The journal, Autism, enjoys a wide readership that extends far beyond academia. We set out here, for the benefit of the whole readership, to debunk the myth that autism causes criminal behaviour. We review the little research on this topic and describe how easily negative stereotypes can be reinforced by press reports.
King and Murphy (2014) conducted a thorough review of the research in this area. They found that on the whole, there is no evidence that people with autism are more likely to engage in criminal activity than people without autism. The studies they reviewed presented conflicting information, however. Some studies have found that people with autism are less likely to commit offences such as probation violations and property offences (Cheely et al., 2012; Kumagami and Matsuura, 2009), and another study reported that people with autism are no more likely to commit violent crime than the general population (Woodbury-Smith et al., 2006). On the other hand, some people with autism may be more likely than the general population to commit certain types of offences such as arson (Hare et al., 1999; Mouridsen et al., 2008), sex offences (Cheely et al., 2012; Kumagami and Matsuura, 2009) and assault and robbery (Cheely et al., 2012).
Research on autism and offending needs to be interpreted with caution, however. Most studies rely on information from small samples that do not represent the general population. These studies also rarely include people without autism for comparison. This makes it inappropriate to attempt to generalise these studies to the autism population at large. For example, two studies found a disproportionately high prevalence of autism in high security hospitals (e.g. Hare et al., 1999; Scragg & Shah, 1994), but this does not mean that the autism population as a whole includes a disproportionate percentage of people who present a danger to society.
There are also several case reports of people with autism engaging in criminal behaviour (e.g. Baron-Cohen, 1988; Mawson et al., 1985). However, generalisations cannot be made on the basis of individual cases regardless of whether these reports originate in the research literature or in the press, not least because it is often the unusual characteristics in such cases (e.g. the bizarre and random acts of violence noted by Mawson et al., 1985) that initially draw attention for analysis.