone call between Donna Terman, Center for the Future of Children, and Susan Chadwick, research analyst, Colorado Division of Criminal Justice, June 8, 1994. 99. Ariz. Rev. Stat., §13-902 (D). 100. Telephone call between Donna Terman, Center for the Future of Children, and Laurie Scott, supervisor of Special Sex Offender Unit, Maricopa County (Arizona) Adult Probation, May 23, 1994. 101. See note no. 5, Scott. This chapter describes the Maricopa County, Arizona, sex offender program in detail, including special behavioral rules limiting paroled sex offenders’ contact with children.

197