The Motivation-Facilitation Model of Sexual Offending

Motivation

Consistent with dictionary definitions and with a consensus definition from Kleinginna and Kleinginna’s (1981) review, I also define motivation here as a psychological process that energizes and directs behavior; for example, a desire for sex is a motivation to engage in sexual behavior. Motivation for sex can influence perceptions, intentions, and other psychological phenomena, but is distinct from these other phenomena. Motivations to sexually offend can vary from offender to offender and offense to offense, but the MFM emphasizes three primary sexual motivations: (a) paraphilias, (b) high sex drive, and (c) intense mating effort.

Facilitation

Being motivated to sexually offend as a result of paraphilia, high sex drive, or intense mating effort is not sufficient in the MFM because—in line with the general theory of crime—someone who is high in self-control would be able to resist these motivations. In other words, even very strong motivations to sexually offend can be countered by sufficiently strong inhibitions. Self-control can vary both as a trait (across individuals) and as a state (within individuals); for example, within individuals, self-control might vary with mood, exposure to stressors, and other dynamic factors such as intoxication. Facilitation factors are defined as those factors that overcome any trait or state inhibitions against acting upon motivations (see Figure 1). In the following sections, I discuss self-regulation problems and hostile masculinity as two examples of trait facilitation factors, and negative affect and alcohol use as two examples of state facilitation factors.

Vulnerable Victims

There is a reluctance in the sexual offending literature to look closely at adult victim characteristics, perhaps out of concern that this might be seen as victim blaming or shifting of the responsibility from the perpetrator. There seems to be less resistance to looking at child vulnerabilities, perhaps because children are less likely to be viewed as culpable agents. In either case, I consider the identification of potential victim vulnerabilities to be important in terms of developing better prevention strategies. Vulnerable child victim characteristics include father absence or living in a blended family, because children are relatively at greater risk from step-father figures than from genetic fathers ¹ ; lower income households or other family stressors; and loneliness, social isolation, or rejection by peers (Finkelhor, 1980; Finkelhor, Ormrod, Turner, & Hamby, 2005; Holmes & Slap, 1998; Snyder, 2000). Vulnerable adult victim characteristics include being female, having a history of childhood or adolescent sexual abuse, and being under the influence of alcohol or drugs (Barnes, Noll, Putnam, & Trickett, 2009; Senn et al., 2015; Testa, Livingston, Vanzile-Tamsen, & Frone, 2003). These vulnerabilities represent both historical or trait factors (e.g., being female) and state factors (e.g., being under the influence of alcohol or drugs).

Presence of Guardians

This situational factor does not refer to the presence of guardians in the formal or legal sense, but in the sense of a responsible person who could intervene if a sexual offense is taking place (Leclerc, Smallbone, & Wortley, 2015). Within the family home, this could be a parent or (formal) guardian, an older sibling, or a coresiding adult; outside the family home, guardians could include other relatives, teachers or other professionals, older friends, and bystanders. The nature and number of potential guardians will change with the potential victim’s age and circumstances. Leclerc et al. (2015) found that the nearby presence of a guardian was associated with a shorter duration of child sexual abuse incidents and a lower likelihood of penetrative acts.

Time and Place

Sexual offenses are not randomly or uniformly distributed across time or space. Certain times and certain places are associated with greater risk. For example, McKillop, Brown, Wortley, and Smallbone (2015) found that sexual offenses against children were more likely to occur in the afternoon or early evening than at other times, and usually in the residence of the child or perpetrator, if they lived separately. This reflects the impact of routine activities, where children are less vulnerable to sexual offenses while at school, surrounded by other adults and children. Similarly, adult victims are more vulnerable at different times and places, for example, in the evenings, at their own residence or the perpetrator’s residence, or after gathering at parties or bars.

Much more work is needed to elucidate the kinds of situational factors that are most germane, and the way these situational factors interact with individual perpetrator (and potential victim) factors. Many sexual offense prevention strategies—walk home services, bystander interventions, supervision rules banning unsupervised contact with minors, and relapse prevention strategies—are predicated on the importance of situational factors in sexual offending. Individual differences in risk are targeted by perpetrator-focused interventions, such as treatment for at-risk persons or identified offenders, and in supervision models, such as the Circles of Support and Accountability, containment, or multi-agency protection plans (Beier et al., 2009; English, Jones, & Patrick, 2003; Hanson, Bourgon, Helmus, & Hodgson, 2009; Scott et al., 2006).

Strengths

The MFM incorporates evidence from community studies of self-identified sexually aggressive men, as well as clinical and forensic studies of identified sex offenders (e.g., Lalumière et al., 2005; Malamuth et al., 1995). The most distinctive feature about sex offenders is the prevalence of paraphilias, as other offenders usually score higher on antisociality (though some sex offenders can be quite high in antisociality, and sex offenders still score higher than nonoffending men, on average, after excluding criminal history variables). The MFM also connects to research on situational risk factors and work to identify psychologically meaningful risk factors (Mann, Hanson, & Thornton, 2010).

I suggested earlier that the MFM could be applied to other types of offenses, such as noncontact sexual offenses involving exhibitionism or voyeurism. Child pornography offending is a compelling test for the MFM because child pornography offenders are likely to have pedophilic sexual interests, yet many have not had any known sexual contact with children. The difference seems to involve facilitation, because child pornography offenders are lower on antisociality indicators such as psychopathy or criminal history than contact offenders (Babchishin, Hanson, & VanZuylen, 2015). Moreover, child pornography offenders who do have more antisociality (e.g., more prior criminal justice involvement) are more likely to sexually reoffend than those who are low in antisociality, whether by committing another child pornography offense or a contact sexual offense (Seto & Eke, 2015). This suggests that many child pornography offenders are individuals who are motivated to engage in sexual behavior involving children because they have pedophilic or hebephilic sexual interests, but they are high in self-control or low in facilitation factors and thus are unlikely to commit contact sexual offenses involving children. Their self-control is not sufficient, however, to inhibit acting on their pedophilic or hebephilic motivations by committing child pornography offenses. Consistent with rational choice and routine activity theories, child pornography offenders have more access to Internet technologies, but less access to children than contact offenders (Babchishin et al., 2015).

Weaknesses or Gaps

In the current description of the MFM, all motivations are sexual in nature, contrary to ideas that anger, revenge, desire for power or control, or other nonsexual motivations play key roles in sexual offending against adults (Pullman, Stephens, & Seto, 2016; Seto, 2008, 2013). This is an assumption that requires further empirical evaluation. Typologies developed to describe the heterogeneity of offenders against adults distinguish between sexual and nonsexual motivations (e.g., Barbaree, Seto, Serin, Amos, & Preston, 1994). For sexual offending against children, there has long been a distinction between sexual and ostensibly nonsexual motivations. Groth, Hobson, and Gary (1982) suggested there was a difference between those who were motivated by a sexual attraction to children (fixated) in contrast to an emotional affinity for children and/or distance from adults (regressed). Similarly, Finkelhor (1984) suggested some offenders are primarily motivated by a sexual attraction to children, whereas others are primarily motivated by their emotional congruence with children. I have described emotional congruence or affinity with children as ostensibly nonsexual because newer conceptualizations suggest that desire for erotic activity and desire for closeness or intimacy are facets of sexual desire (e.g., Chadwick, Burke, Goldey, Bell, & van Anders, 2017; van Anders, Goldey, & Kuo, 2011).

Consistent with feminist conceptualizations of male sexual violence toward women, and with early typological work on rape and rapists, hostile masculinity may represent a motivation rather than facilitation factor, or could both motivate and facilitate acting on other motivations (Brownmiller, 1975; Groth, 1979; but see Lalumière et al., 2005, and Palmer, 1988). For example, endorsement of rape myths and other offense-supportive beliefs about women and sex can be viewed as aspects of hostile masculinity. These attitudes and beliefs could both motivate (e.g., believing that women who wear revealing clothing are “asking for sex,” increasing sexual desire) and facilitate (e.g., believing that women are subordinate to men and should therefore acquiesce to men’s sexual demands) sexual offending. One way to test whether hostile masculinity can be a motivation for sexual offending, particularly offenses against women, is to see whether it is associated with sexual offending in the absence of paraphilia, high sex drive, or intense mating effort. In other words, would someone who is high only in hostile masculinity be at greater risk of committing sexual offenses, in conjunction with other facilitation factors such as negative affect and alcohol use?

A related and important question is whether ostensibly nonsexual motivations are essentially sexual in nature. For example, emotional congruence with children may reflect a romanticized sexual attraction to children, rather than a focus simply on sexual gratification. As another hypothetical example, a man who is viewed as motivated by a nonsexual desire for power and control over an unwilling woman may in fact be expressing nonconsensual sexual sadism or biastophilia, wherein having that power and control over a woman is sexually arousing. Another possibility that can be explored in further research is that nonsexual and sexual motivations are intertwined in complex ways; for example, a more distal, nonsexual, and misogynistic motivation to have power and control over women may be associated with a more proximal, sexual, and still misogynistic motivation to engage in coercive sex.

Another question for the MFM is how to address offenders who victimize both children and adults. Such individuals are less likely to be have pedophilia (given some sexual offenses targeting adults) but they might be more likely to be have biastophilia or sexual sadism (where this activity paraphilia trumps age or gender preferences), high sex drive, or antisociality. I recently discussed in a different paper how sexual attractions to gender, age, and sexual activities might be co-organized (Seto, 2017). Another gap or limitation is whether the MFM can explain incest offending. Individuals who commit offenses against related children are unlikely to have pedophilia or hebephilia, and do not show evidence of high sex drive or intense mating effort, on average (Seto, Babchishin, Pullman, & McPhail, 2015). So what then motivates their sexual offenses against related victims?

The focus of the MFM is individual differences associated with risk to sexually offend, reflecting the clinical and forensic focus of the original model development. The MFM focuses on the (potential) perpetrator’s psychology, but sexual offending involves both a perpetrator and a victim. It is increasingly clear that we also need to consider protective factors, situational factors, and systemic or structural factors, integrating psychological and criminological theories and findings. Protective factors are less understood than risk factors, but they might include intelligence, social support, and positive community ties (e.g., as a result of work, religious observance, volunteerism). The MFM also does not explicitly incorporate victim characteristics, vulnerabilities, or resistance. Although the model can address offending against children and against adults, there are important differences. For example, sexual offenses against genetically related adults are rare, or rarely reported to authorities, whereas sexual offenses against genetically related children are a major subset of sexual crimes against children.

Finally, a major limitation of the MFM, and of other proposed models to explain sexual offending, is that causal inferences are temptingly made from correlational evidence. These causal inferences need to be tested. We need research to demonstrate that the onset of putative motivations precedes the onset of sexual offending, preferably from longitudinal follow-up as well as retrospective recall studies. And we need both multivariate and experimental work to rule out alternative explanations. For example, laboratory studies conducted in the 1980s showed that experimental manipulations to induce anger at a female conference or to activate alcohol expectancies could increase sexual arousal to depictions of nonconsenting sex, suggesting anger and alcohol use are states that could facilitate sexual response to the sexual cues in those depictions (Barbaree, Marshall, Yates, & Lightfoot, 1983; Yates, Barbaree, & Marshall, 1984). Could similar experimental manipulations facilitate sexual arousal to depictions of children? Such studies have not been conducted yet. Social priming paradigms might also be able to temporarily manipulate sex drive and mating effort to see whether they have the expected effects on laboratory analogs such as genital response or performance on implicit cognition tasks such as the implicit association task (e.g., Gray, Brown, MacCulloch, Smith, & Snowden, 2005).

Future Directions

The identification of gaps and weaknesses in the MFM points to a variety of different directions for further research, to test the model, fill in the gaps, and possibly modify it. Focusing on the model as it applies to sexual offenses involving children, we need research exploring whether the model explains the onset and/or persistence of sexual offending among self-identified perpetrators, that is, using anonymous surveys with community samples. Online survey technologies assuring confidentiality could shed light on whether the MFM is best suited for clinical or forensic populations.

Most research on the sexual motivations for sexual offending against children has focused on pedophilia. However, hebephilia is a relevant and distinct sexual interest in pubescent children, and is likely to also play a major role in sexual offending against children (Stephens, 2015). Indeed, I have suggested that the prevalence of hebephilia should be higher than for pedophilia, because the latter represents a greater departure from the species-typical sexual interest in sexually mature young adults (Seto, 2017). The MFM would predict a strong association between paraphilic motivations and sexual recidivism in terms of victim characteristics and the nature of the offending. Follow-up studies typically look at how paraphilia indicators can predict sexual recidivism, but we would expect a correspondence where evidence of pedophilia specifically predicts future offending involving children (child pornography or sexual contacts with children). Such studies would probably require aggregation of multiple studies with detailed follow-up information (beyond criminal records indicating new charges) to detect these effects.

For sexual offending against adults, there is still a great deal of debate about the importance of paraphilic motivations, though the existing evidence would suggest these motivations play a role that is comparable to that of pedophilia for sexual offending against children (cf. Lalumière et al., 2005; Seto, 2008). A coercive paraphilic disorder was not, however, included in the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5; APA, 2013), whereas sexual sadism was specifically listed. An important question is whether the distinctive sexual arousal patterns shown by rapists, as a group, compared with nonrapists, is explained better as biastophilia (nonconsent) or nonconsensual sexual sadism (violence/suffering). Arousal to nonconsent is different from arousal to violence/suffering because nonconsent does not require cues of pain, injury or suffering, and there are self-identified sexual sadists who prefer consenting interactions. Our research suggests that rapists are more influenced by nonconsent cues than by violence/injury cues, whereas the reverse pattern is observed among self-identified sexual sadists living in the community (Harris et al., 2012; Seto et al., 2012). There is much less controversy about the role of paraphilias such as exhibitionism and offenses involving indecent exposure or voyeurism and offenses involving invasion of privacy, but we still need follow-up research to determine whether having these paraphilias does indeed predict the same kinds of sexual offenses if recidivism occurs. Much more work is needed to understand the role of paraphilias in sexual offenses against adolescents and adults.

There is surprisingly little work on the role that high sex drive plays in sexual offending. There is some empirical support for the use of sex drive reducing medications for some pedophilic sex offenders who have demonstrated an inability to control their sexual urges, though methodologically stronger evaluation studies are clearly needed (Rice & Harris, 2011). The rationale for this medical intervention does not require high sex drive, however, as even reducing normal sex drive could be associated with a reduction of sexual behavior toward children. A similar logic has been applied to the use of sex drive reducing medications for offenders motivated by sexual sadism directed at nonconsenting persons (Thibaut et al., 2010).

Impersonal sex, which is related to intense mating effort, is an important dimension in Malamuth’s (1998) confluence model and seems to play an important role in explaining sexual aggression toward women in studies involving community samples. But what about criminal justice samples? The focus has been on paraphilias and, to a lesser extent, high sex drive, but it would be very interesting to ask questions about and collect information regarding the proportion of lifetime number of sexual partners (already used as an indicator of high sex drive) who were casual, short-term, and relatively impersonal. I noted earlier that many individuals who have sexually offended against children do not have pedophilia or hebephilia; many of these nonpedophilic offenders might instead be characterized by high sex drive or intense mating effort.

From the MFM, I would predict that a typology of sexual offending based on a clear assessment of primary motivation(s) could lead to better predictions of treatment needs and sexual recidivism outcomes than one based on other aspects of personal history such as maltreatment history, mental health, or criminal history, including sexual offending history. Beyond the expected correspondence between paraphilias and sexual offending behavior and victims, I would expect that individuals who are primarily motivated by high sex drive will target adolescents and adults corresponding to their gender sexual orientation, including individuals who are already known to them. I expect individuals who are primarily motivated by intense mating effort will target peers who they have never been sexual involved with before, particularly in school, friendship, and work contexts.

We do not yet have strong tests of the MFM with adolescent offenders. There is limited work on pedophilia or other sexual motivations, with some evidence that a small minority of male adolescents, particularly among those who offend against boys, show relatively greater sexual arousal to children than to peers or adults (see Seto, 2008). I am not aware of any studies of biastophilia or excessive sexual preoccupation in adolescent sex offenders. Similarly, the research literature on female offenders is growing, but we know much less about women who offend, either against children or adults. In addition, most research on offending involving adults has focused on cases where men perpetrate against women. We have an extensive literature on men who offend against boys, but very little evidence about men who offend against other men.

Although the MFM has strong intuitive appeal as an explanation of noncontact sexual offending involving exhibitionistic or voyeuristic activity, there is very little empirical research on these forms of sexual offending, irrespective of tests of the MFM. Policy and practice has mostly focused on contact sexual offending, though there has also been a second line of research on online sexual offending over the past decade (see Seto, 2013). Also, though community samples have been studied, more work needs to be done, for example, examining different theoretical frameworks in community studies that focus more on social and situational factors (e.g., alcohol use and expectancies, sexual scripts.)

Longitudinal studies of general or at-risk cohorts of children and young adolescents are required to demonstrate whether the MFM can explain the onset of sexual offending, as opposed to the persistence of sexual offending. These studies are expensive, difficult, and time-consuming, but essential if we are to correctly explain the origins of sexual offending. In addition, the field is in dire need of more experimental manipulations of motivation or facilitation via sexual arousal, mood induction, and social/cognitive manipulations of risk-taking behavior. This is a surprising gap in the literature because early sexual offending studies did experimentally examine the impact of factors such as pre-exposure to mainstream pornography, anger induction, and alcohol expectancies on sexual arousal to non-normative stimuli (e.g., descriptions of rape). Research drawing on these paradigms is needed to understand the impact of these factors, and their potential interactions with sexual motivations and trait facilitation factors with sex offenders against children and other forms of sexual offending. Finally, as is true of so much social science research, most of what we know about sexual offending comes from developed, English-speaking countries. There is a critical need for work from other countries so the cross-cultural applications of the MFM can be explored.

A developmental perspective is needed because there is good evidence that the factors discussed in the MFM change over time. We do not expect age-related changes in paraphilia per se, though the intensity of paraphilic sexual thoughts, fantasies or urges may decline over time in conjunction with an overall decline in sex drive. Mating effort is life-course dependent, changing with age and with life circumstances. It is usually highest among young men, which may help explain why young men are more likely to engage in sexual aggression than older men (see Lalumière et al., 2005). I would also expect age-related changes in the intensity and expression of facilitation factors, including self-regulation problems, hostile masculinity, negative affect, and alcohol use (e.g., Hare, 2003; Hare, McPherson, & Forth, 1988).

Final Comments

The MFM has similarities to other models or theories proposed to explain sexual offending, but it is distinctive in its emphasis on primary sexual motivations, distinction between trait and state facilitation factors, and potential versatility across types of sexual offending. Considering the criteria for judging whether a model or theory is scientifically useful, this article argues that the MFM has promise: It helps unify different facts and theories about sexual offending, is internally consistent, is parsimonious, and generates testable ideas for future research.