In 1998, Bruce Rind, Philip Tromovitch, and Robert Bauserman published a major scientific paper that reviewed the effects of child sexual abuse. They looked at scores of studies that had been conducted over several decades; in the end, they suggested that the term "child sexual abuse" was too strong, as it implied serious harm where harm did not necessarily exist. Instead, they advocated the neutral term "adult-child sex" and suggested that the term "child sexual abuse" should be reserved for cases in which actual harm had been demonstrated.
As you might expect, this proposal caused a tremendous amount of controversy. Support groups for abuse victims were enraged (quite understandably), fearing that sexual abusers would use this study to justify their actions. Their worries proved to be justified when the North American Man-Boy Love Association (NAMBLA...God help us) proudly proclaimed that Rind and his colleagues had vindicated their lifestyle choice--that sexual contact between an adult and a child was perfectly normal and not harmful in the slightest. Soon the controversy spread to the entire nation. The egregious Dr. Laura preached about the study on one of her radio programs, claiming that the study was invalid because the results of scientific research should be rejected when they contradict common sense.1 (Of course, this preposterous statement has no basis in reality, and I'm half-tempted to claim that it would justify revoking her PhD in physiology.) Columnists around the country adopted thoughtful and sonorous tones, wondering how the ivory-tower academics could possibly be so wrong. The United States Congress even weighed in with a resolution that condemned the study and expressed its support for victimized children everywhere.
Some of these arguments would've been reasonable if the article had actually claimed that sexual contact between adults and kids was A-OK. (Dr. Laura's comment would still have been cretinous, of course, and Congress would still have no business passing a resolution about the study, but nobody expects rational behavior from these people anyway.) As it happens, though, the article makes no such claim. At all. Ever. In fact, the overwhelming majority of the data they report indicate that adult-child sexual contact harms kids. In short, the people who commented on the article--Congress, NAMBLA, Dr. Laura, victims' organizations, columnists--did not bother to read it, or if they read it they did not understand it.0
I have read the article. I've read it several times. I have a background in clinical research that helps me to understand both their methods and the statistics they used to evaluate their data. I will explain what they did, what they found, and why their conclusions do not support the claims that were made in the popular media. With few exceptions, I won't discuss the truth or falsity of their claims; I just want to show that even if everything they said is true, their study does not support the claim that adult-child sexual activity is harmless.
Rind and his colleagues were trying to answer a fairly simple question. They stated that most people, whether researchers or laymen, believe that child sexual abuse has several properties:
- it causes harm;
- the harm is widespread and pervasive;
- the harm is intense;
- boys and girls react equally--that is, child sexual abuse is pervasively and intensely harmful for both genders.
In their article, Rind and his colleagues argued that these statements were false given the standard legal and clinical definition of child sexual abuse.
The emphasized portion of the last sentence is probably the most important part of their argument, and I'll get back to it in a moment. Right now, though, I'd like to point out that Rind and his colleagues set the bar pretty low. If you're going to argue against a statement of the form "x is y," you only need to show a single example in which x isn't y to prove your point.2 So their hypothesis would be proven if most (but not all) people said that their experience had been harmful.
But let's get back to the phrase I emphasized. In this study, Rind and his colleagues are using a specific definition of "child sexual abuse," and their conclusions only apply to that specific definition. It may not be your definition or my definition or Dr. Laura's definition or the commonsensical definition, but it's the definition they used, and it's based on the definition used in the majority of studies of child sexual abuse. And that definition encompasses any kind of sexual contact between an adult and a minor. As they wrote, it includes "the repeated rape of a 5-year-old girl by her father and the willing sexual involvement of a mature 15-year-old adolescent boy with an unrelated adult." Moreover, it generally includes noncontact sexual experience--acts like exhibitionism, or an adult showing a pornographic magazine to a teenage boy.
The breadth of the definition might seem a bit silly. If a 5-year-old is raped by her father, that's a nightmare (at least I hope most of us would so consider it); if a 15-year-old boy gets it on with an 18-year-old woman, that's more of a wet dream. And plenty of non-contact sexual experiences seem like they wouldn't necessarily be traumatic at all. Indeed, while I was reading this article, I realized that under this definition, I had been "sexually abused." One day, when I was thirteen or so, I was walking around town. I turned into an alley to take a shortcut to a comic-book store and froze when I saw a sketchy-looking man in a topcoat standing next to a dumpster. He turned towards me and opened it, revealing his naked body. Sure enough, he had a penis. It wasn't terribly interesting to me; I had certainly seen penises before (after all, I have one) and there wasn't anything unusual about his. I cursed at him and left. At the time, I thought it was slightly gross and slightly funny, and that's it. I didn't think about it much at all, and I don't think it's ever affected my sex life.
Anyway, Rind and his colleagues wanted to figure out whether child sexual abuse, as defined above, was really associated with the four properties they mentioned. Most studies of sexually abused kids involve only a handful of people, which makes it hard to draw firm conclusions, so they decided to combine data from lots of different studies, all of which looked at college-aged students who had been abused as kids. Now, one can argue endlessly about the ways in which they combined data or the methods they used to analyze it...but as I said, I'm not going to quibble about that here. Instead I'll just go over the conclusions they drew.
First they looked to see whether students who had been abused as kids had more psychological problems than those who hadn't been abused. On the whole, they did, though the overall difference was small; in particular, they were more likely to suffer from depression and anxiety. So on the whole, Rind and his colleagues found that sexual abuse did cause harm, even under the broad definition they used in their study. Notice that this result may underestimate the effects of the more vicious forms of child sexual abuse, because as I noted above (and will note again throughout this writeup), it includes noncontact and consensual sexual relations.
They also averaged participants' ratings of the immediate effects of the abuse. Here are the results:
Females: 11% positive, 18% neutral, 72% negative
Males : 37% positive, 29% neutral, 33% negative
First, note that for females, the experience was almost always harmful or neutral, supporting the contention that most girls are harmed by sexual experience during childhood. For the males, the picture is different--the percentages were roughly equal across all three categories. So these results support the claim that under the broad definition of sexual abuse, boys and girls react differently.
But what about the results for males? In my view, the difference between 37% positive and 33% negative is probably not large enough to be meaningful, and does not justify the claim that "sexual abuse of boys is usually positive." Even if the difference were meaningful, I scarcely think we could ignore the 33% of boys who report that the experience was unpleasant at the time. (If a given act is going to harm a third of the boys who experience it, can anyone seriously claim that it's an OK thing to do?3) Furthermore, we don't know exactly what happened to the boys in the "positive" category. We do know that on the whole, girls were more likely to experience abuse from close family members, and more likely to experience violent abuse; maybe they were also less likely to have "consensual" experiences4. Thus, males' experiences may have been less traumatic on the whole. For example, I would rate my immediate reaction to my own "abuse experience" as neutral or slightly positive (since it was funny). So the results for boys provide no support for the claim that "man-boy love" is a good thing. They do support the claim that "man-boy sexual interactions, including consensual or noncontact interactions, are not always harmful." That's a completely different notion, and while it supports Rind and colleagues' thesis, it's not even close to a ringing endorsement for pedophilia.
Next they examined the long-term effects of abuse. Here are the results:
Females: 16% positive, 25% neutral, 59% negative
Males : 42% positive, 32% neutral, 26% negative
Here again, the data cannot be used to support the claim that sexual contact between adults and children is harmless, as many of the victims clearly suffered harm. Moreover, it seems possible that these victims had felt harmed for a while, then managed to overcome it (through therapy, their own efforts, or the simple passage of time). Children can overcome many traumatic events, including a serious illness or violence or the death of a parent, but that doesn't mean that the events are good. But the results do support some of Rind and colleagues' assertions--namely that the harm is not necessarily pervasive and is not the same for males and females.
Note that this is a good thing: Even if these people have been traumatized and hurt, many of them can get past it. I think that's something that we would want to tell victims--that they can get through it and live a happy, healthy, and unimpaired sex life.
Finally, Rind and his colleagues examined the family environments of people who had been sexually abused. They noted that people who had been sexually abused had also experienced poor family environments. Specifically, they were more likely to live in homes with nonsexual abuse or neglect, conflict, and the like. In a way, this makes sense--if Dad sexually abused you, but also beat you and locked you in the closet and starved you, it might be hard to figure out exactly which kind of abuse caused what. Of course, that doesn't mean that sexual abuse causes no harm; it just means that it's hard to sort its effects out from the effects of other harmful acts.5
Thus the real results of the study provide no support whatsoever for the claim that sexual contact between adults and children is harmless. Rather, they show that such contact is frequently harmful and often has lasting negative effects. Rind and colleagues simply showed that one should not assume that any adult-child sexual activity is harmful and immediately throw the victim into therapy. Rather, one should be willing to accept the possibility that the victim is quite all right, especially if the victim (a) is an older male, (b) experienced noncontact sexual activity, or (c) actually had some desire to have the sexual experience. It ultimately supports a rather mild and even commonsensical conclusion--that the "standard" definition of child sexual abuse is probably a bit too broad for practical purposes. And it shows that it would be nice if people actually read the studies that they set out to condemn.
0Some people would argue that one should not publish an article that lends itself to this kind of misinterpretation. But that would be like banning Darwin because of the people who misuse evolutionary theory to justify claims of racial superiority.
1 I hope I may be forgiven my failure to look up the original reference. I can't listen to Dr. Laura without getting angry.
2For practical purposes, of course, you really want several cases, since a single case can always be some sort of bizarre exception. For example, a friend of mine (who had been adopted at a young age) discovered that she was three months younger than she thought she was, because someone had misinterpreted her birth certificate ("3/6" was read as "March 6th," when in her country of origin it really meant "June 3rd.") Suppose she had entered into a consensual relationship a month before her real sixteenth birthday. It would have been illegal, but neither the "abuser" nor the "victim" would have known it.
3Moreover, the statement that it was "pleasant at the time" is not necessarily straightforward. Abusers frequently use cajolery and bribery to seduce their victims, and they often seek out victims who seem vulnerable to such methods. (I shall not provide more details as I do not wish to provide everything2 with a child rape manual; those who need to understand can probably figure out what I'm talking about.) So the victim may have seen the events as positive at the time, only to realize the horror of it later on.
4"Consensual" is a tricky term because these kids can't legally consent. But underage boys can certainly be interested in sex and want to have a sexual experience. That's mainly what I'm referring to here.
5I should note that this is really my own interpretation of their results, and that their interpretation differs slightly; they essentially state that child sexual abuse has no detectable effect after the effects of other family environment variables (FEVs) are accounted for. While this is technically true given the methods they used, I think it's misleading. The explanation requires some understanding of statistics, though, which is why it's in this footnote. Essentially, the authors looked at a number of studies that used statistical techniques (usually multiple regression) to analyze the data. Typically, the studies entered the FEVs into the regression equation first, then entered the abuse variable. They often found that after entering all the FEVs (in some cases as many as 13), the effect of child sexual abuse dropped to 0. But this just means that the variables are intercorrelated--if one entered child sexual abuse first, then the FEVs, one might well find that the FEVs had no effect. So I think the proper conclusion is that the two are hard to tease apart, not that family environment is the real cause of the adjustment difficulties in people who experienced child sexual abuse.
Rind, B., Tromovitch, P., & Bauserman, R. 1998. A meta-analytic examination of assumed properties of child sexual abuse using college samples. Psychological Bulletin 154, 22-53.