part0012

Meta-analytic reviews of experimental studies of pornography have found that exposure to mainstream adult pornography has a medium-size significant effect on negative attitudes about women and on physical aggression in the laboratory among nonoffenders (Allen, D’Alessio, & Brezgel, 1995; Allen, Emmers, Gebhardt, & Giery, 1995); this effect is also observed for adult pornography containing violent content. There is also evidence that frequency of pornography use and likelihood of engaging in sexual coercion are positively correlated in samples of nonincarcerated men (e.g., Malamuth, Addison, & Koss, 2000). Allen, D’Alessio, and Emmers-Sommer (1999) conducted a meta-analysis of 13 studies and concluded that sex offenders did not significantly differ from nonoffenders in their age at first exposure to pornography or frequency of pornography use. However, there was some evidence that sex offenders were more likely to report engaging in sexual behavior— masturbation, consensual intercourse, or sexual offending—after using pornography. There are no experimental data on the effects of exposure to child pornography, so the impact of exposure to child pornography is not known.

The experimental literature on the effects of adult pornography and the effects of depictions of violence in television programs, film, and video games suggest that exposure to child pornography is likely to have a negative effect on subsequent behavior, especially among individuals who seek out such content and are already predisposed to act on their sexual interests. It is highly unlikely that viewing child pornography has a cathartic effect and thereby reduces the likelihood that an individual will commit a sexual offense against a child. A cathartic effect of child pornography would not be consistent with evidence regarding the impact of sexually explicit media (see Malamuth et al., 2000; Seto, Maric, & Barbaree, 2001). Thus, men who view child pornography are likely to be pedophiles and may be more likely to exhibit an effect of exposure to child pornography on their positive attitudes about sex with children and on the likelihood of engaging in sexual behavior with a child.

There are three methodological limitations to the studies of adult pornography use and sexual offending that were reviewed by Allen et al. (1999): (a) They are retrospective and therefore might be subject to recall biases, (b) they are based on self-report and therefore are vulnerable to social desirability, and (c) they do not usually distinguish between types of pornography. Most studies have focused on mainstream pornography (e.g., nude photographs in magazines such as Playboy or Penthouse or sexually explicit depictions of intercourse available through photographs, videos, and digital media). It is possible that sex offenders do not differ from controls in their overall pornography use but differ in their use of pornography that corresponds to their particular preferences. For example, Nutter and Kearns (1993) reported that their group of 25 nonincarcerated child molesters did not differ from volunteers in their use of pornography, but the researchers only asked questions about mainstream, commercially available content and did not ask about use of child pornography. W. L. Marshall (1988) found that a higher proportion of sex offenders against children than nonoffenders reported frequent use of pornography; however, only 7 of the 51 sex offenders against children admitted they had ever looked at child pornography, and only 3 admitted they collected child pornography. One third of the sex offenders against children reported that viewing pornography was a precursor of a sexual offense, but they reported that this pornography usually depicted adults. Moreover, sex offenders against children reported that they could begin sexually fantasizing about children simply by seeing children, without any use of pornography. Wheeler (1997) surveyed 150 sex offenders with child victims who were participating in outpatient treatment. Most (93%) of the sex offenders with child victims admitted to sexual fantasies involving children, typically girls. One third of the sex offenders with child victims claimed that they used pornography before committing an offense, and similarly, one third claimed that pornography had influenced their subsequent offending. They admitted using child pornography, but this use was infrequent, and most of the content consisted of nude children rather than explicit depictions of sexual activities. At the same time, the sex offenders against children were more likely to report using adult mainstream pornography than a comparison group of 122 other offenders.

Finally, Langevin and Curnoe (2004) found that sex offenders against children (21%) were more likely to use pornography than rapists (8%) as part of their sexual offenses; offenders against unrelated children (26%) were more likely to use pornography than incest offenders (17%). Among the sex offenders against children who used pornography, approximately one half showed pornography to their victims during the offenses. Approximately one half of the pornography they showed was commercially available adult pornography, but 6 of the offenders showed child pornography, and some of the other offenders took photographs of their child victims. Of the 33 sex offenders against children who took photographs of their victims, 9 subsequently distributed them to others.

Depictions of sex with children have existed for a long time; for example, Tannahill (1980) described a Chinese sex manual from the 1400s called The Admirable Discourses of the Plain Girl . In the 20th century, prior to the emergence of the Internet and other digital technologies, child pornography offenders would have had to contact individuals or commercial distributors to purchase or trade materials, or they would have had to produce their own content. Each contact or production might have led to an arrest by law enforcement officials (see Jenkins, 2001; Lloyd, 1976; O’Brien, 1983). Moreover, the content that was available consisted of photographs, magazines, books, or films that would eventually deteriorate with age. Because child pornography involved physical media, tougher law enforcement could have a substantial impact on the availability of child pornography, as it did in the 1970s and the 1980s (Jenkins, 2001). In the past 15 years, however, the production and distribution of child pornography has become much easier and cheaper with the decreasing cost of digital cameras and related technologies. These digital images rarely deteriorate (files can become corrupted as they are copied) and can continue to be distributed long after the children in the images have become adults. The relative difficulty in obtaining child pornography before the widespread availability of the Internet may have been reflected by Howitt (1995), who interviewed 11 adult male pedophiles who reported that they infrequently used child pornography. Instead, they created their own pornographic materials using images from catalogues, magazines, and other freely available and legal sources.

Child pornography makes up only a small percentage of the total pornography content available on the Internet, but given the total number of images and video clips that are available, this still represents hundreds of thousands of unique images and hundreds, if not thousands, of unique video clips (see Jenkins, 2001; Taylor & Quayle, 2003). 4 This estimate does not take text or audio files into account. Mehta (2001) examined 9,800 randomly selected images from 32 Usenet newsgroups between July 1995 and July 1996 and found that 4% depicted prepubescent children and 11% depicted pubescent children. Mehta, Best, and Poon (2002) examined 507 randomly selected film clips obtained through the a peer-to-peer distribution network and found that 4% depicted children in sexual situations. This can be compared with Lebegue’s (1991) analysis of over 3,000 pornographic magazine and book titles; less than 1% of the titles suggested pedophilic content (titles from the 1970s or 1980s include Nymph Lovers , Lollitots , and Joyboy ), and another 4% of the titles suggested depictions of incest.

Taylor et al. (2001) have proposed that “collections of child pornography are not accidental; they result from deliberate choices by an individual to acquire sexual material” (p. 99). In their article, Taylor et al. described a database of more than 80,000 child pornography images and 400 video clips that had been collected by the COPINE (Combating Paedophile Information Networks in Europe) project. This database, now part of Interpol’s Child Abuse Image Database (Noble, 2007), consisted of pictures known to be at least 15 years old (e.g., scans from child pornography magazines) and images downloaded daily from more than 60 newsgroups that sometimes distributed child pornography (averaging 1,000 images per week and images depicting two previously unposted children per month). More than one half of the child pornography images were of girls. Among the more recent images in the database, 41% of girls and 56% of boys were judged by the researchers to be between the ages of 9 and 12; the rest of the depicted children were judged to be younger. 5 Although age is an imperfect indicator of pubertal status, as discussed in chapter 1 of this volume, the judged ages of the children in the COPINE database suggest that a majority of the images depict prepubescent children (especially those judged to be below the age of 9). Taylor et al. (2001) reported anecdotally that child pornography offenders interviewed by COPINE investigators said they preferred thin, fair-complexioned children whose genitals were clearly visible and who did not show secondary sexual characteristics.

Police investigations have also revealed information about child pornography content. For example, Operation Cathedral, a large international police investigation of a child pornography ring called the Wonderland Club, resulted in the arrests of 107 men in 14 different countries (see L. Kelly & Regan, 2000). The participants in this international ring were high-volume traders in child pornography, with the club requiring an initial contribution of 10,000 images to join. At its peak, the Wonderland Club had an estimated 200 members. Almost a million images of 1,263 children were seized, and all of the children were estimated to be under the age of 16; some images were of infants or toddlers.

Finally, approximately four fifths of the child pornography offenders in Wolak et al. (2005) had images of children between the ages of 6 and 12, two fifths had images of children between the ages of 3 and 5, and one fifth had images of children under the age of 3. Many offenders had images of both prepubescent children and adolescents, but 17% had pictures exclusively of children ages 12 and younger. Four fifths of the possessors had images depicting the sexual penetration of children, and approximately one fifth had images depicting violent content such as bondage, rape, and torture.