~     THE PREVALENCE OF PAEDOPHILIA     ~

"What, though, of those adults who experience a sustained preference for children as sexual partners, but who never -- whether from fear or 

[Page 28]

moral scruple -- touch a child with erotic intent, finding more or less satisfying outlets in fantasy, or in 'substitute' activities varying from masturbation to marriage? 

It is impossible to know how many adults fall into this category, but on the speculative assumption that they constitute no more than 1% of the population, their total numbers would approach half a million ... 

I am taking it for granted that feelings of sexual attraction in adults towards at least some young people from about 15 upwards are so common as to be virtually universal, and that similar feelings for boys or girls in the years immediately following puberty are experienced by rather more adults than most of them are prepared to acknowledge." 
(Righton, 1981, pp. 24-25) 

There is no doubt that substantial numbers of underage people are exposed to unwanted or unwarranted sexual experiences with adults. The only uncertainty is in specifying the exact number. The precise figures depend on decisions, such as whether non-contact abuse is included and how and where the research is carried out. Nevertheless, tentative but useful data is available (La Fontaine, 1990). 

In contrast, little is known about paedophilia in the community rather than in institutional settings. Simple information such as the numbers of active paedophiles in the general population is unavailable. More complex questions such as the number of men with unexpressed paedophile leanings are impossible to discuss knowledgeably. 

Research by Temoleman and Stinnett (1991) would be dismissed as trivial were more substantial data available. Nevertheless, it gives some indication of the extent of sexual interest in underage people. Undergraduate volunteers averaging 22 years of age from largely white rural areas or small towns took part. Their sexual histories were assessed using the Clarke Sexual History Questionnaire supplemented by additional questions. 

Illegal sexual behaviour was fairly common: 

In all, two-thirds of the sample had done at least one of these things and over half had committed an arrestable office. Very few of them, just 3%, reported being a victim of sexual abuse as a child.

[Page 29]

Their sexual arousal to different types of sex and age group was measured using the Farrenkopf Arousal Portfolio and Cardsort. The portfolio consists of separate pictorial collages of 0-6-year-olds, 6-12-year-olds, 12-18-year-olds and adults, as well as violence towards women, exhibitionism, voyeurism, dressing in female underwear and bestiality. Sexual arousal to each collage was rated by the men on a scale from 0 to 100. The cardsort involves a set of cards describing sexually explicit activities with a third person: rape, exhibitionism, voyeurism, frottage, obscene phone calls, and sex with boys, girls, adolescents and adults. 

Although fairly low levels of arousability were found on average to child stimuli, sexual arousal to depictions of teenagers was considerably higher and to adult women very high.  
While this is significant, other data are more revealing: 

Such data are at best suggestive because of the poor sampling methods used. 

Finkelhor and Lewis's (1988) epidemiological study of child molestation ought to have provided far better evidence. They used a sample of 2500 Americans over 18 years of age contacted by random telephone dialling. Because of the strong risk that people would be unwilling to disclose over the telephone that they had sexually abused a child, Finkelhor and Lewis used a variation of the Randomized Response Technique. This involved putting pairs of questions to the respondent. For example, some were asked to answer "yes" or "no" to either member of the following pair, but not both: 

The respondent chooses which one of the pair to answer at random I (e.g. by tossing a coin). Since the interviewer does not know which of the two questions the respondent is answering, the respondent may feel free to answer truthfully. 

This is not the pointless exercise that it would at first appear, since so long as the researcher knows 

statistical formulae can calculate the proportion of people who have sexually abused a child. Of course, this is based on probabilities, so it is never known whether a particular individual was an abuser or not.

[Page 30]

This sounds fine in theory -- though perhaps a little statistically complex. A difficulty arises because the sample was divided into two, and slightly different pairs of questions were asked. This was to enable the percentages answering "yes" to the "neutral" question when presented on its own to be found -- each sample answered directly the other sample's "neutral" randomized question. 

So, the second half of the sample were posed the following pair of questions: 

Within the margins of statistical variability, the two alternative pairings of the sexual abuse question should generate identical rates of abusing. In fact, radically different percentages of abusers were obtained -- one estimate was 17% sexual abusers in the general population, the other 4%! 

Finkelhor and Lewis rather implausibly explain this by suggesting that the so-called "neutral" question about union membership may have been rather more sensitive than had been intended compared with that about renting. Unfortunately, this post hoc argument is not convincing. For example, the two "neutral" questions showed equal rates of refusal to answer when asked as direct questions individually, and so do not appear to differ in sensitivity. 

Finkelhor and Lewis found no relationships between admission of abusing and family background or prior childhood victimization. Since previous research has found such relationships, this may cast more doubt on the method's validity. 

These problems notwithstanding, it is unclear what Finkelhor and Lewis mean by sexual abuse. Phrased as the abuse question is, possibly some respondents included their childhood sexual activities with other children. 

Discussion of the issue of the non-sexually active paedophile similarly lacks relevant information. One commentator, Righton (1981), discussed 57 clients who were attracted to boys. About a third of them had never made any sexual contact with boys, although sometimes such thoughts filled their fantasy. It cannot be assumed from a clinical sample that a similar proportion of non-offending paedophiles could be found in the general population. 

The possibility of passive paedophilia is rarely raised other than in individual case studies (Hurry, 1990). 

Some support, albeit rather tentative, for the view that paedophilia is relatively uncommon, can be found in research on the distribution 

[Page 31]

of paedophilia related material available from pornography retailers. Although this cannot directly reveal the extent of paedophilia, it may indicate the relative interest in paedophiliac material compared with other types. 

Lebegue (1991) examined pornographic titles collected in the USA during the mid-1980s. Using the standard psychiatric diagnosis scheme (DSM-111-R), the titles were classified according to 
the paraphilias they indicated. Any references to schoolgirls were included in the paedophilia category. Partialism (i.e. an obsession with a particular body part such as breasts or bottoms) is not a diagnostic category in DSM-111-R. 

Similar results were found in Australia (Lebegue, 1985); 

In this latter study the incest and paedophilia categories were combined but amounted to less than 4%. 

Of course none of this is definitive proof of the small scale of active paedophiliac interest in the population; it merely suggests that such activity is relatively uncommon. Quite clearly, the question of the extent of paedophiliac interest in the population is a vexed one. It would be of considerably theoretical and practical interest to know how many men keep their paedophiliac urges under control and do not express them through offending. As it stands, virtually all that we know comes from offenders.