There are 500-600 homicides annually in England and Wales. Perpetrators and victims are predominantly young males, especially when the victim is unknown to the perpetrator. In such ‘stranger homicides’ perpetrators are less likely to have a lifetime history of mental illness, symptoms of mental illness at the time of the offence, or contact with mental health services.

In the UK the total number of both total homicides and stranger homicides increased between1973 and 2003 but neither category increased in people with mental illness.⁽¹³⁾ Similar trends have been noted in work from both the UK,⁽¹⁴⁾ and in other countries.⁽¹⁰⁾

The commonest method of homicide is with a sharp instrument; shooting is relatively rare, accounting for less than 1 in 10 homicides in the UK.

Around half of all convictions are for murder and just under half for manslaughter. One in 25 receives a verdict of Section 2 manslaughter, diminished responsibility.

Despite an increasing rate of homicides in the general population, convictions for infanticide and the rate of infant homicide has remained relatively constant, at around 4.5 per 100 000 live births.⁽¹⁵⁾ Infanticide has become a generic term for killing of infants, even though the criminal charge in England applies to a crime for which only a woman can be indicted.

Although the risk of homicide is higher in the first year of their life than at any other time, the rarity of infant homicide in absolute numbers means that there is a lack of high quality, systematic data at a population level which incorporates clinical characteristics.⁽¹⁶⁾

Data from the National Confidential Inquiry from 1996 to 2001 shows that 1 in 25 of the 2665 homicide perpetrators identified were convicted of infant homicide. Half of these infants were killed by their father and around a third by their mother. A quarter of perpetrators had symptoms of mental illness at the time of the offence and a third had a lifetime history of mental illness. Perpetrators of neonaticide were predominantly young, unmarried mothers experiencing symptoms of dissociation at the time of the homicide.

There were significant differences between male and female perpetrators, with males being more likely to have previous convictions for violent offending. Females were more likely to kill within a month of the birth and they were more likely to have affective disorder and symptoms of mental illness at the time of the offence but few of these women were under the care of mental health services.
Most males received a custodial sentence, whereas three quarters of women received a community sentence or hospital disposal.⁽¹⁶⁾

Multiple homicides, in particular serial homicides, have generated a great deal of public and media interest over recent decades yet this phenomenon is rare in the UK. The rarity of these events means that there is a lack of empirical evidence about the characteristics of perpetrators and victims in the UK, with most evidence emanating from the United States. Even then, however, there is an absence of systematic, robust evidence, with many studies being limited by small sample size.

Most definitions of multiple homicides include three criteria; number of victims, which can vary from 2 to 10 in different definitions,⁽¹⁷⁾ time, and motivation. The temporal relationship distinguishes subcategories: mass murder consisting of a single episode and location, with serial, and spree murders occurring over time in separate locations. The latter two are differentiated by an emotional ‘cooling-off ’ period, which is present in serial homicide. Other authors have discussed motivation, such as sexual gratification and internal psychological gratification, but the lack of robust evidence means that it seems premature to include motivation as part of any such definition.⁽¹⁷⁾

Mass murder has been classified by victim type such as family annihilators and classroom avengers. Mullen⁽¹⁸⁾ proposed a category of ‘autogenic (self-generated) massacre’, which encompassed perpetrators indiscriminately killing people in pursuit of a highly personal agenda, arising from their own specific social situation and psychopathology. They were characterized by social isolation, being bullied in childhood and personality traits such as suspiciousness, obsessional behaviour, grandiosity, and persecutory beliefs. He concluded that these murders are essentially murdersuicides, where the intention is to kill as many people as possible before killing themselves. It would now appear, particularly with recent events in Virginia, that this form of multiple homicide is an established form and concerningly appears to becoming more common. Cantor et al.⁽¹⁹⁾ propose that media-related modelling is a potential factor in the emergence of this crime, with perpetrators often seeing themselves as lone warriors, themselves modelled on media images, and well informed about previous, similar, massacres.

An exploratory study, incorporating a nested case control study, showed that serial homicide offenders were more likely to be male dominated, compared with single homicide offenders, and were more likely to use strangulation. Moreover, victims of serial murders were significantly more likely to be females who were unknown to the perpetrator and the motivation being sexual.⁽¹⁷⁾ Unfortunately most classification systems of serial murder, including the FBI classification, have been criticized as being inherently flawed due to weak operational definitions and unsubstantiated assumptions regarding behaviour and characteristics.⁽²⁰⁾