[Zum Deutsche ‹bersetzung]
Sexuality & Culture, 4(2), 67-81 (2000)
Rind, Tromovitch, and Bauserman:
Four years later, Finkelhor (1979) proposed an ethical justification for prohibiting adult/child (defined as a prepubertal youngster) sexual behavior. The reason for using an ethical justification was that the justification based on psychological harm lacked cogency. According to Finkelhor, it was empirically weak since "it is possible that a majority of these children are not harmed" (p.693).
More recently, the Past President of the APA, Martin Seligman (1994), argued that the case for CSA being a "special destroyer of adult mental health" (p. 232) was far from proven. The existing research indicating harm, according to him, "abandoned methodological niceties" (p.233). These studies were characterized by sampling bias, lack of adequate control groups, and a failure to consider alternative explanations for the findings. He wrote: "Once the ideology is stripped away, we still remain ignorant about whether sexual abuse in childhood wreaks damage in adult life and, if so, how much" (p.234).
Of significance is the fact that the weight of the evidence, when objectively considered, has supported the notion that CSA is neither necessarily nor typically harmful. For example, Constantine (1981) reviewed 30 studies. He found that
In his review of 25 studies, Conte (1985), taking issue with Constantine's using the research "to make a case for 'legitimate instances of child-adult sex,'" concluded that "a review of the literature describing the effects of sexual abuse on children leads irrefutably to the ambiguous conclusion that sexual abuse appears to affect some victims and not others" (p. 117).
Similarly, Browne and Finkelhor (1986) reviewed 28 studies. They found that among adults who had experienced CSA less than 20 percent evidenced serious psychopathology. They noted with concern the efforts of child advocates to exaggerate the harmful effects for political purposes because of its potential to harm the victims and their families:
Kendall-Tackett, Williams, and Finkelhor (1993) reviewed 45 studies. They found that up to 49 percent of the sexually abused children suffered no psychological harm. They concluded that a lack of symptoms could not be used to rule out sexual abuse since "there are too many sexually abused children who are apparently asymptomatic" (p. 175). Further, among the children who were symptomatic, symptom abatement occurred for most within two years with or without treatment. These authors also found that when sexually abused children in treatment were compared with nonabused children in treatment, the sexually abused were less symptomatic than their nonabused clinical counterparts (p. 165).
In 1997, Rind and Tromovitch conducted a meta-analytic review of seven studies on the effects of child sexual abuse. Unlike prior reviews which were based primarily on clinical samples, this review involved studies that used national probability samples: four were from the United States, and one each, from Great Britain, Canada, and Spain. The findings indicated that child sexual abuse "is not associated with pervasive harm and that harm, when it occurs, is not typically intense" (p. 237). The findings of Rind, Tromovitch, and Bauserman, which caused the recent maelstrom, simply confirmed this earlier study.
Moreover, it has not been demonstrated that CSA has any influence upon the adult personality. For example, Beitchman, Zucker, Hood, DaCosta, Akman, and Cassavia (1992) reviewed 32 studies. They concluded that the evidence suggested that CSA has serious long-term effects, but that it was not clear to what extent these effects were due to CSA per se (p. 115). Levitt and Pinnell (1995) concluded, based on their review of the literature, that "the traditionally accepted link between childhood sexual abuse as an isolated cause and psychopathology in adulthood lacks empirical verification" (p.151). The Rind, Tromovitch, and Bauserman study(1998) indicated that CSA is non-causative. They reported that CSA-adjustment relations became nonsignificant when family environment was controlled for. Indeed, the evidence tends to confirm Seligman's earlier conclusion that
Thus, contrary to the APA's assertion that CSA should never be considered harmless, quite the opposite is the case. That is, the empirical evidence gives no reason to consider CSA as necessarily or even usually harmful.
Second, based on their findings, Rind, Tromovitch, and Bauserman made the important recommendation that the scientific community use more neutral terms to study the phenomena of adult-child and adult-adolescent sexual behavior. In their view, abusive sexual behavior would be reserved to situations involving an unwanted sexual encounter with negative reactions. Those situations involving a willing encounter with positive reactions would be labeled simply adult-child sex or adult-adolescent sex (p. 46). One might wish to further refine this recommendation (e.g., abuse should be defined when the child/adolescent is unwilling regardless of whether their action was negative or not).
Nevertheless, their recommendation was designed to move the scientific community beyond the victimological paradigm that has dominated the study of and response to CSA thus far. In this paradigm, the child or adolescent is viewed as a passive "victim" (Feierman, 1990). This is based on the conviction that the child or adolescent is incapable of experiencing sexual desire or initiating sexual contact. According to Okami (1990), this conviction "attributes participation in peer sexual behavior to 'curiosity' and participation in adult/nonadult sexual behavior to 'coercion'" (p. 93). Even behavior that is self-reported as positive by the child or adolescent is defined by the victimologists as abusive. According to Okami (1990), the victimological paradigm reflects a Victorian idealization of children as sexless innocents. This is politically correct, but is both historically incorrect (Bullough, 1990) and scientifically incorrect (Ceci, & Bruck, 1995; Friedrich, Grambsch, Broughton, Kuiper, & Beilke, 1991; Friedrich, Fisher, Broughton, Houston, & Shafran, 1998; Lamb, & Coakley, 1993).
The victimological paradigm has been responsible for much of the biased and polemical research that, as Seligman noted, has characterized this area of study (Okami, 1990). An alternative paradigm is needed; one which, as recommended by Crittenden (1996) in the handbook on child maltreatment of the American Professional Society on the Abuse of Children, considers this behavior "more as a common variant of human behavior than as abnormal behavior"(p. 166). Perceiving adult/non-adult sexual behavior in this manner would increase the likelihood of an unbiased approach to the study of adult/non-adult sexual behavior. It also would permit, as Crittenden points out, the application of what is known about normal sexual behavior to cases of what is now referred to as CSA. This is hardly possible if the scientific community insists on referring to all adult-child and adult-adolescent sexual behavior as abusive.
To recognize the distinction between abusive and non-abusive does not preclude identifying the interaction as immoral and/or illegal. Conte (1985) pointed out that decisions concerning the appropriateness of adult/nonadult sexual interactions involve ethical, legal, and religious principles. By way of example, robbery is unlawful not because it results in psychological harm but because society has decided that people have a right to their own property. Put another way, the question of the effects of child sexual abuse should not be confused with the moral and/or legal issue of dealing with this behavior. As argued earlier by Kilpatrick (1992), at the very least, in professional and scientific discussions, if not in moral and legal ones, abuse is something to be established as a conclusion rather than simply being accepted as a premise.
Why, then, might many in the scientific and professional community take the position that CSA is harmful and ignore the suggestion for the use of more neutral terminology in the study of CSA? Part of the answer, I believe, is the effort to avoid being vilified by the victimologists. Their attacks on anyone who seeks to bring a measure of rationality and objectivity to this problem is well known (Okami, 1990; Neimark, 1996). Consequently, anyone who calls for rationality and objectivity with respect to CSA will typically preface their remarks along the line taken by Seligman (1994): "So this preface: I believe sexual abuse is evil. It should be condemned and punished" (p. 232).
But there is an additional reason - money. As noted by Dineen (1999), the psychology industry (which she defined broadly to include psychologists, psychiatrists, psychoanalysts, clinical social workers and psychotherapists) needs victims to justify the expansion of its domain and, thus, it "manufactures victims." A similar point was made earlier by Tavris (1993) with respect to the incest-survivor recovery movement. CSA is a problem widely exploited by professionals according to Costin, Karger, and Stoesz (1996):
According to Dineen, Tavris, and Costin et al., then, child sexual abuse has become an arena of opportunism for and exploitation by some in the mental health industry. And it is a quite lucrative arena. For example, according to Nathan and Snedeker (1995),
Further, in most settings, children who have been sexually abused are routinely offered treatment even if asymptomatic (Beutler, Williams, and Zetzer, 1994). Finklehor and Berliner(1995) estimated that among substantiated cases of child sexual abuse, from 44 percent to 73 percent receive psychotherapy. A recent report of the National Institute of Justice (Miller, Cohen,& Wiersema, 1996), indicated that victims of child sexual abuse were much more likely than victims of other crimes to receive mental health care. These data showed that up to 50% and more of CSA victims received mental health care as compared with no more than 4% of victims of other crimes. And the average cost of mental health services for the typical victim of child sexual abuse was nearly sixty times greater than that for the victim of another crime ($5,800 vs. less than $100).
There is also the cost associated with the derivative of CSA, viz., repressed memory therapy. For example, in Washington state's Crime Victims Compensation Program, the average cost associated with treating adults whose claims were based on repressed memory of childhood sexual abuse was approximately four times the average claim in other mental health claims (Loftus, 1997; Parr, 1996). The average cost of non-repressed memory claims was less than $3,000, while that of the repressed memory claims was more than $12,000, with one claim exceeding $50,000. In just over four years, the citizens of Washington paid out over 2.5 million dollars for 325 repressed memory claims. The primary diagnosis in most of these claims was Multiple Personality Disorder (MPD). It was not unusual for the claimant to have dozens or even hundreds of personalities - one claim involved over 700 alter states and another over 3000. All thirty were still in therapy three years after their first memory surfaced, and sixty percent (18) were still in therapy five years after their first memory surfaced. As noted by Piper (1994), treatment for MPD entails long and costly therapy. It is not, according to him, cost-effective. Repressed memory therapy offers the perfect example for what Campbell (1994) referred to as therapists operating "rent-a-friend" agencies with long-term leases (p. 20). These well serve the therapists' interests but not those of their clients.
This suggests the question: what do the consumers and public at-large get in return for their money. With respect to the treatment of children and adolescents who have been involved in adult/nonadult sexual behavior, they get little or nothing. Finkelhor and Berliner (1995) reviewed 29 studies concerned with the effectiveness of treating sexually abused children. Of the 29 studies, 17 used a pre-post design. While nearly all reported positive improvement, it cannot be said that the improvement was due to the treatment. As Finkelhor and Berliner note, longitudinal studies have shown that sexually abused children improve over time with or without treatment (p. 1409). Three of the 7 experimental design studies compared treatment and no-treatment groups. These found significant effects of treatment, but the reviewers commented that their "relatively small-scale designs ... detract from their scientific weight" (p. 1414). Among the quasi-experimental studies which had equivalent groups (3 of the 5 reviewed) there was no advantage for children receiving therapy compared with children not receiving therapy. While these reviewers took an optimistic posture with respect to the outcome of therapeutic intervention, they noted that current research is methodologically flawed and concluded that the effectiveness of sexual abuse treatment has yet to be proven (p. 1415).
However, the weight of the evidence in this review parallels those found in naturalistic studies on the effectiveness of child and adolescent psychotherapy, namely that they have little or no effect (U.S. Department of Health and Human Services, 1999; Weisz, Weiss, & Donenberg,1992; Weisz, Weiss, Han, Granger, & Morton, 1995). These findings are reinforced by evidence from studies on continuum of care programs for children and adolescents. One of the more ambitious of these was the Fort Bragg Project (Bickman, 1996). The U.S. Army spent 80 million dollars to demonstrate that "a continuum of mental health and substance abuse services for children and adolescents was more cost-effective than services delivered in the more typical fragmented system" (p. 689). The project offered in- and out-patient services to more than 42,000 child and adolescent dependents of military personnel in the Fort Bragg (North Carolina) area for more than five years from June 1990 to September 1995. It was considered a model program by the American Psychological Association's section on Child Clinical Psychology and the Division of Child, Youth and Family Services Joint Task Force.
The study showed that the program produced better access to treatment, higher levels of client satisfaction, and fewer restrictions on treatment. The cost, however, was higher and the clinical outcomes were no better than those at the comparison site. The findings led Bickman "to question the assumption that clinical services provided in the community are effective" (p. 699).
But if it is not effective, can therapy for CSA be harmful? According to Seligman (1994), the answer to this question is: yes. He cautioned against therapy for the sexually abused and noted, for example, that it is often asserted that the sexually abused need to relive the experience and experience a catharsis in order to improve. Despite the fact that catharsis has a long history as a therapeutic technique, there is no evidence that it works (Bushman, Baumeister, & Stack,1999; Seligman, 1994). On the contrary, as Seligman suggested, reliving the event may be harmful since it heightens the event in the child's mind and interferes with the natural healing process (pp. 234-235).
In the area of treating adults with repressed memory therapy, this also may be harmful (Stocks, 1998). A study of the state of Washington's Crime Victims Compensation Program is suggestive, though not probative, of the harm that can occur in therapy for repressed memories of sexual abuse (Loftus, 1997; Parr, 1996). Between 1991 and 1995, in the state of Washington, 325 repressed memory therapy claims were awarded victim compensation. Loni Parr, a nurse consultant, and staff employees reviewed 183 of these claims. They randomly selected 30 from these in order to gain a preliminary profile of the cases. Their findings are alarming.
Overall, the status of these claimants deteriorated during treatment. Before recovering memories, three (10%) had attempted or thought of suicide; after recovering memories, 20 (67%) were suicidal. Before memories, only 2 (7%) had been hospitalized; after, 11 (37%) had been. Before the emergence of memories, only one woman (3%) had engaged in self-mutilation; after, 8 (27%) had mutilated themselves (Loftus, 1997).
Further, before entering therapy, twenty-five (83%) of the patients had a job; after three years of therapy, only 3 (10%) were still employed. Twenty-eight (93%) were married when they entered therapy; within five years, 18 of the 28 (64%) were divorced or separated. Twenty-one of the patients had minor children and one-third (7) lost custody of their children during therapy. All were estranged from their extended families (Loftus, 1997; Parr, 1996).
These patients were in therapy longer than other mental health patients, and evidenced a high rate of mental and emotional problems, all of which arose and worsened during therapy. In fact, the longer the patients were in therapy the more disabled they became. The primary diagnosis in these cases was Multiple Personality Disorder, and it was not unusual for claimants to have dozens or even hundreds of personalities; one person had over 3000! The findings of this study buttress the conclusion of Ofshe and Watters (1994):
Rather than distancing itself from the Rind, Tromovitch, and Bauserman study, the APA as well as the scientific and practice communities could have used the opportunity to:
1. Educate the community about the myths surrounding the problem of CSA. This includes laying to rest the myth that because a sexual activity violates a moral and/or a legal code that it is thereby necessarily or even usually psychologically harmful. In other words, it is time, as suggested by Rind and Tromovitch (1997), to stop equating wrongfulness with harmfulness in sexual matters.
The perpetuation of this myth is unethical and has possible iatrogenic effects, as noted sometime ago by Schultz (1980). He wrote:
An example of this "pitting" of the professional against the child was provided by Germaine Greer in 1975. She wrote of the experience of one of her school friends:
2. Undertake research in the area of adult/nonadult sexual behavior that is shorn of the ideological bias that has contaminated much of the research in this area. A beginning move in this direction necessitates limiting the label "child sexual abuse" in the scientific literature to those instances where the sexual behavior is abusive. Abusive sexual activity can be defined as an unwanted sexual experience that may involve coercion, threat, and/or demonstrable harm.
3. Stop automatically referring the sexually abused for therapy. CSA is not a psychiatric disorder or a syndrome (Finkelhor, & Berliner, 1995). Rather it is an event or series of events in a person's life. Treatment is indicated only when there is a currently demonstrable harm. To treat the asymptomatic child/adolescent is comparable to a physician treating child/adolescent for bicycle accidents. Many who have a bicycle accident do not require treatment. When they do need treatment, it is for the clinical condition rather than the event responsible for that condition. In other words, the asymptomatic child or adolescent should not be treated.
However, even when there is demonstrable harm, treatment should be recommended only with caution since it may, as pointed out by Seligman, only worsen the harm by interfering with the natural healing process. According to Seligman, the overreaction of parents and police, and early therapeutic intervention to undo "denial," and later therapeutic intervention to recover the "repressed" memory and then reliving the experience may do more harm than good. Thus, he recommended to parents whose child has been abused or who were themselves abused that they "turn the volume down as soon as possible" (p. 235).
The excessive and unnecessary provision of CSA treatment also takes resources from other victims and other victim needs (Costin et al., 1996). Lastly, and most importantly, it also makes the accurate evaluation of treatment effectiveness impossible since the treatment pool is contaminated by including those who do not need treatment in the first place.
4. Advise prospective clients of the risks of serious side-effects associated with therapy. They have the right to know the probabilities of a successful outcome versus a non-successful outcome, i.e., of getting worse and of not improving. Prospective clients have a right to know whether the treatment they are to be exposed to is empirically validated, is still experimental or has been discredited by sound research. With this information, prospective clients can make an informed decision as to whether or not to subject themselves or their children to the risks associated with therapy.
The Rind, Tromovitch, and Bauserman study of the impact of CSA among college students is politically incorrect but scientifically correct. It has a number of important implications for the research and practice communities. Among the more important is the need to stop exaggerating the negative impact of adult/nonadult sexual behavior, as suggested earlier by both Browne and Finkelhor, and Seligman. Another important implication is for conducting research that does not approach the issue of adult/nonadult sexual behavior with a political ideology as often has been the case thus far. And finally it is time to stop the common practices of 1)assuming that CSA causes psychological harm, and 2) routinely recommending psychotherapeutic intervention.