‘Any male victim?’ The legacy of conflating paedophilia and homosexuality and the differential seriousness of sexual abuse by victim gender

Abstract

This article presents the findings of a historical, qualitative thematic analysis of archival clinical records of 24 men convicted of child sexual abuse and referred for civil commitment to the Massachusetts Treatment Center between 1959 and 1984. Drawing on the perspective of Cicourel, the study examined the differential ways men convicted of child sexual abuse were constructed by various criminal justice actors based on the gender of their victims. Overall, the study found that men with male victims were constructed as more deviant than men with female victims. The study’s findings thus contribute to the research literature on the historical pathologisation and criminalisation of homosexuality.

Keywords:

• Any male victim
• child sex offenders
• child sexual abuse
• child sexual abusers
• civil commitment
• differential treatment
• homosexuality
• paedophilia
• sex offender
• victim gender crossover

Introduction

Homosexuality was once heavily pathologised and considered a sexual perversion and mental disorder (akin to paedophilia), which required sex offender-specific therapeutic intervention (Bernstein, 2004; Conrad & Angell, 2004; Foucault, 1976/1990; Freedman, 1987; Krafft-Ebing, 1886; Woods, 2015). In an infamous study conducted during this time, Ellis and Brancale (1956) concluded that homosexual men (including those who engaged in consensual sexual acts with other males) were ‘truly deviant’, the riskiest and most likely to reoffend. Recent literature on paedophilic offenders, as well as the current Diagnostic and Statistical Manual of Mental Disorders (DSM–5), still commonly distinguish paedophilia based on victim gender preference (American Psychiatric Association, APA, 2013; Blanchard et al., 2021; Blanchard et al., 2009; Fedoroff et al., 2013; Hall & Hall, 2007; Seto, 2012, 2017). While homosexuality is now regarded as neither a psychiatric condition nor a crime in most countries, the lasting effects of the conflation between the two endure today (Bernstein, 2004; McDonald, 2016).

Child sexual abuse is a gendered social issue that is predominantly perpetrated by adult men, against female children and young people (Ioannou et al., 2017; Lippert et al., 2009; Stoltenborgh et al., 2011). Over 50 years of empirical research on men convicted of sexual offences against children has demonstrated that having male victims increases the risk of sexual recidivism (Abel et al., 1987; Hanson & Bussiere, 1998; A. J. R. Harris & Hanson, 2004; Helmus & Thornton, 2015; Långström & Grann, 2000; Prentky et al., 1997; Sandbukt et al., 2021; Vess & Skelton, 2010). Hanson and Bussiere’s (1998) influential meta-analysis found that sexual deviancy (e.g. operationalised as ‘the perpetrator has boy victims’) was the best predictive factor for sexual recidivism in child sexual abusers. Similar findings have been reported throughout the literature, with research consistently finding that those who sexually abuse boys have higher rates of recidivism. This has resulted in the inclusion of ‘any male victims’ as an item on popular actuarial risk assessment tools (e.g. Beech et al., 2003; Hanson, 1997; Static-99 (Hanson & Thornton, 2000); A. J. R. Harris & Hanson, 2004; Risk Matrix 2000 (Thornton et al., 2003); Vess & Skelton, 2010).

It was assumed previously that over the offending careers of child sexual abusers, victim preference remained relatively consistent, which is important for assessing risk and classification level (Abel et al., 1987; Gebhard, 1965; Guay et al., 2001; Lamade & Prentky, 2019; Soothill et al., 2000). However, numerous studies exploring offender ‘crossover’ have established that men who sexually abuse children may not have specific or exclusive age, gender or ‘relationship’ victim preferences, and are known to abuse children outside of those previously assumed preferences (Bradford et al., 1988; Cann et al., 2007; Heil et al., 2003; Heil & Simons, 2008; Kleban et al., 2013; Levenson et al., 2008; Stephens et al., 2018; Wilcox et al., 2005). Victim gender crossover in samples of child sexual abusers in clinical settings, across several countries, is estimated to occur in 10–15% of perpetrators (Bradford et al., 1988; Heil et al., 2003; Kleban et al., 2013; Scurich & Gongola, 2021; Weinrott & Saylor, 1991). A systematic review conducted by Scurich and Gongola (2021) examined the prevalence of crossover among 1371 men who had sexually abused children and found the rate of victim gender crossover to be between 7.1% and 33.5%. However, studies using self-report information have estimated victim gender crossover to be between 36% and 53% (Heil et al., 2003; Weinrott & Saylor, 1991). In an Australian sample of child sexual abusers, Vess and Skelton (2010) reported that 17% of those men previously known to abuse only female children abused male children in a follow-up offence, and 10% previously known to only abuse boys abused girls in a follow-up offence.

Child sexual abusers who do not have any clear victim gender preference have been found to have the highest risk of sexually reoffending (Cann et al., 2007; Levenson et al., 2008). This makes sense, as lacking strict gender preferences increases the overall pool of potential victims and thus opportunities to engage in child sexual abuse (Beauregard et al., 2012; Heil et al., 2003).

Against this backdrop, the present study examined the historical case files of 24 men convicted of sexual offences against children: 10 men who were known to have exclusively female victims and 14 men who were known to have exclusively male victims. The groups were compared to explore the extent to which victim gender appeared to influence criminal justice actors’ construction of their sexual and psychological dangerousness. The aim of this work is to contribute to a better conceptual understanding of perpetrators of child sexual abuse and their construction by a range of criminal justice actors (e.g. clinicians and correctional officers).

Conceptual orientation

To address this aim, we draw on the perspective developed by Cicourel (1968), which focuses attention on how criminal justice actors (e.g. police, clinicians, therapists, social workers, psychologists and psychiatrists) define forms of ‘deviant’ behaviour. This perspective represented an advance on existing work (especially by Merton, 1957) that highlighted the unreliability of official crime statistics. Merton (1957) and other sociologists had already demonstrated that official crime statistics were unreliable since they only reflected those crimes brought to the attention of police and other criminal justice authorities. For Cicourel, however, the point was not merely that official statistics were ‘unreliable’ or masked a ‘dark figure of crime’ but that we should direct our attention to the sociological fact of this unreliability (Kitsuse & Cicourel, 1963). According to this perspective, official statistics and other records of crime and deviance are a better reflection of the views and behaviours of those defining and counting deviance than they are of offending or deviant behaviour itself (Nelken, 2020). Subsequently, similar behaviours resulted in some individuals but not others being defined as deviant (Kitsuse & Cicourel, 1963). Rates of deviant behaviour must, according to this perspective, be viewed as social artefacts:

 . . . rates of deviant behavior are produced by the actions taken by persons in the social system which define, classify and record certain behaviors as deviant. If a given form of behavior is not interpreted as deviant by such persons it would not appear as a unit in whatever set of rates we may attempt to explain. . . . From this point of view, deviant behavior is behavior which is organizationally defined, processed, and treated as “strange,” “abnormal,” “theft,” “delinquent,” etc., by the personnel in the social system which has produced the rate. (Kitsuse & Cicourel, 1963, p. 135; italics in original)

Critically, Cicourel (1968) draws our attention to descriptions recorded by professionals in case files that seek to justify professionals’ decisions about individual clients. According to this perspective, ‘reports of decision-making often provide evidence not of the way that it was but of the way that it had to be’ (Nelken, 2020, p. 541). This article thus focuses on the views and behaviours of those defining deviance in the criminal justice system and those tasked with managing and treating that deviance at the Massachusetts Treatment Centre (MTC).

Participants and data source

The present study examined the redacted archival, clinical records of men who had been referred to the MTC for assessment. Individuals referred to the MTC were subject to 60 days of observation and were ultimately either determined to be a sexually dangerous person (SDP) according to the sexual psychopath legislation (and thus civilly committed and detained at the MTC until they were no longer considered sexually dangerous) or determined not to be an SDP (and thus either released or returned to prison; Freedman, 1987; Prentky et al., 1997). The case files contained extensive histories of the men, including detailed police reports, clinical notes and reports by treating psychiatrists and psychologists (including diagnosis and progress), court documents and other correspondence. The initial sample from which the current sample was drawn consisted of 812 men referred for treatment to the MTC between 1959 and 1984. This facility was established in response to the sex crime panic of the time and provided an assessment of anyone referred for civil commitment and treatment and anyone determined to be an SDP (Freedman, 1987; Prentky et al., 1997). Most of the men were white (89.6%), and of those who disclosed their religious denomination, the majority identified as either Catholic (59.3%) or Protestant (34.4%). Interested readers are referred to D. A. Harris (2008) for more detail.

There were 341 individuals whose sexual offences involved children exclusively. The mean age of child sexual abusers at the time of their index offence was 32.5 years (range = 15.7–70.8 years, SD = 10.6). Victim information was only available for 227 perpetrators, with a median number of three victims each (range = 1–18). Descriptive statistics for victim gender indicated there were 123 men whose offences involved at least one male victim in their crimes (54.2%) and 104 men whose offences did not have any male victims (45.8%).

The sample

A purposive sampling technique was used to narrow the sample of 227 child sexual abusers for whom victim gender preference was recorded at the MTC to those who: (a) had at least three child victims and (b) had extrafamilial victims exclusively. As repeat child sexual abusers with extrafamilial victims are known to have high rates of recidivism (Doren, 1998; Petrunick & Deutschmann, 2008), limiting the sample according to these two variables allowed us to minimise variation across the two groups and to control for key potentially confounding variables. This process ultimately resulted in a total of 24 men whose records were available and who had offended exclusively against 3+ extrafamilial, child victims. Fourteen men were known to have only abused boys (men with male victims, hereafter MMV), and 10 were known to have abused only girls (men with female victims, hereafter MFV).

Descriptive statistics were calculated on the two groups (MMV and MFV) to ensure their similarity on a range of variables (e.g. perpetrator age, number of victims, age of victims; see Appendix, Table A1). The mean age of the 10 MFVs was 35.8 years (range = 21–53 years) at the time of their referral to the facility, and they were primarily white (80%) and Catholic (60%).

The collective total number of officially recorded female victims was 59, making the average number of reported victims per individual 5.9 (range = 3–8 victims). Four files indicated at least some self-reported (not officially recorded) or alleged (not substantiated) additional victims. According to available records, these 10 men admitted to an estimated cumulative total of 63 additional unreported victims (average of 6.3 victims per perpetrator; range = 0–40 children). The average age of female child victims was 7.3 years old (range = 3–16 years). Half of the MFVs (n = 5) were officially diagnosed with paedophilia, and 70% (n = 7) were ultimately determined to be SDP and civilly committed indefinitely at the MTC.

There were 14 men with convictions for offences exclusively against multiple extrafamilial prepubescent boys. The mean age of this group of men was 29.9 years (range = 18–64 years), and they were all white and primarily Protestant (57.1%; seen in Appendix, Table A2).

Although the average age of MMVs was younger than that of MFVs, the difference was not statistically significant. The total number of officially reported victims for MMVs was 85 male victims, with a mean of 6.1 victims per participant. This is comparable to the MFV’s average number of 5.7 victims mentioned above. According to the case files, there was an estimated total of 61 unreported child victims for the MMVs (average of 4.3 per perpetrator). The average age of their male victims was 8.9 years old (range = 4–16 years). Of the 14 MMVs, 57.1% (n = 8) were diagnosed with paedophilia, and 78.6% (n = 11) were deemed SDP and were civilly committed to treatment.

Analytical strategy

As this project draws on previously collected, de-identified data that received ethical approval (see D. A. Harris, 2008, for more detail), Griffith University Human Research Ethics committee determined that this study was exempt from requiring ethical approval. The redacted and deidentified files of the 14 MMVs and 10 MFVs were thematically analysed (Braun & Clarke, 2006, 2012). The process of thematic analysis is rooted in grounded theory (Glaser & Strauss, 1967) and includes a six-step process (data familiarisation, generation of initial codes, searching for initial themes, reviewing themes, theme defining and final report; Braun & Clarke, 2006, 2012).

Consistent with Braun and Clarke’s (2012) approach, the initial stage of thematic analysis began with data familiarisation. Data familiarisation was achieved by the first and second authors reading and reviewing each case file. This yielded an initial 33 codes using an inductive strategy. Examples included ‘victim blaming’, ‘sexist language’, and ‘lack of violence’. The 33 codes were then arranged and sorted by similarity, leading to the identification of three interrelated key themes: reification of injury to males; amplification of victim status of males; and relative dangerousness of the perpetrator. These are discussed in turn below.

Results

This section presents the results from the thematic analysis, with findings arranged into three themes.

Reification of injury to male victims

The extent to which the victim’s injury or harm was denied (minimised) or confirmed (upheld) was a key theme identified in the archival files. This theme concerned how language was used in police reports, court transcripts and clinical assessments to describe the seriousness of the offence and injury to the victim. What is noteworthy here is that report writers not only reified injury to males but also minimised injury to girls. For example, in the case below, the report writer emphasised the absence of violence and used the word ‘merely’ to minimise the seriousness of an indecent assault on a very young female child:

The present offense involved a 5 ½ year old girl, and again there was no violence. He merely touched her private and then exposed himself and masturbated until he reached a climax. (F7)

Even when other variables (such as age of victims and relationship to victims) remained constant, the injury to female victims was minimised. For example:

It might be added that though he did commit sex offences and involved children during his past episodes, there, nevertheless, is no evidence that he is prone to violence or that he even hurt anyone. (F9)

In contrast, when boys were abused, the injury was more likely to be affirmed, and the offences were considered far more devastating for the victim. Boys were generally portrayed as victims of psychological and physical harm and were frequently depicted as having been seriously harmed. For example, the language that was used to describe the impact on the subsequent psychological development of the following boy is clear:

These acts have a harmful, emotional impact on the young mind. The injuries are of a type you can’t see. Important is the potential effect upon a child of immature and ugly sex stimulation and induction. (M10)

Descriptions of male victimisation also carried insinuations of more violence, force and serious impacts than the victimisation that girls experienced. For example:

One of the victims was verbally threatened and he pushed another in the stomach. Many of the victims, reportedly 30 [in number], came from broken homes. It is felt that he is a sexually dangerous [person] because of the psychotic process and also the evidence of aggression. (M3)

He maintained that he would never hurt anybody, although he did employ some force when they [male victims] refused by going on and kissing them anyway. (M2)

This is consistent with the historical belief that men’s sexual abuse of boys – or ‘homosexual abuse’ as it was sometimes termed (Wright, 2007) – was aberrant and thus more damaging to the victim than the abuse of girls, which had been viewed as so ubiquitous that it was unremarkable (Wright, 2007; see more broadly Davies, 2002).

Amplification of victim status of males

Another theme to be extracted from the thematic analysis was how the child victim was either downplayed or affirmed (sustained), depending on their gender. Boys were frequently portrayed as particularly vulnerable victims, while girls were often seen to have precipitated their abuse in some way. Relatedly, victim-blaming by criminal justice officials was also much more common for MFVs than for MMVs. Two main indicators evidenced this theme. First, it was evident in how the victims (or the circumstances of the offences) were described in the police reports. Second, it was apparent in the way the men themselves were depicted (in the clinical assessments written by the facility staff).

Female victims were almost universally portrayed to have more sexual agency than boys of the same age. Below are excerpts from two of the case files in which young girls are depicted as being much more mature than their chronological age:

It is reported that the alleged victim [age 13] of the offence . . . unzipped her clothes and was behaving seductively. She was 13, but she said that she was 20. [The police] record states that she unzipped her clothes, and according to him, was behaving extremely seductively. (F8)

[One of the victims (13)] is very matured and developed for her years. She looks more like a woman than a child. . . . [F6] said that the girl aged 9 belonged to a ‘sex ring’ and that she seduced him in order to extort money from him. (F6)

Describing female victims in a way that makes them appear much older than they are is evident with the use of words such as ‘seductive’ or ‘mature’. In many of the MFV cases, at least some of the responsibility for the abuse is attributed to the victim. This was even observed in cases involving much younger victims. For example, a 42-year-old man’s repeated abuse of a 9-year-old girl was described as follows:

[Victim] admitted that on at least six separate occasions she did accompany [F6] to his room and permit him to make sex advances on her. (F6)

This accords with literature that shows that the public (O’Hara, 2012) and judicial decision-makers (Wright, 2007) construct female victims of child sexual abuse in similar terms. The linguistic differences were perhaps most pronounced in the descriptions of the offences. For example, for one MFV, the medical examinations of four young girls (one of whom was only 4 years old) described medically confirmed physical evidence of sexual assault in the following way:

A doctor’s report showed four of the little girls had physical signs consistent with sex relations. (F1)

It is undeniable that today, such physical indications would ideally be described as incontrovertible proof of penetrative child sexual abuse. But in the MFV cases analysed, it is clear that sexual assault of female children as young as four was minimised and described as ‘having sexual relations’.

The language used in the MMV reports implied enhanced deviance and criminality on the part of perpetrators. For example, the repeated use of the word ‘assault’ (rather than ‘relations’ or ‘advances’) much more forcefully implies a lack of consent by the victim. Further, other words such as ‘sodomy’, ‘pervert/se’, and ‘aberrant’ were commonly used. The perception of the greater seriousness of offences against male victims is evident in the MMVs below:

One cannot minimise or overlook however, the factor of emotional violence of despoiling children, of indoctrinating susceptible children, of perverting immature minds, and of causing his victims to become a danger to other children. (M10)

He indoctrinated and seduced young susceptible minds through his vile, despicable practices. This man has shown no sense of decency, no remorse, no regard for the rights and welfare of others. (M11)

Male victimisation was frequently seen as more damaging, with clinicians and police officers fearing that the boy victims (but not girl victims) would be ‘indoctrinated’ into becoming sexual offenders themselves. There were also fears about male victims becoming homosexual because of same-sex victimisation with strong support for the ‘abused–abuser’ hypothesis (Garland & Dougher, 1990):

 . . . [the victim] is now a sexual pervert as a result of [subject’s] actions with him. . . . (M10)

Many of these children were around seven and eight years old and had been quite ignorant of such homosexual behaviour. Furthermore, he [police officer] feels that some of these children-victims have been conditioned to such behaviour and might result in there [sic] becoming homosexuals. (M3)

Inferences to the ‘infectious’ nature of homosexuality in MMVs pertained to apparent interfering with the preconceived heteronormative expectations of children’s sexuality by altering the direction of the otherwise assumed future heterosexual citizen (Robinson & Davies, 2008). The assumed process of indoctrinating male victims to become sexual predators and homosexuals is consistent with the historical construction of the ‘folk devil’ homosexual paedophile (Robinson, 1970). Research examining the contemporary constructions of sexual harm against children has understood child sexual abuse as a future-oriented and gendered construct (Small, 2019). Attorneys conceptualised girls who had been sexually abused as children as in a chronic state of vulnerability, where child sexual abuse was understood to lead to promiscuity as an adult (Small, 2019), whereas boys who had been sexually abused were conceptualised to be in a chronic state of aggression, where sexual abuse was understood to lead to sexual deviancy and sexual offending as an adult (Small, 2019). Ultimately, the future potential heteronormative otherwise lifestyle of children, especially boys who have been sexually abused, was at risk (Small, 2019). The conflation of paedophilia and homosexuality further exacerbated the perceived threat that ‘homosexual paedophiles’ posed to children (Bruhm & Hurley, 2004; Douard & Schultz, 2013; Jenkins, 2004; Robinson & Davies, 2008; Smaal, 2013).

The dangerousness of MMV was further emphasised by the differential perceived believability of male and female victims. This is especially clear in the two men whose files included crimes against both girls and boys (M2 and M6). In both these cases, although the men admitted to offending both male and female children, it is particularly instructive to note that charges were only pursued for offences that were perpetrated against boys. (To be clear, these men were considered MMVs for this analysis because they were officially convicted of crimes with exclusively male victims. Only after their archival files were reviewed did their self-reported victimisation of girls become known.) Male victims were generally more likely to be believed by police, and police appeared to be more sceptical of female victimisation. This was most notable in the language used by police in their reports. For example, below, a boy’s ‘testimony’ is compared with a girl’s ‘claim’. Girls’ accounts were frequently peppered with equivocating words such as ‘claim’ or ‘accused’ or ‘alleged’, which challenged the victim’s credibility.

Girls claim subject asked them if they wanted to. . . . (F10)

One victim testified that he was forced to commit the acts of sodomy. (M3)

At times, male victims would even request (and be granted) the opportunity not to talk about some aspects of the offence because it was causing them discomfort. Such relief was never granted to any female children in the MFV files. This reflects the well-documented disbelief of female complainants in the criminal justice system, including the emblematic ‘corroboration requirements’ contained in common law, which historically held that in sexual assault trials, a conviction should not be made without evidence corroborating the accounts of women and girls (Ben-David, 2019; Sheehy, 2002). Again, following Cicourel (1968), constructions of female unbelievability are used to justify decision-making about the required criminal justice response to perpetrators.

Dangerousness of the perpetrator

The third theme identified concerned the relative dangerousness of and risk posed by the perpetrator. MMVs were often constructed as especially risky and warranting extensive treatment for their inherent deviance compared to MFVs.

It was evident that the patient does not have proper controls in the sexual area. He is in need of extensive psychiatric treatment on an in-patient basis. (M8)

MMVs were seen as ‘doubly deviant’ because of the intersectionality of their sexuality (homosexuality) and sexual deviance (paedophilia). At the time, this elevated the seriousness of their offence and thus the perceived dangerousness of the perpetrator. This is unsurprising given the well-documented pathologisation of homosexuality (Conrad & Angell, 2004; Foucault, 1976/1990), which was only removed as a mental disorder from the DSM in 1973 (Silverstein, 2009). By way of example, M10 and F8 are juxtaposed below. With a comparable number of similar-aged victims, both men’s offences are described as ‘repetitious’. Whereas the MMV (M10) is ‘sexually dangerous’, is ‘odd, erratic’ and demonstrates ‘abnormal behaviour’, the MFV (F8) is merely a ‘nuisance’ who should not be committed for treatment:

He has long been a frank homosexual. Sexually dangerous, and odd erratic person with a propensity to engage in sexual misbehaviour with young boys. He has a repetitious abnormal behaviour. (M10)

The patient is a repetitious, nuisance offender (pedophile) and it is the general consensus that he is not committable. (F8)

It is essential to note the conflation of paedophilia and homosexuality in many case files. Homosexuality was often used interchangeably with paedophilia, making it difficult to determine whether the individual was being diagnosed solely with paedophilia or was a gay man who happened to abuse children. It was clear that MMVs were consistently labelled as homosexual, sexually deviant, predatory, vile, perverse and abnormal, whereas MFVs were instead considered to be merely immature or nuisance offenders, which frequently translated into an assumption of them being less risky and dangerous. In Cicourel’s (1968) terms, the construction of heightened deviance (and thus dangerousness) of MMVs was used by clinicians to justify their protracted treatment requirements and indefinite detention. As noted above, nearly 79% of MMVs were deemed SDPs, compared with only 70% of MFVs. While this is perhaps a less pronounced difference than we might expect based on the clinicians’ narratives, it nonetheless indicates the increased dangerousness attributed to MMVs.

Discussion

This study utilised a qualitative approach that explored whether victim gender could usefully differentiate how men convicted of child sexual abuse were historically constructed by criminal justice system actors. The study examined how victim gender influenced treatment objectives and clinical assessments of the sexual and psychological dangerousness of men who had sexually abused children. Despite the similarities between the demographics and offence characteristics of MMVs and MFVs (and noting the small sample size), the thematic analysis revealed many differences in how the men were constructed, assessed and processed by criminal justice actors and by the therapeutic practitioners at the time. The differences were most evident in the language used to describe the offences, the offenders and their victims. The frequent conflation of paedophilia and homosexuality led to an increase in the perceived level of severity, sexual deviancy and therefore the risk of recidivism for men who had abused male children.

Our qualitative results indicate that MMVs’ offending behaviours were considered to be the most serious, and the most sexually deviant, and they were assumed to be at a higher risk of sexually reoffending. The way MMVs were socially constructed is reminiscent of the predatory paedophile, perceived to be particularly deviant, emerging out of the sex offender panics throughout the 1970s (Best, 1993; Levine, 2006; Small, 2015). Small (2015) examined the legal construction of sexual offenders among practising attorneys who, in their experiences, primarily drew on the notion of the socially constructed predatory paedophile contingent on the age and gender of the perpetrator and victim.

Comparatively, MFVs were deemed immature, nuisance offenders whose offences were less severe and less damaging to their victims. Their treatment recommendations frequently indicated a need to ‘find someone a bit older’ and assigning at least some of the responsibility for the offence to the precocious or provocative behaviours of the female victim was not uncommon. Considering female children are consistently reported to be sexually abused at higher rates than males (Barth et al., 2013; Moore et al., 2015; Stoltenborgh et al., 2011), the differences between the perceived seriousness with male victimisation seemingly normalised the sexual abuse of females by men and reinforced male sexual entitlement over female bodies regardless of their age. These findings are consistent with current research that has highlighted the minimisation and normalisation of female sexual victimisation compared to males (Armstrong et al., 2018; Banton & West, 2020; Chim et al., 2020).

Victim blaming manifested in several ways throughout the results and consisted of attacking the victim’s credibility, questioning their victimisation, and attributing blame to the victim by suggesting they acted in a sexually provocative manner that encouraged the offender. The extent of victim-blaming was considerable (see further Lonsway & Fitzgerald, 1994; Theimer & Hansen, 2020). The female child victims in the study were extensively blamed and treated as having more sexual agency than the male victims (of similar age), who were regarded as defenceless and unprovoking targets.

Similarly, this study found that treating clinicians reserved the harshest opinions, treatment and diagnostic prognoses for patients who victimised male children while attributing less blame to male victims. Therefore, it is highlighted by this study that sexual crimes involving male victimisation further contributed to narratives of rape myths, gender stereotypes and child sexual abuse misconceptions that influenced how men were treated in the criminal justice system.

Although we now rightly distinguish between homosexuality and paedophilia, it was evident at the time that there were inherent societal biases and perceptions of same-sex victimisation drawn through the conflation between homosexuality and paedophilia. The lasting effects of this conflation continue to stigmatise the queer community (Lancaster, 2011; McDonald, 2016; Meyers, 1965). For example, a study by Hunt and Moodie-Mills (2012) found that queer young people convicted of non-sexual offences were unfairly labelled as sex offenders and referred for sex offender treatment at higher rates than for heterosexual young people convicted of similar offences. These multiple factors combined create a particular subgroup of men, perceived as perilous, dangerous and ‘doubly deviant’, which has been portrayed through decades of heteronormative assumptions and expectations, and through the historical pathologisation and criminalisation of homosexuality.

Limitations

The findings from the current research should be viewed in light of some important limitations. First, although the participant criteria were intended to be as inclusive as possible, these criteria were not exhaustive. As a result, some offence and offender characteristics that were not included in the current set of cases could have an impact on criminal justice actors’ language and decision-making.

One could argue that the data source is quite dated. Certainly, that would be the case if we were attempting to generalise our findings to contemporary examples. On the contrary, we aimed in this article to explore the historical construction of male perpetrators of child sexual abuse according to their victim gender preference.

Conclusion

This article contributes to the literature on child sexual abuse by examining the differential way that child sexual abuse perpetrators have historically been constructed by criminal justice system actors based on the gender of their victims. In sum, the intersection of homosexuality, same-sex victimisation and paedophilia resulted in MMVs being seen as more serious and dangerous than MFVs.

Our work highlights, in line with Cicourel’s (1968) perspective, the inherent biases (in this case, homophobic biases) of those defining deviance. In doing so it contributes to an understanding of the historical construction of male perpetrators of child sexual abuse.

Ethical standards

Declaration of conflicts of interest

Madeleine Bennett has declared no conflicts of interest.

Danielle Arlanda Harris has declared no conflicts of interest.

Kelly Richards has declared no conflicts of interest.

Ryan Shields has declared no conflicts of interest.

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee [current Griffith University Human Research Ethics Committee and previous United States National Institute of Justice Institutional Review Board and San Jose State University Institutional Review Board] and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Informed consent

Informed consent was not obtained from all individual participants included in the study under the grant of ethical approval by the San Jose State University Institutional Review Board, which concluded it was impracticable to receive informed consent from the participants at the time the data were created (between 1954 and 1989).

Appendix

Table A1. Demographics of men with multiple, exclusively female child victims (MFV).

Case	Age (years)	Ethnicity	Religion	Official victims	Unreported victims	Total victims	Sexually dangerous person
F1	49	White	Protestant	8	2	10	No
F2	34	White	Catholic	8	0	8	Yes
F3	26	White	Protestant	7	15	22	Yes
F4	53	White	Catholic	5	0	5	Yes
F5	37	White	Protestant	6^a	6	12	No
F6	43	White	Protestant	7	0	7	Yes
F7	24	White	Catholic	5	40	45	Yes
F8	29	Black	Catholic	3	0	3	No
F9	42	Mixed	Catholic	7	0	7	Yes
F10	21	White	Catholic	3	0	3	Yes
Total	Mean			59	63	122	7

^aMean substitution was used to approximate the number of unofficial victims in one case where ‘some’ victims were indicated, yet the exact number was unstated. Estimate was rounded to the nearest person.

Table A2. Demographics of men with multiple, exclusively male child victims (MMV).

Case	Age (years)	Ethnicity	Religion	Official victims	Unreported victims	Total victims	Sexually dangerous person
M 1	29	White	Catholic	7	4^a	11	No
M2	22	White	Protestant	3	4^a	7	No
M3	44	White	Protestant	6	30	36	Yes
M4	21	White	Protestant	3	1	4	Yes
M5	19	White	Catholic	4	2	6	Yes
M6	34	White	Protestant	9	2	11	Yes
M7	27	White	Protestant	6	3	9	No
M8	31	White	Catholic	4	15	19	Yes
M9	32	White	Catholic	6	0	6	Yes
M10	21	White	Protestant	11	0	11	Yes
M11	21	White	Catholic	9	0	9	Yes
M12	35	White	Protestant	4	0	4	Yes
M13	64	White	Protestant	12	0	12	Yes
M14	18	White	Catholic	6	0	6	Yes
Total				85	61	146	11

^aMean substitution was used to approximate the number of unofficial victims in two cases where ‘some’ victims were indicated, yet the exact number was unstated. Estimate was rounded to the nearest person.