THE DAMAGING EFFECTS OF CHILD PORNOGRAPHY
by Diana E.H. Russell, Ph.D.
Written September 2004

DO NOT QUOTE WITHOUT OBTAINING MY PERMISSION.

Introduction and Background: I sent this manuscript to my editor at Routledge Publishers in September 2004. Unfortunately, I was very late in meeting my deadline for many reasons, especially my intense involvement in initiating a campaign against the richest landlord in Berkeley (after UC Berkeley) for his 15-year exploitation of minor sexual slaves imported from India -- among other crimes. My editor had just quit her position at Routledge to accept a job at the New York University Press, so my manuscript remained unread while Routledge searched for a new editor. By January of 2005, a new editor had still not been hired, so I wrote to Mary McGinnis, the Vice President of Routledge to ask her what I should do about this. I feared if I revised the manuscript before I had an editor, she might well request that I revise it yet again. Ms. McGinnis told me to go ahead and revise it. Meanwhile, she said she would also like to see my manuscript. Less than a week later, she called me and declared that, "There is no way that Routledge will be associated with a book of this nature." "Why not?" I asked her. "It's the branding issue," she replied. "What do you mean?" I asked her, but she didn't explain. I interpreted her statement to mean that she didn't want Routledge to become known for publishing such a shocking book. It included many sexually explicit child pornography stories written for pedophiles, as well as descriptions of child pornography, including gross cartoons, all of which were legal. I told Ms. McGinnis that I was willing to remove the material that bothered her, but she insisted that she had discussed the issue with members of the staff, and she wasn't willing to reconsider. She resolutely held to her position despite my continued pleas. Since my manuscript was late, I had broken our contract, so I knew she would have this excuse to disregard it. Of course, this wasn't a genuine concern for her, since she had told me that I should go ahead and revise my manuscript. She said that she would help me find another publisher for this book. However, she did not follow through on this promise. I contacted my previous editor at New York University Press to ask if she and this publishing house would be interested in publishing Stolen Innocence. It so happened that New York University Press had published the major social scientific book on child pornography in recently. So she said that there would be no interest in publishing a book that would be in competition with this volume. I considered suing Routledge for breach of contract, since the lateness issue was obviously not the real reason for refusing to give me a chance to revise my manuscript. A respected colleague advised me not to, because she believed this would make it next to impossible to find another publisher, as well as jeopardizing publishers' interest in future projects of mine.

Stolen Innocence: The Damaging Effects of Child Pornography
Chapter 8: Introduction to Russell's Theory of Pornography
as a Cause of Child Sexual Abuse

As previously mentioned, there is almost universal agreement among researchers that child pornography created by photographing children constitutes child sexual abuse. The hypothesis that I will describe, explicate, and attempt to substantiate in this and the following chapter is that a causal relationship exists between adult or juvenile males'* [*footnote: from this point on, the terms "adult man," "men," or "males" should be understood to include juvenile and adult males as potential or actual sexual perpetrators] exposure to child pornography -- including computer-generated, written and oral forms of it -- and their perpetration of child sexual victimization.

However, since there are still professionals and many more non-professionals who believe that exposure to pornography is cathartic, that is, it provides "a release of wishes, desires or drives such that they do not have to be acted on in reality" (Kelly et al., 1995, p. 23), it is important to evaluate this theory before launching into my very contrary theory.

Catharsis vs. Intensified Desire

According to the catharsis theory, the repeated exposure of males to pornography "leads to a steadily decreasing interest" in the material (Bart and Jozsa, 1980, p. 210). This theory is frequently described as the "safety valve" theory. Applying it to child pornography, it assumes that repeated viewing of child pornography would lower the desire for sex with children of pedophiles and other consumers of this material. Hence, according to this theory, watching child pornography would diminish the prevalence of child sexual abuse.

In evaluating the validity of the catharsis theory, an old but very influential experiment by Howard, Reifler, and Liptzin (1971) will be described. This poorly designed experiment has been widely quoted as proof of the validity of the catharsis theory (Diamond, p. 199).

Howard et al.'s (1971) experiment was based on a small sample of 23 white college males and nine comparable controls. The researchers exposed the subjects and the controls to a pornographic movie, then tested both these groups for their sexual arousal to this material. The subjects were then exposed to adult pornography for 90 minutes a day for 15 days, whereas the controls viewed two non-pornographic movies over the same period. Twenty of the 23 subjects were then shown a third pornographic movie.

Howard et al., (1971) found that "all of the subjects reported initial stimulation by the pornography," followed by "a marked decrease in interest in it as a result of the exposure" to large amounts of the same kind of pornography (cited by Osanka and Johann, 1989, pp. 173-174). Many pro-pornography researchers have cited Howard et al.'s interpretation of their flawed experiment as if it had provided solid evidence that pornography is harmless because it initially causes sexual excitement but on repeated viewing interest in acting out the sex acts portrayed dissipates rather than stimulating or intensifying this interest.

Zillmann and Bryant conducted an experiment that showed why Howard et al.'s conclusion is totally invalid. These researchers gave the male subjects a greater range of pornography to view than the limited material available to the subjects in Howard et al.'s experiment. Zillmann and Bryant found that the subjects' boredom after repeatedly looking at the same pornographic material motivated them to switch to viewing different and more extreme pornography, such as sadomasochism and bestiality (cited by Osanka and Johann, 1989, p. 175). Zillmann and Bryant go on to say that the sexual brutalization of women holds "the greatest promise of exciting men" (Ibid.). Howard et al., had failed to consider this possibility, resulting in their flawed methodological design. Osanka and Johann cite other research that, like Howard et al.'s, draws the same fallacious conclusions based on similarly flawed experimental designs, e.g., O'Donohue and Geer (1985) and Schaefer and Colgan (1977).

Although masturbation is not mentioned in the experiments of Howard et al. and Zillmann and Bryant, it seems likely that male viewers' acts of masturbation (a major goal of pornography) to the pornography they were watching, disipated their interest until the next time they became sexually aroused. Would they always be satisfied to masturbate to this material? Clearly, many would not because most males consider masturbation a very inferior alternative to sex with the type of individuals they desire. Furthermore, the ejaculatory pleasure obtained from masturbation would intensify the association between it and the pornography viewed. Hence, the catharsis theory is not substantiated by the experimental research. (see Summers and Check, 1987).

Research aside, probably very few people would support a proposal to solve the problem of parents physically abusing their children by watching movies that show parents battering and torturing their children, and many of the children enjoying this abusive behavior. Likewise, perhaps no one would favor movies showing African American women and men being lynched as a way of diminishing racial discrimination. So why do so many individuals -- including researchers -- believe that only in the case of male misogynistic pornography that exposure to it dissipates the problem. Since believing such an inconsistency is totally irrational, this alone is sufficient "proof" that pornography is no safety valve.

Men's Propensity to Sexually Victimize Children

Because it is important to know the proclivities and the state of mind of those who read, hear, and view child pornography, I will start by describing some of the research on men's propensity to sexually victimize children. This question is of prime importance. If, for example, only one percent of men have a proclivity to victimize children sexually, whether or not child pornography intensifies this interest or causes these men to act out their proclivities, would be an issue of relatively little importance. If, on the contrary, a large percentage of men have such a proclivity, then ascertaining whether or not men's exposure to child or adult pornography promotes their acting out these proclivities is vitally important.

Some of the studies to be described below focus on the extent of male desire for sex with children. Sexual interest in children is a critical prerequisite for acting out child sexual victimization in most instances.* [*Footnote: I say "most instances" because there are undoubtedly some cases in which males sexually victimize children for other reasons. For example, in South Africa some African males rape baby girls because they believe this act can cure AIDS. In this AIDS era, some males sexually assault young girls rather than adult women in many countries because they believe they are more likely to be virgins -- and therefore AIDS-free -- or at least that girls will have had less exposure than older females to this fatal disease.

The Research of Kurt Freund and Kevin Howells

While some clinicians (e.g., Wyre), law enforcement officers (e.g., LA cop), as well as the public at large, consider all perpetrators of child sexual abuse to be pedophiles, most researchers -- including myself -- do not subscribe to this view. Nor does Kurt Freund (1981) who contends that

Freund noted elsewhere that "Pedo- and hebephilic [male] patients* [*Footnote: see the definition of hebephilics in the Appendix] [presumably non-patients as well] are ... often married [to women] but ... in the course of time intercourse tends to become conspicuously infrequent" with their spouses (p. 165).

Although the research of both Freund and Howells (to be described shortly) is very old, it remains very salient and still relevant to the issue of men's propensity to sexually victimize children (more recent research will be cited later). Before describing Freund's illuminating experimental findings, it is important to know that he defines a pedophile as an individual who has a

Freund (1981) used phallometric testing in his experiments to diagnose what he called "true pedo- or hebephilia" as opposed to "normal" heterosexual males by measuring their level of sexual arousal when viewing pictures of naked girls and boys. He chose this measure as the best because he considered a diagnosis of pedophilia or non-pedophilic child molesters to be impossible from "a person's known sexual history alone" (p. 162). [However, he also acknowledged that there is a "problem of response suppression or feigning of spurious responses" with phallometric testing (p. 162).]

In one experiment Freund (1981) assessed the penile volume changes of so-called normal heterosexual males on viewing "colour slides and movies of nude females and males of various ages" (p. 162). The results showed that although the "normal" heterosexual males showed a larger penile response to adult females than to children," (Howells, 1981, p. 79), "children have some arousal value even for normal males" (Freund, 1981, p. 137). In addition, Freund found "that normal heterosexual males respond even to very young girls substantially more than to males of any age group" (pp. 161/2).

The term "normal" as a descriptor of heterosexual males is placed in quotes because use of this concept assumes that the sexuality and personalities of heterosexual males are not greatly distorted by the patriarchal character of the societies in which this term is used. For example, research by Malamuth, Briere, Check and others has shown that it is common for males in the United States and Canada to acknowledge some likelihood of raping women if they could be assured that they would get away with it. More specifically, Briere, Malamuth and Ceniti (1981) reported that 60% of a sample of 356 male college students indicated that there was some likelihood that they would rape and/or force "a female to do something [sexual] that she really didn't want to do" if they would not be punished for it and if no one would know about it. On the basis of this research, mens' proclivity to rape/force a woman to have intercourse/sex could be considered normal for men in the United States -- given their socialization, their exposure to the patriarchal culture, etc.

Similarly, many other studies have shown that high school boys feel they would be entitled to rape females under many different circumstances. However, to consider this sense of entitlement as normal for heterosexual males would be exceedingly sexist. Although all contemporary societies are patriarchal to varying degrees, there are also significant cultural differences regarding what sexual behavior is considered "normal" for heterosexual males. For example, in India, despite being illegal, it is considered acceptable in some regions for old heterosexual males to marry and have intercourse with very young girls. The same is true among the Masai of Central Africa. However, most other countries have an age of consent for females of 16 years or older. In many non-Western societies it is believed to be "normal" to have more than one wife, sometimes many more. Hence, quotes will be used for the term "normal" when this term is not used in actual quotes by other researchers, etc.

Kevin Howells (1981), who defined pedophiles as "persons with a dominant and sustained sexual interest in children" (p. 62), maintained that

Howells' situational offender, who "prefers adult partners" and only "becomes involved with a child when there is 'some challenge to his sexual adequacy or threat to his sense of competency as a man'" (p. 78; this unreferenced quote comes from Nicholas Groth) is the equivalent of Freund's category of "normal" heterosexual male. Elsewhere, Howell's describes situational offenders as those "whose behaviour is precipitated by unusual life circumstances" (p. 62), or whose behavior is "situationally induced" (p. 67), or as "those of a normal orientation" (p. 77), as compared with "sexual preference induced pedophilic behaviour" (p. 67).

Howells cites Swanson's examples of important situational factors predisposing some "normal" males to select a child "as a sex object" who serves "as a substitute for an adult woman" (p. 77), for example, "marital disruption, loss of sexual partner through the wife's illness or work requirements, the use of alcohol, and multiple life stresses" (p. 77). He also quotes Groth's examples of situational factors or stressful precipitating events as "physical, social, sexual, marital, financial and vocational crises to which the offender fails to adapt" (p. 78).

The notion of "normal" situational sexual perpetrators against children markedly contrasts with the view that all such males are pedophiles. Indeed, Howells maintains that "There is good reason to think that such persons [pedophiles] form a minority in the total population of people who become sexually involved with children" (p. 76; emphasis added). Although Howells' cites other researchers (e.g. Mohr et al., 1964; Swanson, 1968) who have come to a similar conclusion (p. 77), he does not explain what the "good reason" is.

The Research of John Briere and Martha Runtz

Psychologists John Briere and Martha Runtz (1989) undertook a provocative study in an attempt to determine the percentage of "pedophiles" (their application of two definitions will be described below) in a sample of 193 undergraduate males. These researchers recruited male students in class for a study on "sexual attitudes" (p. 66). The students were assured of complete anonymity and confidentiality, and were also informed that they could discontinue their participation in the study at any time.

Briere and Runtz's information about their methodology was very inadequate. For example, they failed to describe the percentage of the class who refused to participate in their study, and the possible impact of the refusal rate on their findings. Nor did they mention the class description, the class status of the students, their demographic characteristics, and the limited generalizeability of the study. [Footnote: Unfortunately, this failure is the norm in the reporting of many studies by psychologists.] Bearing these limitations in mind, Briere and Runtz (1989) reported the following findings in response to the four questions they asked the students' about their sexual interest in children:

1. Just over one fifth (21%) of the male undergraduates "admitted to at least some sexual attraction to some small children" (p. 71).

2. "9% reported fantasies about sex with a child" (p. 71).

3. Over half of the students who reported fantasizing about sex with a child (5% of the total sample) "stated that they had masturbated at least once to such fantasies" (p. 71).

4. Seven percent "indicated [that there was] at least some likelihood of having sex with a child were it possible to do so without detection or punishment" (p. 71).* [*Footnote: This figure is substantially lower than the percentage Malamuth obtained in a study using the same or similar question. He related that 10-15% of male students reported some likelihood of sexually abusing a child if they could be sure of getting away with it (Malamuth, personal communication, July 1986).]

When Briere and Runtz (1989) applied David Finkelhor's very broad definition of pedophilia requiring that "the adult has had some sexual contact with a child" or that "the adult has masturbated to sexual fantasies involving children," they estimated that at least 5% of the university males in their sample were pedophiles (p. 71). When they applied an even broader definition of pedophilia requiring only that the students have at least some sexual attraction to children -- the figure for pedophilia in Briere and Runtz' sample rose to 20% (p. 71).

Briere and Runtz (1989) hypothesized that, "given the probable social undesirability of such admissions ... the actual rates of pedophilic interest in this sample were even higher" (p. 71). They maintained that their findings support Vernon Quinsey's conclusion, "based on a review of anthropological and historical data" that adults sexual behaviors with children are ubiquitous (p. 71).

Despite the very broad use of the term pedophilia, Briere and Runtz's study confirms Freund and Howells' findings that a significant percentage of so-called "normal" heterosexual males have some sexual interest in children. These researchers conclude with good reason that:

"The current data offer strong support for the notion that male sexual interest in children is relatively common in our society, even among "normal" (non-incarcerated and nonclinical) males" [i.e. non-pedophiles -- by the terminology Freund, Howells and I favor]. (p. 7).

Also significant is the fact that Briere and Runtz's question asked about sexual attraction to "small children" -- a phrase that suggests pre-pubescent rather than adolescent children. Had they asked their sample of male students about their sexual attraction to adolescent children at least five years younger than themselves, the percentage presumably would have been very much higher. Since we live in a culture that eroticizes teenage girls (for example, film star Brooke Shields was described at the age of 12 years as the most beautiful woman in the world), a good case can be made for the notion that only a small percentage of men experience no sexual attraction to children. Indeed, after defining hebephiles as "persons particularly attracted to postpubescent children (adolescents)," Dietz and Sears (1987-1988) comment:

"Whether such attraction ought to be regarded as abnormal is a debatable point, for attraction to sexually mature members of the opposite sex of the same species is biologically normal. It is not even clear that our culture, which proscribes sexual activity with those below the age of consent, condemns sexual attraction to these persons." (p. 28, fn. 47)

Hence, Briere and Runtz's 20% figure for males who admitted at least some attraction to small children would be infinitely higher if it included males who have some attraction to adolescent females under the age of 18.

The Research of Claude Crépault and Marcel Courtoure

Crepault and Courtoure (1980) undertook an experiment in which they studied Canadian mens' erotic fantasies. Their sample consisted of 94 French-speaking adult male subjects aged between 20 and 45, who resided in the province of Quebec and "who had lived with a woman for at least a year" (p. 266). They recruited their subjects -- 61% percent of whom were married -- by placing posters in public places and "advertising the research in a major French language newspaper" (p. 567). They obtained their data by administering two-hour long semi-structured interviews and self-administered questionnaires about the subjects' erotic fantasies.

Crepault and Courtoure reported finding that slightly more than three fifths (61.7%) of these men imagined a "scene where you sexually initiate a young girl" (p. 571), and 3.2% imagined a "scene where you sexually initiate a young boy" (p. 572). Although the non-random method of subject recruitment renders it impermissible to generalize these figures to a wider population, these researchers' finding that over 60 percent of the men had such erotic fantasies about young girls is surprisingly high. Perhaps the wording of the advertisements and posters encouraged males with such fantasies to respond (the authors provide no information on this wording). However, it is not clear why subjects who responded to posters and an advertisement would be more prone to have erotic fantasies about young girls than men of comparable age in the general population [check article for evidence of a comparison group].

The Research of Philip Jenkins

Philip Jenkins (2001) also emphasizes that "a sexual interest in children is not confined to a tiny segment of hard-core ... 'pedophiles'" (p. 25), and rejects the notion that there is "a fundamental gulf dividing 'child-lovers' (sic) from 'normal' people" (p. 27).

Referring to the sizeable legal market in pseudo-child pornography in which adult women masquerade as young teens (Jenkins, p. 27) on adult sites titled "lolitas" or "child porn" (p. 29). Jenkins infers that "The popularity of such materials indicates a mass popular market for teen sexuality" in the United States (p. 28). Jenkins is struck by the significance of pornography merchants assuming "that a substantial audience would be interested in something that notionally lies so far beyond the pale" (p. 30). From these observations he infers "that those interested in child pornography might not be so far removed from the 'normal' population" (p. 30).

Pornography researcher Gail Dines takes issue with Jenkins' assumption that the sizeable popular interest in pseudo-child pornography is only a reflection of a mass popular market for teen sexuality. She maintains that the mass media also constructs the sizeable male interest in this market (Personal communication, Friday, March x).

Research Report by the Office of the United Nations High Commissioner for Refugees (UNHCR) and Save the Children-UK

A report released in February 2002 by UNHCR and Save the Children charity in Britain documents the "rampant" sexual victimization of children in refugee camps in Guinea, Liberia and Sierra Leone in exchange for money and food by "U.N. staff, security forces, staff of international and national NGOs, government officials and community leaders" (p. 3). This report was based on "interviews and group sessions conducted with 1,500 adults and children" (UNHCR and Save the Children-UK, 2001, p. 2).

The interviews revealed that "the very people [men] who are meant to be providing services, are the exploiters themselves," stated Jane Gibril, a senior official with Save the Children charity (Maharaj, 2002, p. A9). Journalist Davan Maharaj (2002) notes that "The report ... is replete with accounts by children who said they were forced to have sex with relief workers to get basic humanitarian aid" (p. A9). "Children who refused to have sex were sent to the back of food lines (p. A9). The assessment team reported that childrens' involvement "in sexually exploitative relationships has become a mechanism for survival for many refugee families" (p. 8).

According to Maharaj (2002), 40 humanitarian aid groups were implicated in creating "cesspools of sexual exploitation of children" (p. A9). More specifically, the assessment team reported that

The majority of children involved in "the exchange of sex for money or gifts" were "girls between the ages of 13 and 18 years" (p. 3). In addition, "girls between the ages of four and 12 were also reported as being sexually harassed, either verbally or through touching of buttocks, breasts, or genitals" (p. 12). The opinion of many of the men who were interviewed -- "including agency workers and community leaders" -- was "that younger girls are more desirable as sexual partners.... Some also believed that sex with a virgin could cleanse a man from infection" (p. 4; emphasis added).

According to the report, some peacekeepers "were alleged to have had sex with the girls without using condoms" (p. 6). To what extent the peacekeepers' contributed to the "50% pregnancy rates of all the teenage girls in the camps" was not addressed. Most of these girls "had become mothers between the ages of 13 and 16" (p. 10). Most of them found that "their families and care-givers reject[ed] them when they become pregnant" (p. 10). Furthermore, "abortion is illegal" in Guinea, Liberia and Sierra Leone; it constitutes "a felony offence punishable with life imprisonment" (p. 10). Hence these girls become single mothers at a very young age with no means of supporting themselves and their child or children.

The irony is that many pedophiles do not treat their victims in the callous manner described in this report. This is a suggestive indicator that most of the predatory male peacekeepers probably qualify as "normal heterosexual men." On the other hand, since adult women and children were equally available for exploitation, it is difficult to understand why these so-called normal men would prefer to have sex with children.

A major shortcoming of the interview-based report is its failure to provide information on the numbers and percentages of preadolescent girls and teenagers who were sexually victimized by peacekeepers. Earlier we pointed out that some researchers consider adult males' sexual attraction to teenage girls to be normal. However, it is illegal to act out this attraction in the United States and many other countries. The peacekeepers who exploited teenage girls in Guinea, Liberia and Sierra Leone were in such relatively powerful positions over the impoverished semi-starving refugees, that they may have felt that they need not be constrained by any local laws. However, in the absence of more information on the ages of the children, this vital question cannot be answered.

Nevertheless, the assessment team's report corroborates the statements made by the researchers whose work was described above that many so-called normal non-pedophilic adult males sexually victimized children. Although there were presumably some pedophiles who participated in the sexual violation of children in these countries, it seems exceedingly unlikely that a disproportionate percentage of these sexual predators would have sought jobs helping the poverty-stricken refugees in the African countries mentioned above.

Finally, in many traditional societies studied by social anthropologists as well as in some ancient civilizations like Greece, it is, or was, customary for adult males, and sometimes mothers, to have sex with children. For example, Finkelhor (1984) notes that

Ford and Beach (1951) also report that in a few societies,

In contrast to most nations today, members of the Lepcha people clearly do not consider it abusive for adult males to have sex with young girls. Nor is the incestous behavior of the other pre-industrial societies cited by Ford and Beach considered a harmful breach of the incest taboo. Assuming that the biology of the males in these societies is the same as the biology of males in all other societies, these practices suggest that many "normal" males in the United States and elsewhere, may have a propensity to sexually abuse children.

The Meaning of "Cause"

Given that the major goal of Stolen Innocence is to demonstrate that the consumption of child pornography plays a causal role in child sexual victimization, it is vital to define the term "cause."

George Theodorson and Achilles Theodorson (1979) differentiate between the term "simple causation" and "multiple causation." They define simple causation as

An event (or events) that precedes and results in the occurrence of another event. Whenever the first event (the cause) occurs, the second event (the effect) necessarily or inevitably follows. Moreover, in simple causation the second event does not occur unless the first event has occurred. Thus the cause is both the SUFFICIENT CONDITION and the NECESSARY CONDITION for the occurrence of the effect (p. 40).

By this definition, the consumption of child pornography clearly does not cause child sexual victimization because some pornography consumers do not sexually victimize children and because many incidents of child sexual victimization are unrelated to child pornography. However, Theodorson and Theodorson's (1979) concept of multiple causation (defined below) is applicable to the relationship between the consumption of child pornography and child sexual victimization.

With the conception of MULTIPLE CAUSATION, various possible causes may be seen for a given event, any one of which may be a sufficient but not necessary condition for the occurrence of the effect, or a necessary but not sufficient condition. In the case of multiple causation, then, the given effect may occur in the absence of all but one of the possible sufficient but not necessary causes; and, conversely, the given effect would not follow the occurrence of some but not all of the various necessary but not sufficient causes (p. 40).

As I have already presented considerable research on males' proclivity to sexually victimize children, I will next discuss the theories relating to the causes of those proclivities.

Causes of Males' Proclivity to Sexually Victimize Children

There are many factors that contribute to the causation of child sexual victimization. Adults' exposure to child pornographpy is only one of them. I will not attempt in this book to evaluate the relative merits of different causal factors, but merely to present evidence that exposure to child pornography is a significant one.

Because not all adult male consumers of child pornography are equally negatively affected by it, some people conclude that this is evidence that merely consuming this material cannot be playing a causal role in child sexual victimization. This is similar to the tobacco industry's claim that since many smokers do not die of lung cancer, this proves that smoking does not cause lung cancer. But this reasoning is faulty since there are no grounds for assuming that the proponents of smoking as a cause of lung cancer believe that smoking is the only cause. In addition, the tobacco industry's fallacious defense is based on their focus on individual rather than group differences. In contrast, the proponents of smoking as a cause of lung cancer focus on the statistically significant number of smokers who get lung cancer compared with non-smokers.

Whereas the individual level of analysis is more relevant for psychologists and clinicians, the group level of analysis is more relevant to social policy makers. Although it is important for psychologists to try to explain individual differences, this information is not needed to determine the impact of pornography or smoking on individuals.

David Finkelhor's theory re: causes of child sexual victimization

Sociologist David Finkelhor (1981) developed a very useful multicausal theory to explain the occurrence of child sexual victimization. According to his model, four preconditions must be met in order for child sexual victimization by an adult to occur. First, "the adult must have sexual feelings for a child" (p. 2). Second, the adult's internal inhibitions against acting out his sexual feeling(s) must be overcome. Third, the adult's social inhibitions against acting out his sexual feelings (e.g., fear of being caught and punished) must be overcome. Fourth, the adult must overcome the resistance or attempts at avoidance by the child, if these occur (p. 2).

According to Finkelhor's theory, "(T)he presence of all four prior conditions" must be met to explain the occurrence of child sexual victimization (p. 9; emphasis added). In addition, he maintained that there is a logical sequence to these four preconditions:

Only some individuals have sexual feelings about children. Of those that do, only some overcome their internal inhibitions to act on these feelings. Of those who overcome their internal inhibitions, only some overcome external inhibitions -- the surveillance of other family members or the lack of opportunity -- and act on the feelings." (p. 8)

Finally, the potential child victim's resistance to the sexual victimization must be overcome.

Finkelhor's formulation of the fourth precondition is problematic since, as he himself acknowledges, "sexual abuse can occur in such a way that resistance by the child is irrelevant to whether the abuse occurs" (p. 7). For example, perpetrators can make a surprise assault on an unsuspecting child, and/or they can also use force. In such cases, this precondition is not necessary for child sexual victimization to occur.

Whereas Finkelhor's model relates to the causes of child sexual abuse in general, my three-factor theory to be presented in the next chapter applies to only one cause: the exposure of adult males to child pornography is a significant cause of child victimization. As will become evident shortly, my theory draws heavily on Finkelhor's model.\

Finkelhor and Jones' Research on the Decline in Child Sexual Cases

Finkelhor and Jones (2003, March) note that, "The number of sexual abuse cases substantiated by child protective service (CPS) agencies dropped a remarkable 40 percent between 1992 and 2000, from an estimated 150,000 cases to 89,500 cases" (p. 1). In order to shed light on the causal factors involved in this decline, these researchers undertook a detailed evaluation of "the strengths and weaknesses of six possible explanations for the decline by using data from a number of different sources" (p. 1). Finkelhor and Jones conclude that, "Evidence from a number of different sources, including NCVS [National Crime Victimization Surveys] data showing a 56-percent decline in self-reported sexual assault against juveniles, is consistent with a real decline in sexual abuse" (p. 2).

Because my theory (explicated in the next chapter) shows a causal relationship between exposure to child pornography and child sexual abuse, and because there is a consensus among researchers that there has been an enormous increase in child pornography on the Internet, many individuals are likely to interpret Finkelhor and Jones' conclusion that there has been a real decline in child sexual abuse as evidence that my theory must be incorrect. Therefore it is important for me to evaluate the validity of this interpretation.

The Validity of a Decline in Child Sexual Abuse

I and co-author Rebecca Bolen (2000) undertook a thorough analysis of the methodology of the National Crime Victimization Surveys -- one of the studies on which Finkelhor and Jones base their analysis and conclusion that there has been a real decline in child sexual abuse between 1992 and 2000. While Bolen and I focused on NCVS's findings about the national incidence of rape in the United States, Finkelhor and Jones excluded all cases of juvenile rape, focusing instead on NCVS' data on sexual assault. The NCVS for 1994 defined sexual assault as follows:

"A wide range of victimizations, separate from rape or attempted rape. These crimes include attacks or attempted attacks generally involving unwanted sexual contact between victim and offender. Sexual assaults may or may not involve force and include such things as grabbing or fondling. Sexual assault also includes verbal threats." (p. 149)

Given that Finkelhor and Jones' report was designed to evaluate the possible explanations for the decline in child sexual abuse, the NCVS's definition of sexual assault is an inappropriate data base to include. Not only is the NCVS' definition seriously flawed as a definition of sexual assault, but it does not even purport to serve as a definition of child sexual abuse. The term child sexual abuse typically includes a great range of abuses from non-contact abuses (e.g., witnessing genital exposure) or relatively mild forms of contact abuse (e.g., touching of sexual body parts) at one end of the continuum to rape at the other end. In contrast, the NCVS' definition of sexual assault excludes rape and attempted rape, but includes verbal threats. No other incidence or prevalence study of child sexual abuse excludes rape and attempted rape and includes verbal threats -- for good reason. It makes no sense!

Moreover, the NCVS surveys are limited to children aged 12 and older. Furthermore, the two youngest age groups in which their age data are categorized are 12- to 15-year-olds and 16- to 19-year-olds. Clearly, 18- and 19-year-olds are not children. Finkelhor and Jones fail to mention these serious methodological problems with the NCV Surveys and how they dealt with them. Nor do they mention these kind of methodological problems in any of the other studies on which they base their analysis and conclusions about the decline in the number of substantiated incidents of child sexual abuse.

Rebecca Bolen and I (2000) undertook a thorough evaluation of the methodology of the NCVS surveys efforts to provide national data on the incidence of both reported and unreported rape (see Chapter 9). We concluded with the statement that,

"Currently, the NCVS's massive underestimate of the incidence of rape is not only useless, but dangerous, because it misinforms the public regarding the magnitude of the rape problem. We would actually be far better off without these surveys. Hence, unless their methodology is repeatedly revised until it yields more realistic assessments of the incidence of rape, it would be preferable for them to discontinue their questions on rape." (p. 85)

For reasons explained above, it appears that our conclusion is even more appropriate with regard to use of the NCVS surveys as a basis for estimating the national incidence of child sexual abuse.

Aside from one other interview-based study that is limited to the state of Minnesota, Jones and Finkelhor (2001) base their analysis of "the trends in reported and substantiated cases of child sexual abuse ... on data from child protective service CPS) agencies" (p. 10). These studies include aggregate data from the National Child Abuse and Neglect Data Systems (NCANDS) and "more detailed child protective service data from Illinois, Minnesota, Oregon, and Pennsylvania" (pp. 1-2). Hence, it is important to understand the limitations of the CPS data for measuring the incidence of child sexual abuse. Following is a description by Jones and Finkelhor (2001) of how the CPS agencies obtain their data.

"All States have laws that require different classes of professionals to report any suspicion of child maltreatment. Reports by these professionals or by other community members come to the agencies' attention locally or through a centralized State intake process. CPS agencies investigate those reports that seem valid and fall within their jurisdiction, evaluate the level of evidence that maltreatment has occurred, and decide whether or not to substantiate the reports.".... (p. 10)

Data stored in the CPS administrative system represent only those cases of maltreatment that come to CPS attention. Many cases of child maltreatment never dome to official attention at all. Furthermore, most States limit the role of CPS agencies to cases of maltreatment by caretakers. This would include investigating reports of abuse by parents and guardians, babysitters, teachers, and other individuals who share responsibility for caring for the child.... Despite capturing only a certain proportion of cases, CPS data are the best source for analyzing national trends in child maltreatment because aggregate information is available on an annual basis." (p. 10)

Child sexual abuse is one form of "child maltreatment" in addition to physical abuse and neglect.

Bolen and I (2000) made the following criticisms of researchers who use CPS data as a basis for ascertaining the incidence of child sexual abuse:

1. Narrow definition: Because the CPS' definition of child sexual abuse limits these acts to those that are perpetrated or permitted by a parent or caretaker, most experiences of child sexual abuse do not qualify as such; for example: sexual abuse by siblings, neighbors, and acquaintances who are not in a caretaking role, as well as sexual abuse by strangers, gangs, or peers, and "even numerous seductions by adults or adolescent acquaintances of the child where the parents were not neglectful" (Finkelhor, 1994, p. 35). Furthermore, several studies have found that sexual abuse by nonrelatives is much more prevalent than sexual abuse by relatives (e.g.,......). Given the exceedingly narrow definition of child sexual abuse, it comes as no surprise to find such a low incidence of child sexual abuse.

2. Reported cases biased: Incidence estimates based on reported cases of child sexual abuse are inherently biased because they are limited to suspected cases of child sexual abuse and the small minority of children who choose to disclose their abuse experience(s) to someone who then has to report it to the appropriate authorities. It is well-known that few cases are reported than actually occur.

3. Only substantiated cases: Until recently, only cases that have been substantiated or indicated (i.e., there is a very strong suspicion that they are valid, but they do not quite meet the substantiation guidelines) are included in the official incidence statistics. Finkelhor and Jones note that it is now necessary for the CPS to decide between only two categories: substantiated and unsubstantiated. Presumably, this has resulted in a lowering of the incidence of substantiated cases.

The impact of the false memory syndrome movement in the 1990s has likely decreased the number of reports of child sexual abuse to CPS agencies as well as increasing the number of cases they unsubstantiate.

4. Comparison with FBI statistics on rape: Although the FBI statistics on rape are typically described as based on cases reported to the police, the statistics are in fact based only on the number of rapes that are not "unfounded". "Unfounding" refers to "the percentage of complaints determined through investigation to be false" (UCR, 1993, p. 24). The criteria used by the police to determine which complaints are false or baseless are not explained in the Uniform Crime Reports.

Both the FBI's statistics on founded cases of rape and CPS's statistics on substantiated cases of child sexual abuse grossly underestimate the incidence of these crimes, making them appear relatively rare. Both founded cases of rape and substantiated cases of child sexual abuse are equivalent to mere tips of icebergs that are highly unrepresentative of the vast numbers of unreported cases.

5. Excluded cases of child sexual abuse: In addition to the cases of child sexual abuse that we have already noted as missing from the CPS statistics on substantiated cases, there are several other types of sexual abuse that are largely or completely missing. For example:

a. Pornography-related child sexual abuse

b. Child sexual abuse of child prostitutes

c. Sexual slavery: Domestic and international trafficking

d. Recently disclosed cases in the Catholic church

Conclusion: The statistics quoted by Finkelhor and Jones as indicating a decline in child sexual abuse cases between 1992 and 2000 are based on such deficient data that their observation cannot be taken at face value. Furthermore, their attempt to evaluate several different explanations for why these defective figures show a decline, are all based on similarly defective data. Therefore their conclusion that there has been a real decline in the incidence of child sexual abuse must be discounted.

Home | About Diana Russell | Pornography As a Cause of Rape (book excerpt) | Publications | Other links |