Introduction: Defining High and Low Intellectual Functioning
Intellectual functioning has typically been quantified by scores obtained on widely recognised intelligent quotient (IQ ) tests. Currently the most frequently used measure of IQ is the Wechsler Adult Intelligence scale (WAIS), which is in its fourth edition (WAIS-IV; Wechsler, 2008). Scores obtained on IQ tests such as the WAIS have largely been viewed as synonymous with an individual’s intelligence, a concept for which there are various definitions. In reviewing over 70 published definitions, Legg and Hutter (2006) presented their own definition which encapsulated key themes identified within those reviewed: ‘Intelligence measures an agent’s ability to achieve goals in a wide range of environments’ (p. 8).
The extremes of intellectual functioning can be defined through reference to upper and lower standard deviations of IQ scores. Thus, IQs which fall two standard deviations away from the mean are considered ‘extreme’, with an IQ of 69 and below representing ‘extremely low’ and 130 and above representing ‘very superior’ intellectual functioning. These scores have subsequently been used as thresholds to categorise individuals. For example, criteria for diagnosing an intellectual disability includes an IQ below 70 (British Psychological Society, 2000) and membership of Mensa, the High IQ Society, requires an IQ over 130 (MENSA, 2019).
This chapter reviews the theoretical history and research relating to individuals who fall within these two extreme groups and who have committed sexual crimes. In doing so, it acknowledges the various limitations of the IQ approach. In particular, it is acknowledged that groups of individuals with similarly assessed IQs are likely to be heterogeneous. In the case of populations with significantly low IQ, limitations in intellectual functioning may be associated with global deficits in functioning characteristic of an intellectual disability , genetic disorders such as fragile X syndrome or cognitive deficits acquired via illness or injury. Groups of those with superior IQs are also likely to differ considerably. For example, it is not uncommon for those diagnosed with Asperger syndrome to have high full scale IQs, or to be considered ‘intellectually gifted’ (e.g. Henderson, 2001). However, studies have found that these individuals’ scores on the Verbal IQ subscales are typically much higher than their scores on Performance subscales (Chiang, Tsai, Cheung, Brown, & Li, 2014; Klin, Volkmar, Sparrow, Cicchetti, & Rourke, 1995), suggesting these individuals may differ significantly from those with globally high IQs.
This chapter does not seek to discriminate between subgroups of those with extreme IQs, rather, its purpose is to offer an overview of the literature concerning individuals populating these broad categories.
Methodological Issues in the Study of Intellectual Functioning and Crime
Before offering an overview of the literature pertaining to intellectual functioning and sexual crime, a number of methodological issues within the field must be highlighted. Firstly, empirical studies which have sought to investigate prevalence figures are hindered by so-called dark figures of crime. In terms of those with low intellectual functioning, ‘true’ figures of crime perpetrated by this group are likely diluted throughout all stages of the criminal justice system. In the first instance, crimes may go unreported by members of the public or those occupying a caring position. This may be motivated by a dismissal of the severity or illegality of the behaviour or the belief that the individual should not be deemed culpable of their inappropriate sexual behaviour (Holland, Clare, & Mukhopadhyay, 2002; Lyall, Holland, & Collins, 1995; Turk & Brown, 1993). Professional services caring for these individuals may feel reporting is unwarranted, given that the individual is already in receipt of therapies or has restrictions imposed upon them which address the behaviour of concern. All these possibilities may subsequently lend to a culture of non-reporting. In instances where sexual offences are alerted to the police, further drop out in recording is likely to occur due to Crown Prosecution Service decisions not to prosecute. Such decisions will be influenced by the predicted likelihood of conviction and what is deemed to be within the public interest (Holland et al., 2002). Again, judgements regarding the culpability of the individual will likely mediate such decisions.
For different reasons, the real extent of sexual crime perpetrated by those with high intellectual functioning is also likely to be hidden. First, those with high IQs are perhaps more likely to evade detection when perpetrating crimes. These individuals may be better resourced than those with average or below average IQs to plan, carry out and conceal their criminal acts. Indeed, as Jones (2010) points out, reconviction rates only provide information on detected offences and, as a consequence, exclude those who have employed successful detection and conviction evasion skills. This hinders developments in criminal practice such as risk assessment, formulation and treatment which often draw upon knowledge of reconvictions. Where their crimes are detected, those with high IQs may experience more lenient treatment upon apprehension, trial and sentencing (e.g. Bennett, Levinson, & Hioki, 2017; D’Alessio & Stolzenberg, 1993) and may be better financially resourced to defend their case. In fact, McCord, McCord, and Zola (1959) argue that the long observed association between low IQ and crime is primarily due to this group being at greater risk of both detection and punishment.
Apropos sentencing, a large body of research has investigated the effects of various factors on sentencing decisions, including the defendant’s personal characteristics. This area of research is largely motivated by the desire to explain significant disparities observed in sentencing decisions. Race and sex are amongst the most commonly investigated variables, with studies consistently reporting males and those from African American backgrounds being subject to harsher decisions (e.g. Desantts & Kayson, 1997; Hessick, 2010). The influence of the defendant’s intellectual capacity in sentencing decisions does not appear to have been investigated as a standalone factor; however studies concerning privileged groups’ experiences of the justice system are perhaps indicative of at least some influence. For example, Bennett et al. (2017) present empirical evidence of judges’ avoidance of harsh sentencing when dealing with white-collar fraud cases, with the majority opting for the minimal possible sentences available to them. The authors consider one explanation for this is that these judges are more empathic to those committing white-collar crimes due to greater identification with these individuals, lending to greater leniency. Additionally, the recent sentencing of Paul Manafort, advisor to President Donald Trump, for a series of fraudulent crimes offers a case example of this supposed elite bias. Critics have argued that his sentence of 47 months imprisonment for a string of offences estimated to cost the Internal Revenue Service millions of dollars fell significantly below the sentencing guidelines put forward by prosecutors of 19.5 to 24 years.
Thus, what we are left with is a skewed picture of crime, which is likely to underestimate the incidence of crime perpetrated by intellectually superior persons. Our understanding is further marred by a significant lack of research investigating the incidence and nature of offending committed by these individuals as well as their treatment within the Criminal Justice System. It is for this reason Oleson (2016) advocates for the use of self-report methodology to gain better insight into crimes perpetrated by those considered ‘elite’.
At this point the link between social standing and IQ is worthy of discussion. The positive relationship between class mobility and performance on intelligence tests represents a robust research finding (e.g. Nettle, 2003); so too does the link between cognitive ability and academic achievement, occupational success and wealth (e.g. Murray, 1998; Wai, 2013). As such, the commentary within this chapter draws parallels between IQ and social eliteness. In doing so, the author does not seek to deny the various limitations in such an approach, not least the role of cultural bias in both the construction and testing of IQ, nor the presence of high IQs within socially deprived groups.
The second methodological issue concerns the definition of the population studied. As described above, any review of literature concerning high and low intellectually functioning groups is likely to be faced with heterogeneous populations. Literature concerning those with low intellect is particularly varied, covering those with learning disability, brain injury and genetic disorder, with historical writings describing groups such as ‘imbeciles’, ‘idiocy’, ‘feeblemindedness’ and mental defectives’.1 The nature of the sexual crimes perpetrated by those studied offers a further layer of complexity. Some studies have sought to separate sexual offending into groups which theoretically gather together. For example Vera, Barnard, and Holzer (1979) compared IQs of individuals convicted of rape, nonviolent sexual, violent non-sexual and nonviolent non-sexual offences, all of whom were incarcerated within a Canadian high secure setting. Other studies have grouped together all those who have committed a sexual crime, regardless of its nature. For example, Guay, Ouime, and Proulx (2005) compared the IQs of individuals with sexual convictions to those with non-sexual violent convictions. A further issue when researching sexual offending is that the offence the individual is charged with may not always reflect the criminal act which took place, particularly in instances of plea bargaining. All of these issues lend to an extremely varied evidence base in which comparisons between studies are difficult to make.
Early Interest in Intellectual Functioning and Criminality
Interest in intelligence and its relationship with adaptive and maladaptive behaviours was perhaps first spark...