Killing Children

Contents of this page

  1. Data for England & Wales
  2. Sex of the Perpetrators
  3. Infanticide
  4. Omissions from the Data? Infanticide and Neonaticide
  5. Sudden Infant Death (SID, or “Cot Death”)
  6. USA Data

1. Data for England & Wales

Data on child killings in England and Wales appears to suffer from considerable obfuscation, shockingly so given the importance and sensitivity of the subject. This issue has been well described by Robert Whiston and Nigel Hawkes.

Year Child Deaths Reference given below
90-91 122 Ref.1
91-92 115 Ref.1
92-93 116 Ref.1
93-94 112 Ref.1
94-95 120 Ref.1
98/99 78 Ref.2
99/00 65 Ref.2
00/01 95 Ref.2
01/02 62 Ref.2
02/03 91 Ref.2
02/03 99 Ref.5
03/04 70 Ref.2
04/05 60 Ref.2
05/06 48 Ref.2
06/07 64 Ref.2
07/08 63 Ref.2
07/08 210 Ref.3
08/09 174 Ref.4
09/10 90 Ref.7
10/11 56 Ref.6
11/12 47 Ref.6
12/13 67 Ref.6

Child Homicides v Year

Sex of the Perpetrators

Information on the sex of the perpetrators of child killings in the UK is sparse. ONS data does not specify. This is odd because in other cases, e.g., for intimate violence, the sex of perpetrators looms large.

In 2013, the NSPCC web site included some very limited information, as follows,

  • Killings of children by a natural parent are committed in roughly equal proportions by mothers and fathers, but where the child is killed by someone other than a [birth] parent, males strongly predominate.
  • The proportion of child homicides by a parent, or parents, is about 70%.

These statements were attributed to a report “Reducing homicide: a review of the possibilities” by Brookman and Maguire, 2003. However, the link to this report is now dead, and the above statements no longer appear on the NSPCC web site. However, the 2004 Dewar Research report by Yarwood includes data from the period 1995 – 2001 from which he draws similar conclusions,

  • A parent is the principal suspect in about two-thirds of cases of child killings. (Ref.6, above, confirms for years 2010 to 2013 that a parent or step parent is the perpetrator in two-thirds of cases);
  • Where a biological parent was responsible for child killing, the father was the killer in 54% of cases and the mother in 46% of cases. However, if non-biological carers/parents are included these figures become about 60% and 40% respectively.

The above sources suggests that mothers are responsible for roughly one-third of child killings and that biological fathers are also responsible for roughly one-third of child killings, the bulk of the remaining one-third of perpetrators being other males. However an analysis of Serious Case Reviews carried out in October 2016 has led to a rather different picture emerging in which mothers appear responsible for more child killings than fathers and male partners combined (further details below).

There is an influential report called 29 Child Homicides widely used by support services. It details the case histories of 29 instances of men who killed their children. It was written by Women’s Aid in 2004. Its purpose is to reinforce the view that fathers are dangerous (and hence to frustrate attempts by fathers to obtain custody rights in the family courts). Karen Woodall observed the following regarding the gender-biased perspective of protection agencies:

How many of those same support services, I wonder, who willingly and blindly accept that the only issue to be concerned about around domestic violence is the safety of women and children, are aware that during the same period covered by the 29 Child Homicides report, NSPCC research shows that some 800 children have died at the hands of resident mothers or carers. And how many of those same support services that rely upon 29 Child Homicides to inform their thinking, have read the 2000 publication “Child Maltreatment in the UK”, which showed that violent treatment was more likely to be meted out by female carers than male ones“.

In 2016 Women’s Aid produced a further report along similar lines: “Child First: Nineteen Child Homicides“, discussed further here. This prompted a review of the Serious Case Reviews (SCRs) published by the child protection authorities over the 7 years 2009 to 2015. In principle, all child killings should result in a SCR, but in practice this is not the case. Consequently it is not entirely secure to assume that the SCRs are representative of the totality of child killings. On the other hand, this review of SCRs is probably the most reliable source of perpetration data currently available. The conclusions of the review of the SCRs was,

  • One or both parents are culpable in the killing of at least 65% of children – and probably a much greater percentage if culpability could be assigned in every case.
  • Mothers’ male partners, where not the father, are culpable for ~11% of child killings. This might be a somewhat larger percentage if culpability could be assigned in every case.
  • Children are killed by someone other than their parents or the mother’s male partner in only ~7% of cases.
  • Where culpability is established, the mother is the lone perpetrator in 36% of cases and either a lone or a co-perpetrator in over half of cases (58%).
  • Mothers are more likely to be responsible for a child death than fathers and male partners combined.
  • Single mothers are the demographic most likely to be responsible for the deaths of children (61 cases versus 55 cases of fathers as sole perpetrators).

However, readers should be aware of the difficulties inherent in assigning culpability in the cases of certain child killings – essentially where the boundary lies between accident and culpably neglect. See the review for further discussion.

Yarwood also concludes,

  • A substantial number of filicides occur following parental separation, primarily by fathers: of these, a substantial proportion then commit suicide and a further significant proportion require psychiatric services
  • Male parents who kill their children are generally treated more harshly and unsympathetically by the legal process than female parents: fathers are more likely than mothers to be charged with murder than manslaughter, and more fathers than mothers convicted of manslaughter are imprisoned; convicted mothers are more likely than fathers to be hospitalised or treated rather than imprisoned.


In English law, infanticide is defined as the killing of a child under 1 year by its mother. Infanticide is recognised as a crime distinct from murder or manslaughter specifically to permit leniency to be shown towards mothers. Women are vitually never given a custodial sentence for infanticide. In fact very few infanticide cases are even brought to court now. In contrast, men killing babies can expect a long prison sentence.

David Thomas (“Not Guilty: In Defence of the Modern Man“, Weidenfeld and Nicolson, London, 1993) drew attention to the curious way in which the criminal justice system in Britain treats infanticide. In the period 1980 to 1990, 293 children less than 1 year old were victims of homicide in Britain. Yet only 42 suspects were charged with homicide/infanticide in the same period. In the years 1989/90, the discrepancy was even more marked. There were 50 child victims but only one suspect was brought to court. This observation implies that the overwhelming majority of killings of children under 1 year are infanticides, i.e., they are committed by mothers. This is because the perpetrator will almost always be known in the case of baby killing – and male perpetrators would be prosecuted virtually every time.

Data on prosecutions for infanticide in the period 2002 to 2013 have been obtained via an FOI request to the CPS (Ref. 13/2014). Data in the form of an Excel file (also copied here in case the link is suspended) provides the following information (relating to England & Wales),

  • In the 11 year period 2002-2013 just 4 prosecutions were brought for a charge of infanticide, at least 3 of which resulted in a verdict of not guilty;
  • Despite this, in the same 11 year period there were 12 convictions for infanticide. This is possible because the convictions for infanticide result from a more serious initial charge, e.g, murder.
  • Of the 12 convictions for infanticide, none resulted in a prison sentence.
  • The sentences awarded for the 12 convictions for homicide were: one conditional discharge, two hospital orders, eight community sentences and one suspended sentence.

The text response to CPS FOI 13/2014 (also copied here) included further information on the original charges which ultimately resulted in convictions for infanticide. Of the five cases of convictions for infanticide for which records had been retained, all five were against original charges of murder.

The picture which emerges from this evidence is that women are very rarely charged with infanticide, but rather infanticide is used as a fall-back when charges of murder are brought against a mother in relation to infant death. The remarkable observation is that the prosecuting authorities consider it worthwhile, in all such cases, of pursuing a charge of murder and yet the outcome is never a prison sentence even after a guilty verdict – because the charge is dropped to infanticide.

Neonaticide is the killing of a baby within its first 24 hours of life. Neonaticide is virtually exclusively carried out by the mother.

Omissions from the Data? Infanticide and Neonaticide

In the above Table of child death data, note that the Ofsted reports, Refs.3 and 4, are lower bounds to the number of deaths because Ofsted was entirely reliant on local authorities sending them serious case reviews. Some authorities failed to do so. Despite this the Ofsted reports Refs.3,4 record substantially greater numbers of child killings than the other sources, notably the ONS homicide statistics. It would appear that some of the categories used in Refs.3,4 are not being picked up by the ONS data. Ref.3 recorded killings in the following categories: suicide, murder by parent or carer, physical abuse, substance misuse identified as a factor, neglect, killed by parent whilst sleeping, domestic violence identified as a factor, killing by another young person, death by accident where neglect occurred, shaken baby syndrome, and other or unexplained causes. I find it odd that Ofsted reverted to ONS data after 2009, rather than continuing to report the more detailed, and larger, statistics. One presumes they continue to receive the local authority serious case reviews.

There is no reliable data on the frequency of neonaticide. To quote Michael Craig, J R Soc Med. 2004 Feb; 97(2): 57–61: “It is very difficult to get accurate figures on the incidence of neonaticide and infant homicide since many cases are never discovered; official figures are likely to be an underestimate”. Resnick, Am J Psychiatry 1970;126: 1414-20, has suggested that “hundreds and possibly thousands of neonaticides still occur in Britain each year”.

Steven Pinker has expressed a similar opinion in the context of the USA, “every year, hundreds of women commit neonaticide: they kill their newborns or let them die. Most neonaticides remain undiscovered, but every once in a while a janitor follows a trail of blood to a tiny body in a trash bin, or a woman faints and doctors find the remains of a placenta inside her.

The suggestion that neonaticide occurs in such large numbers seems inflammatory. But there are arguments that mothers may, in certain circumstances, regard it as their due right to dispose of a new born infant and that this perspective results from evolution. In this respect note that there is a body of opinion amongst feminists that abortions should be legal without any time limit (e.g., Jessica Valenti). Since abortion within days of full term is essentially identical to the killing of a new born baby, this is consistent with some women believing they have a natural right to determine whether a neonate should be allowed to live. The widespread view is that mothers who kill their children must be mentally ill. But, even if true in general, neonaticide may be an exception to this. Phillip Resnick found that mothers who kill their older children are frequently psychotic, depressed or suicidal, but mothers who kill their newborns are usually not.

If it is true that “hundreds and possibly thousands” of neonaticides occur in Britain each year it would dwarf all other child homicides. Indeed, it might even dwarf all homicides, including that of adults.

Another potentially serious source of under-estimation of child killings is due to killings which are passed off as Sudden Infant Death (SID).

Sudden Infant Death (SID, or “Cot Death”)

The ONS Excel file Death Registration Summary Tables, 2013 records 160 deaths due to SID (sometimes called ‘cot death’). Nearly twice as many male babies died from SID (102) than female babies (58).

In the previous year, the ONS Child Mortality Statistics, 2012 recorded a total of 131 deaths from SID. No one really knows what causes SID. But the latter source includes a very interesting breakdown of SID deaths into those pertaining to married women (31) and those for unmarried women (100). To adjust for the different numbers of married and unmarried women, these figures can be expressed as a percentage of the total number of infant deaths from all causes registered in 2012 for married and unmarried women respectively. 2.3% of married women’s infant deaths were due to SID, compared with 6.7% of unmarried women’s. Hence SID is nearly three times more common in infants of unmarried women than for married women.

But it gets worse. The data for unmarried women is also subdivided into those relating to “joint registration”, where the father is involved, and “sole registration”, where information about the father is not available. There were 23 cases of SID in the “sole registration” category, comprising 9.8% of all infant deaths to sole registrants. It would appear that SID is more than four times more common for sole registrants than for married women.

It is not the first time that this has been observed. Yarwood (2004) has the following observations to make regarding Sudden Infant Death,

“In 2001, there were 231 SIDs in England and Wales. The rates of cot death vary with social class and marital status. In 2001, the rate was 1.28 per 1000 live births among children of single mothers, six times the rate of 0.2 per 1000 for babies of married couples. The rate for babies registered by unmarried parents living at different addresses was 0.71 per 1000 live births, over three times the rate for married couples. A report published in April 2004 by researchers from the Foundation for the Study of Infant Deaths (FSID) suggested that as many as one in 10 of sudden infant deaths could be cases of murder or child neglect, amounting to 30 to 40 of ‘covert homicides’ of babies a year in Britain. The number of sudden baby deaths registered in Britain has steadily fallen in recent years. However, the researchers believed that the number of suspected ‘covert homicides’ has stayed about the same, although they admitted that “it is impossible to be certain of the exact figures.” The findings were based on the national “Confidential Enquiry into Stillbirths and Death in Infancy 1993-1996″, a study of the outcome of some half million births.”

The implication is that SID can hardly be due primarily to the intrinsic biology of the infant if it is so strongly correlated with the marital status of the mother. There would appear to be a dominant influence of the standard of care. In the extreme case the above FSID quote suggests that infanticide may be part of the explanation. This is a disconcerting conclusion, and additionally so in view of the fact that nearly twice as many boy babies die of SID than girl babies.

There is no doubt that a non-trivial proportion of SIDs are in fact homicide. Steven Kairys et al, Pediatrics February 2001, 107, Issue 2, observe,

“Recently, well-validated reports of child abuse and infanticide—perpetrated by suffocation and masqueraded as apparent life-threatening events (ALTE) and/or SIDS—have appeared in the medical literature and in the lay press.

Parents of infants with recurrent ALTEs have been observed trying to suffocate and harm their infants.In Great Britain, covert video surveillance was used to assess child abuse risk in 39 young children referred for evaluation of recurrent ALTEs. Abuse was revealed in 33 of 39 cases, with documentation of intentional suffocation observed in 30 patients. Among 41 siblings of the 39 infants in the studies, 12 had previously died suddenly and unexpectedly. Although 11 of these deaths had been classified as SIDS, 4 parents later admitted to suffocating 8 of these siblings. Other cases previously thought to be multiple SIDS deaths within a family have been revealed to be cases of multiple homicide by suffocation.”

Michael Craig, J R Soc Med. 2004 Feb; 97(2): 57–61 quotes Wolkind et al, Acta Paediatr 1993;82: 873-6, writing, “a controversial body of evidence suggests that somewhere between 2% and 10% of cot deaths may ultimately be attributable to infant homicide“. Wolkind et al actually studied families with 2 or more cot deaths. They concluded that 55% of these were probably filicide (infanticide). However, it is not clear what conclusion we can draw from this as regards the frequency of homicide masquerading as SID in cases of just one SID in a family.

USA Data – Killing of Children

In view of the obfuscation in UK data, especially as regards the perpetrators of child killings, it is worth supplementing the UK data with the USA experience, though caution is needed in reading directly across to a different society. The conclusions below are based on “Child Maltreatment 2011“, National Child Abuse and Neglect Data System (NCANDS), U.S. Department of Health and Human Services. There is also a 2013 update report. The data below are for 2011 with the 2013 data in brackets.

  • Parents are responsible, or share responsibility, for killing children in 78.3% (78.9%) of cases.
  • The mother is the sole killer in 26.4% (27.7%) of cases, whereas the father is the sole killer in 15.3% (12.4%) of cases, so the mother is the sole killer twice as often as the father.
  • When the killer involves at least one parent, the mother is involved in the death in 78% (82%) of cases.
  • Overall the mother is involved in the death in 61% (64.8%) of cases.
  • Neglect is the cause, or a contributory cause, of death in 71.1% (71.4%) of cases.
  • Physical abuse is the cause, or a contributory cause, of death in 47.9% (46.8%) of cases.
  • Sexual abuse is involved in fatalities in only 0.7% (1.0%) of cases.
  • Boys are killed significantly more often than girls, 59.1% (58%) boys, 40.6% (41.7%) girls. This is also observed in UK data.

The preponderance of killing by mothers is related to the age spectrum of the victims. The graph below shows the concentration on younger children – especially babies (under 1 year) – 2011 data.

Child Killings USA 2011

Related links (open in this tab)