Skip to content | Skip to institutional links

Common menu bar links

Institutional links

Our Missing Children

National RCMP

Resources

Navigate by

Proactive Disclosure

Proactive Disclosure

The Killing of Canadian Children by a Parent(s) or Guardian(s): Characteristics and Trends 1990-1993

By

MARLENE L. DALLEY, Ph.D.
Ottawa, Ontario
CANADA
January 1997 & 2000

Missing Children’s Registry & National Police Services

Highlights
Introduction

Highlights: Homicide of Children by Parent(s) or Guardian(s)

Since 1990, the incidents of children, killed has increased from 34 to 51 in 1998, with a high of 62 in 1997. Although this statistic remains fairly consistent, it is interesting to note that children are almost as much at risk to be victimized from a parent or guardian as they are from a stranger.
Children less than one year of age were more at-risk in both single and multiple killings. When all children were considered in the analysis, the average age of the victim was 4 years. In this study, as well as others cited in the literature review, age is the clearest factor in terms of absolute risk.
Both mother and fathers kill infant children at the same frequency.
When all age groups were considered, mothers and step-mothers killed more children than fathers and step-fathers.
The gender of the children killed by mother and step-mothers and fathers and step- fathers was evenly distributed. This result varies from other research studies.
With regard to method used, fathers were more physically abusive and mothers strangled or asphyxiated their child to death. This was consistent with other studies.
To kill older children, fathers used a weapon and mothers used a variety of methods.
In multiple killings , the use of a weapon was more prevalent (in 4 of the 11 incidents, a firearm was used to kill 8 children). As well, the average age of a child killed by the mother was 4 years and the father, 8 years. The gender of the children killed was fairly evenly distributed.
Over half of the offenders did not contact the police or request any assistance after killing the child(ren); one quarter of this group committed suicide.
The offender was most often described as mentally ill or severely depressed. This finding is consistent with other research findings cited in the literature review.
Family discord, and relationship instability and ensuing problems were cited as factors contributing to these child deaths. This finding was consistent with other research findings cited in the literature review.

Introduction

Of the many actions punishable by criminal law, the taking of a life of another human being is thought of as the most abominable. As well, when the human life is that of a child, stronger emotions are evoked - sympathy and empathy for those persons whose lives the child has touched - a sense of loss that the book of life is closed before it is written - and, anger at a system that cannot protect a child from such a fate.

When the death is the result of some form of abuse and/or intentional neglect, such anger is intensified. Sadly enough, although the abuse and neglect of children are often reported to police and social service agencies, on occasion, the child is not sufficiently protected. After the shameful death of a 6-year old girl, Elisa, described by the media as a fairytale princess, Time magazine reported "She was not killed by an individual but by the silence of many" (Vol. 146, No. 24, December 11, 1995).

The death of Elisa Izquierdo, shocked the world. Born in a United States homeless shelter, her mom addicted to crack cocaine and under the care of social services, she was murdered viciously by her own mother. Just prior to this tragedy, she was given hope for a brighter future. A chance meeting with a Greek prince initiated the promise of full sponsorship of her educational pursuits. Although many persons may have speculated that she had "beaten the odds," her life ended in a shamefully sad death.

Following a national outcry in the United States, the "Elisa's Law Child Protection Services Reform Act of 1996," NY Soc. Serv. Lay 20 was initiated. This legislation was designed to lift the secrecy traditionally shrouding child abuse cases. Although there were many positive outcomes from this tragedy, one of the most important outcomes was legislative changes. Plans were developed to improve the computer systems which serve family courts so that "Judges are more able to review a parent's history of crime, domestic violence and child abuse before making custody decisions" (Update, Vol. 9, No. 3, 1996).

Canada is not immune to such crimes against children (See Appendix A for an overview of Canadian laws). Many cases have shocked this nation. For instance, in July of 1994, a British Columbia incident involved a father strangling his wife, killing his 13 month old twins and attempting to kill his 3-year-old son. The 3-year-old barely escaped death by running to a neighbour's house. Shortly after the incident, the father wrote about his deep love for his wife. However, following his arrest, the father described the killing with phrases, such as it was "like being stoned," and "above and beyond any kind of drug." It is difficult for anyone to understand a father and husband, feeling euphoria associated with the dreadful deaths of his children and wife.

In 1991, a Canadian father killed himself and his two preschool children by carbon monoxide poisoning in the family car. The couple were in the process of separating. The suicide note stated, "I told you that if you did not want me you would not have my children or my rings." A black rose was placed carefully on the note in the car. An autopsy revealed the father had swallowed his wife's wedding rings. Two children lost their lives when this father sought revenge.

In October 1998, The Ottawa Sun newspaper reported "that Quebec police believe the father of two tiny girls shot his daughters while they were sleeping in a tent and then turned the gun on himself."

In October 1999, The National Post newspaper carried a story of a father who was accused of savagely killing his pregnant wife and 6-year-old daughter and brutally beating his son.

These cases and others have drawn attention to this sad crime - the death of a child by a parent or guardian.

Since collecting the data for this study, the incidents of children killed by parents have fluctuated but reached an all time high in 1997. The frequencies are as follows:

Children Killed By Parents or Guardians

1990	1991	1992	1993	1994	1995	1996	1997	1998
34	37	33	33	47	40	45	62	51

Purpose of the Study

The purpose of this exploratory study was to determine if the state of a marital or cohabitation relationship influenced or precipitated the killing of a child(ren) by their parent or guardian. Research findings support for factors, such as separation, divorce and custody concerns was solicited from investigators who handled each case file.

Importance of the Study

It is postulated that if marital or cohabitation relationship instability is found to be a factor, which influences or precipitates parents' killing their own child then:

factual support will be provided regarding the volatility of the situation;
support will be provided for the enhancement of family and police services and the justification of intervention programs; and
additional more in-depth research on the subject will be required to elaborate on the characteristics, develop indicators and review the laws to determine the circumstances that may put children at-risk.

Literature Review

Canadian public and police expressions of concern have confirmed the need to examine more closely the circumstances of child and youth homicides, especially by parents or guardians. This study attempts to identify if there are any linkages to the marital or cohabitation relationship breakdown.

Canadian filicide, neonaticide and infanticide studies are sparse. Filicide and neonaticide studies are closely linked in that both involve the killing of one's own child, but the explanatory framework and the suspect profiles are very different. Following a review of the literature, Logan (1995) reported that mothers who commit neonaticide (killing of newborns within the first 24 hours of life) do so primarily because the child is unwanted, whereas mothers who murder older children do so primarily for altruistic reasons. Neonaticidal mothers were found to be younger, more often unmarried, and less frequently psychotic. As well, he noted that there were two types of neonaticidal mothers. The first group comprised young, often adolescent women who were immature, passive and likely very fearful of the illegitimate pregnancy and the accompanying disapproval or abandonment of loved ones. They often denied the pregnancy and seldom premeditated murder. The second group of mothers was older, more hardened and had strong instinctual drives and little ethical restraint. This group was more likely to premeditate the crime and make more effort to ensure the evidence was destroyed. This researcher postulated that since the destruction of evidence in neonaticide cases was relatively easy, the frequency of neonaticide in Canada was not accurately known (Logan, pp. 8 & 9, 1995).

A Department of Justice Canada 1992 study on domestic homicides involving the use of firearms reported that nine (9) of the 83 incidents resulted in a parent killing one or more of their children. Of the nine shooting incidents, 15 children were killed. More than one child was killed in seven incidents. In two incidents, the mother shot only one child. In all but one case, the accused committed or attempted suicide following the incident (89%). In one-third (33%) of the child homicide incidents, the victim and the accused had taken up separate residence just prior to the murder (Department of Justice Canada, pp. 39 & 41).

Spousal homicides presented different attributes than was shown in domestic homicide cases, resulting in a parent killing their child. Domestic homicides were characterized less often by passionate anger, fuelled by substance abuse and more often, by depression and deliberation, fuelled by economic difficulties or pending separation. The low occurrence of correlating variables indicated that these incidents were the least predictable. The factors that did correlate with a parent killing their child were as follows: credit problems (44%); recently unemployed (22%); under mental health treatment (22%); on unemployment/welfare assistance (22%); serious physical ailment (13%); and involved in a love triangle (11%) (Department of Justice Canada 1992, p.40).

In their study, Silverman & Kennedy (1988) accessed Canadian homicide statistics, 1961 to 1983. The data represented the incidences of female perpetrated homicide (excluding manslaughter and infanticide) for a 23-year period (230 incidents) and infanticide for a 10-year period (45 incidents). The researchers stated that in Canada, when a female kills, there is almost a two-thirds chance that she will kill her spouse or one of her children. As well, the researchers findings showed that when women kill their own children, the sex of the victims is much more evenly divided (about 50% male/female for non infanticide mothers and 57% female in infanticide cases). For non infanticide mothers (1961 to 1983), 31% of the victims were less than 1 year of age and 74% were less than 5 years of age. For infanticide mothers (1974 to 1983), most victims (84%) were under 6 months of age. The motive for the killings varied. Sixty-seven (67%) of the non infanticide mothers and

36% of those involved in infanticide were classified as "mentally ill." This finding was consistent with earlier findings reported in the literature review. For infanticide mothers, 52% of the motives were reported as " for other reasons." An analysis of the method non infanticide mothers kill their child showed that they most often beat their child to death (22%). Also, other methods used by non infanticide mothers are listed as follows: suffocation (16%), drowning (15%) and strangulation (13%). In the case of infanticide, suffocation (27%), beating (12%) and strangulation (12%) were the methods most often used. Researchers have noted that mothers tended to use their hands to kill their offspring. As well, women who committed infanticide did not tend to follow the act with suicide. Non infanticide mothers were married or in common law relationships (71%) whereas, infanticide mothers were single (69%). Only 9% were separated or divorced. Infanticide mothers tended to be young and ill- equipped to mother a child.

The causes of filicide in Canada vary. Bourget and Bradford (1990) studied 13 Canadian cases of parents who killed their children. These researchers reported psycho social stressors and the presence of a crisis as factors precipitating abuse and death. They suggested that most studies have been directed to the universal phenomenon of child abuse. As related to the Canadian culture, the researchers concluded that the killing of a child was viewed to be more serious than the killing of a newborn infant (infanticide). Studies tended to show that a higher proportion of these crimes were perpetrated by mothers. Homicidal behaviour in parents was also associated with common forms of psychiatric disorders and were manifested as the extended suicide phenomenon. Major depression was diagnosed in 30% of the cases and borderline personality disorder in 39%.

Also, these researchers reported that the mean age of child victims of parental homicide was 2.8 years (range 0 to 12 years). More males (54%) than females were killed. A small proportion were the product of unwanted pregnancies. Sixty-nine percent (69%) of the perpetrators were female. Exposure to psycho social stressors appeared to have been a major factor (67%) in most cases of filicide. Stressors were identified and listed as follows: family stress, marital separation or conjugal stress, unwanted or difficult pregnancy stress, disabled or difficult child stress and stress associated with complicated deliveries and serious financial problems.

A Australian study by Strang (1996) examined the characteristics of all homicide incidents involving children as victims. Data from the National Homicide Monitoring Program at the Australian Institute of Criminology revealed that in Australia, from the period of July 1989 to December 1993, there were 108 known child homicide incidents, resulting in the deaths of 126 children under the age of 15 years. In support of earlier studies, this study showed that children were at a higher risk of homicide from family members than from any other category relationship.

When the offender was a parent, the majority of the murders were committed by the father. This findings differed from other studies reported in the literature review. Male reactions to family dispute took one of two courses: rage or depression. As well, parents and guardians often committed or attempted suicide after the murder.

In Strang's study, thirty-five percent (35%) of the children died as a result of a family dispute, usually relating to the termination of their parents's relationship. Of the 23 family dispute incidents, 17 resulted in the deaths of 43 children and also involved the suicide of the offender. Most of the children murdered in a family dispute were 10 years of age or less. Girls were more often victims of homicide than boys. Twenty-five percent (26%) of the sample were victims of abuse causing death. Young children were more often assaulted, whereas older children were killed with a weapon. Several risk factors were identified namely, age of the victim (under 6 months), the age of the offender (usually under 21 years of age), the over representation of non-biological fathers as well as the conditions of their lives considered by the researcher to be unstable and characterized by poverty.

Strang concluded that the overall picture in Australia was very similar to Canada, England and Wales, with the exception of higher numbers of fathers committing the offence than mothers. Also, this researcher pointed out that the fundamental structural changes in society, namely more marriage breakups, often engenders the rage of the offender and thus, engenders depression (Australian Institute of Criminology, March, 1996).

To summarize, these findings show some relationship between the killing of children by a parent and marital or conjugal relationship problems. A 1992 Department of Justice (1992) study reported that in one-third of the incidents, the victim and the accused had taken up separate residences recent to the incident. Bourget and Bradford (1990) listed family stress, marital separation and conjugal stress as factors precipitating the killing of a child. Researchers also stated that in 61.5% of the cases, there were severe psycho social stressors prior to committing filicide (Department of Justice 1992). Since the psycho social stressors were grouped together, it is impossible to determine the percentage attributed to relationship stress. The Australian Institute of Criminology (1996) study further supported this factor by reporting that 35% of the children died as a result of a family dispute.

These research findings show that children are at-risk in a stressful family situation, especially very young and infant children. Related to the purpose of this study, family dispute is characteristic of most parental abduction situations (For further reading on parental abduction, see Kiedrowski, Jayewardene and Dalley, 1994). Such stress may affect a parent or guardian's mental and relationship stability as well as judgement, placing children more at-risk. Therefore, investigators and social service case workers must help protect children by: recognizing the signs associated with these volatile situations, accessing the necessary tools to act, promptly initiating the necessary steps and sharing information with each other.

Prevalence and Characteristics

Various statistical sources provide information about crimes of violence in Canada. In 1988, the Uniform Crime Reporting (UCR) Survey was expanded to provide extended detail on victims and offenders.

According to the 1995 UCR II survey, children (under the age of 12) were much more likely than youths (12 to 19 years) to be victimized by family members, especially parents. In 37% of the violent incidents involving children, the perpetrator was a family member; 25% were parents and 11% were other family members. By contrast, for those cases involving youth deaths, family members were implicated 17% of the time; parents 7% and other family members 10%. The profile differed for male and female victims. Girls were victimized by family members more often than boys in both age categories and all categories of family relationships (Canadian Centre for Justice Statistics, Juristat, Highlights, Vol. 15, No. 15, p.7, 1995).

Victims of Violent Crimes, Children and Youths Victim - Accused Relationship

Victim	Parents	Other family	Total
Child < 12	25%	11%	36%
Youth 12 to 19	7%	10%	17%
Total	32%	21%	53%

The 1997 Canadian Centre for Justice Statistics, Juristat reported that between 1974 and 1996, there were 1,994 victims of homicide under 18 years of age, accounting for 13% of all homicide victims in Canada during this period. This was an average of 87 children each year. As well, the statistics showed that the homicide rate for children under 18 has remained relatively constant, fluctuating between 1.1 and 1.5 homicides per 100,000 children in the population from 1974 to 1996 (p.9). Despite yearly fluctuations, the homicide rate for children has remained relatively constant. In 1996 the child homicide rate was half the rate of adults. A history of violence was considered to be a factor in 26% of all the child homicides committed by a family member (p.1) (Canadian Centre for Justice Statistics, Juristat, Highlights, Vol. 17, No. 11, pp. 1 & 9, 1997).

The 1995 Juristat Highlights showed that the perpetrator in homicides involving children were family members in the majority of solved cases; 35% were fathers, 31% were mothers and eight percent (8%) were other family members. The smallest percentage was committed by strangers (Canadian Centre for Justice Statistics, Juristat, Highlights, Vol. 15, No. 15, page 14, 1995).

Also, that one in five (20%) violent incidents against children and youths occurred at the hands of family members. Parents were implicated in one half of these cases. Girls were more than twice as likely to be victimized by family members. Rates of homicide were lower for children and youths than for adults. Infants and newborns, however, were significantly at a higher risk of homicide than any other group. Many cases of child homicide seemed to be the extension of child abuse (Canadian Centre for Justice Statistics, Juristat, Highlights, Vol. 15, No. 15, page 1, 1995).

The 1997 Canadian Centre for Justice Statistics, Juristat reported that over the period 1974 to 1996 nearly six in ten (58%) child homicides were committed by family members and the majority (79%) of these were committed by a parent. This was also the case with police reported assaults against children, the youngest victims most likely to be killed by family members (usually parents). The greater likelihood for children under 12 to be killed by a family member was consistent for both sexes. Mothers who killed their children were relatively young. Of the females accused of the child homicide during the past decade, 13% were under the age of 18, and 39% were under the age of 25. From 1991 to 1996, there were 18 cases of infanticide which represented 16% of all homicides of victims under 1 year of age. The average age of this group of perpetrators was 20 years (Canadian Centre for Justice Statistics, Juristat, Highlights, Vol. 17, No. 11, p. 9, 1997).

Also, during this 23 year period (1974 to 1996), 7 in 10 family homicides committed against children resulted in at least one charge being laid against an accused. In one quarter of the remaining cases not cleared by a charge, the accused, usually the parent, committed suicide prior to charges being laid in the incident (Canadian Centre for Justice Statistics, Juristat, Highlights, Vol. 17, No. 11, p. 9, 1997).

Regarding the cause of death, the 1995 Juristat Highlights reported there was a difference between child and youth victims. Almost 60% of child homicides resulted from beatings or strangulation. Fourteen percent (14%) of child victims were shot and eight percent (8%) were stabbed to death. In comparison, hand gun and stabbing instruments were used in nearly two thirds (63%) of all cases of youth homicide. Girls were also more likely than boys to be strangled (Canadian Centre for Justice Statistics, Juristat, Highlights, Vol. 15, page 15, 1995).

Limitations of the Study

The information used to locate the cases of child death(s) by a parent or guardian across Canada was generated by Statistics Canada Uniform Crime Reporting Survey (UCR). A request to Statistics Canada by the researcher for the cases yielded only the location of the incident for the years, 1990 to 1993. The investigating police agency housing the case file was not revealed to protect the identity of individuals at the micro level. Thus, an in-depth search ensued for the cases.

A fact-finding questionnaire was designed to gather information from the primary investigator who handled each homicide case. Thus, the gathering of facts was limited to the information supplied in each case file and the investigator's interpretation of those facts. Although the primary investigator was requested to complete the questionnaire, it is suspected that in some instances, the investigator was not located and the questionnaire was completed by other police personnel.

In some incidences, a clear definition of selections might have been included as part of the questionnaire to ensure the meaning of the responses were well understood and thus, attempt to eliminate any uncertainties.

Methodology and Procedure

The Canadian Centre for Justice Statistics supplied the initial search criteria. Of the 137 reported cases of children and youths killed by their parents or guardians, 1990 to 1993, 132 case locations were forwarded to the Missing Children's Registry for research purposes. The information yielded only the location of the case for confidentiality reasons. An in-depth search followed to find the case and investigator who handled the case. Searches in large urban centres presented more difficulty than in smaller cities and rural areas.

After a review of the literature, a questionnaire was designed to gather information from police case files. The questionnaires consisted of closed and open-ended questions. A request was made for the questionnaire to be completed by the primary case investigator. One hundred and thirty two (132) letters and questionnaires were mailed to police agencies within the range of the locations, 93 (70%) were returned. Five were missing critical details and were excluded from the analysis. The results of this study were based upon 88 incidents or cases which resulted in a total of 99 child deaths (See Appendix A for the definitions). Eleven of the cases involved multiple child killings.

Data was collected on the characteristics of the child victims, characteristics of the offender, causes of death, circumstances of death, circumstances after committing the offences, relationship between the victim and offender, relationship status of the primary care giver, divorce, separation and custody details, information on sentencing as well as other related factors.

An analysis of closed and open-ended question responses was made. Many of the questions in this survey allowed for multiple responses, thus the number of responses may be greater than the number of incidents. In addition, the cumulative percentages may be greater than 100%. Percentages quoted in this study for any one question have been adjusted to reflect non-responses.

Data Analysis

Data were collected from incidents which occurred in the 10 provinces and 2 territories of Canada. No incidents were located in the Yukon. For the years 1990 to 1992, the frequency of child homicide was fairly consistent, 82, 83, 86 respectively (Table 1). In 1993, the frequency lowered to 68.

Table 1

Homicide of Children in Canada Frequency 1990 to 1993

	1990	1991	1992	1993	Total
	f	f	f	f	f
No accused	13	12	6	11	42
Killed by parent	34	37	33	33	137 *
Killed by others	35	32	46	24	137
Unknown	0	2	1	0	3
Total	82	83	86	68	319

f = frequency
* 132 locations of incidents were forwarded to the Registry for analysis

In the years that followed this data collection, the incidents of child homicide reached a high of 96 in 1994, 74 in 1995, 83 in 1996, a high of 96 again in 1997 (Table 1A) and most recently a total of 81 in 1998. The total number for 1990 to 1993 was 319 incidents and for 1994 to 1997, 349 incidents.

Table 1A

Homicide of Children in Canada Frequency / Percentage1994 to 1997

	1994	1995	1996	1997	Total
	f	f	f	f	f
No accused	10	8	5	26	49
Killed by parent	47	40	45	62	194
Killed by others	39	26	33	8	106
Unknown	0	0	0	0	0
Total	96	74	83	96	349

f = frequency

Although homicide data collection is considered extremely accurate, researchers have expressed concerns regarding the underestimation of frequencies. In the scoring of the circumstances of death as accidental or intentionally inflicted, some inaccuracies were identified (Silverman et al in 1988 and Logan in 1995).

Comparison by Province

A cumulative breakdown by province of all children killed by a parent was completed. However, since not all questionnaires were returned, Table 2, shows how many questionnaires were returned and not returned. The total shows the frequency of incidents in each province from the years 1990 to 1993. From the data reported, the most incidents were recorded in the provinces of Ontario (44) and Quebec (39). Table 3 shows a provincial breakdown of 88 incidents and the distribution of the 44 questionnaires, not returned. An analysis of the homicide frequencies of 132 incidents showed more incidents in Ontario and Quebec.

Table 2

Child Homicide Incidents / Killed by a Parent By Province By Questionnaire Returned and Not Returned

Police Jurisdiction	Frequency Questionnaire Returned/used	Frequency Questionnaire Not returned/ used	Number of Incidents
	f	f	t	%
Ontario	38	6	44	33
Quebec	15	24	39	30
Alberta	4	9	13	10
B.C.	9	1	10	8
Saskatchewan	7	0	7	5
Manitoba	6	1	7	5
Nova Scotia	4	3	7	5
N.W.T.	2	0	2	2
N.B.	1	0	1	1
P.E.I.	1	0	1	1
Newfoundland	1	0	1	1
Incidents	88	44	132	100

Data was collected from 88 incidents of homicide, whereby a parent or guardian killed a child(ren). The total number of children killed in the 88 incidents during the period 1990 to 1993 was 99. Some incidents involved the death of more than one child (See Table 3).

Table 3

Children Killed by a Parent/Guardian By Province By Incident By Number of Child Deaths

	Ont	Que	BC	Sask	Man	Alb	NS	NWT	NB	PEI	Nfld	t
Incident	38	15	9	7	6	4	4	2	1	1	1	88
Child deaths	42	17	11	8	7	5	4	2	1	1	1	99

Gender and Age of Victims

Of the 99 children killed by a parent(s) or guardian(s), 51% (50) were female and 49% (49) were males. The average age of all child deaths by a parent or guardian was 3.9 years (See Table 4). The majority (43%) of the children killed were less than 1 year of age (infants). A breakdown of the other groups is as follows: preschool children up to 5 years of age, 34%; children 6 and 12 years of age, 18%; and, teenagers, 13 to 18 years of age, 5%. Although infants are usually under the care of the mother, and as shown in this study the at-risk group, the frequency of a parent or guardian killing their infant children was virtually the same for both mothers as it was for fathers.

Table 4

Children Killed by a Parent(s) or Guardian(s) By Comparing Gender, Year, Average Age
1990 to 1993

Year	Female	Male	Total Children	Ave. Age
1990	11	10	21	4.9
1991	15	12	27	3.6
1992	13	13	26	2.6
1993	11	14	25	2.8
Total	50 (51%)	49 (49%)	99 (100%)	3.9

Table 5

Children Killed by a Parent or Guardian By Gender, 1990 to 1993

Offender	Male	Female	Total Children
Father	19	16	35
Step father	1	3	4
Mother	23	23	46
Step mother	0	2	2
Total	43	44	87

Offender's Relationship to the Child

When all age groups were considered, mothers and step mothers killed more children than fathers and step fathers. This was somewhat consistent with the studies cited in the literature review. Almost the same number of boys as girls were killed (See Table 5). This statistic varies from other studies.

In 48% of the 99 child deaths, the offender was either the mother or step mother and in 39% the offender was either the father or step father. The mother and father planned the killing together in 1% of the incidents (See Table 6).

Table 6

Child(ren) Killed by Relationship to Child 1990 to 1993

Parent / Guardian	Frequency	Percentage
Mother	46	46.0
Father	35	35.0
Both mother and father	1	1.0
Step mother	2	2.0
Step father	4	4.0
Boyfriend	9	9.0
Other	2	2.0
Total	99	100.0

Causes of Death

Investigators were asked to identify the causes of death. Of the 99 children killed, 21% were physically abused and 11% died as a result of the Shaken Baby Syndrome. It is interesting to note that if Shaken Baby Syndrome were considered a form of physical abuse, then these findings would show that 32% of the children were physically abused. Other causes of death were: strangulation (21%); shooting (13%); stabbing (12%); drowning (10%); carbon monoxide poisoning (7%); starvation (2%) and drug induced deaths (2%) (See Table 7).

Fathers were more likely to shake or physically abuse the child whereas, mothers tended to strangle or asphyxiate the child to death. Fathers more often used a firearm to kill older children, 10 years of age or more. Mothers used a variety of methods, as listed below, to kill older children. The average age of children killed by the mother was 3 years, younger than those children killed by the father, 5 years. Offenders described as "a boyfriend" and those male offenders living as common-law partners, most often physically abused children to death.

Table 7

Causes of Death by Frequency and Percentage Years 1990 to 1993

Method	Frequency %
Strangulation	21	21.2
Physical Abuse	21	21.2
Shooting/ firearm	13	13.1
Stabbing	12	12.1
Shaken Baby Syndrome	11	11.1
Drowning	10	10.1
Carbon Monoxide Poison	7	7.1
Starvation	2	2.0
Drug induced	2	2.0
Total Child Deaths	99	99.9

Multiple Killings of Children by Parent(s)

Of the 88 homicide incidents, 11 incidents involved a parent or guardian killing more than one child. In 9 of the 11 multiple killings, the offender committed suicide or attempted to commit suicide after the killing. Of the 9 suicide attempts, 6 offenders were successful in taking their own lives. Over half the parents (55%) were separated at the time of the incident.

A total of 22 child deaths occurred in a multiple killing situation , 6 fathers or guardians killed 12 children (7 males and 5 females) and 5 mothers or guardians killed 10 children (2 males and 8 females). The use a firearm to shoot the child was the most common method of killing (36%), followed closely by carbon monoxide poisoning. The average age of children killed by the father was 8 years and by the mother, 4 years (See Tables 8 and 9, noting each incident involves 2 child deaths).

Table 8

Multiple Child Killings by Father by Incident, Age, Sex, Method

	Child 1		Child 2
	Age	Sex	Age	Sex	Method
Incident 1	10	m	8	f	drug OD
Incident 2	1	m	5	f	stabbed
Incident 3	11	m	7	f	shot
Incident 4	7	m	13	f	stabbed
Incident 5	1	m	4	f	carbon monoxide poison
Incident 6	15	m	8	m	shot

f = female m = male

Table 9

Multiple Child Killings by Mother by Incident, Age, Sex, Method

	Child 1		Child 2
	Age in Years	Sex	Age in Years	Sex	Method of Killing
Incident 1	5	f	12	f	shot
Incident 2	4	f	10	m	carbon monoxide poison
Incident 3	2	f	3	f	strangled
Incident 4	2	f	3	f	shot
Incident 5	8	f	5	m	carbon monoxide poison

f = female m = male

Characteristics of the Offender

In analysing the characteristics of the offender, 164 responses to the question yielded the following characteristics: forty-three percent (43%) were viewed as depressed; 11% were under physician's care; and 10% were diagnosed as having a mental illness. In Silverman and Kennedy's study, based on female-perpetrated child homicide cases 1961 to 1983, Women Who Kill Their Children, the researchers reported that when mothers "violate" the maternal role, they are often viewed as being "crazy", thus, biasing the results. This may be a factor in this study also and is worthy of pursuing with further research.

To gather additional information from the investigators regarding the reasons for the killings, an open-ended question was included. The analysis was based upon 301 responses. Multiple responses were permitted in the analysis and the responses were collated and scored by the researcher. The findings revealed that the offender was viewed as emotionally unstable (52%); lacked parenting skills (19%); and planned to revenge the other parent or guardian by killing the children (19%). As well, it was determined that 47% of the responses showed marital discord and/or cohabitation relationship problems existed.

Reasons and Factors for Killing the Child(ren)

Investigators were also asked to check off from a list of factors, derived from an analysis of representative case files, those factors which the investigator determined had an effect on the murder. This analysis was based upon 178 responses. Multiple responses were permitted and the responses were collated and scored by the researcher. The reasons stated were as follows: angry with their spouse or mate (43%); experiencing family discord (38%); and frustration with an extended family situation (28%). The investigators reported that separation and divorce had an effect on the murders in 23% of the incidents. Custody dispute was identified as a factor in 5% of the incidents.

To further understand and attempt to capture the motive(s) for the killings, the investigators were asked to respond to an open-ended question, "What were the reasons as stated by the parent or guardian for killing the child(ren)?" Multiple responses were permitted and the responses were collated and scored by the researcher. Of the 125 responses (relating to 95 of 99 child deaths) the following reasons were identified: marital discord or cohabitation relationship problems (42%); the child(ren) or infant was killed because it was difficult to handle (16%); the killing was accidental (13%); and there was no reason or admission of guilt (20%).

Family Composition

An analysis of 88 incidents of parents or guardians who killed their child(ren) revealed that 50% were separated from their spouse at the time of the incident and 41% were not. No answer was given in 8% of the incidents.

Offenders Actions After Committing the Crime

An analysis was made to determine if the offender called the police or requested assistance. In the analysis of 98 responses to a fact-finding question , 40% of the offenders reported the incident to the police or called for assistance; 26% committed suicide; and, 33% did not contact law enforcement or request assistance. Multiple responses were permitted in this analysis. It was interesting to note that in the literature review some researchers pointed out that infanticide mothers tended not to follow the child death with suicide. Additionally, that the killing an infant was taken less serious by the Canadian society than killing a child (Bourget & Bradford, 1990). These observations require further research study.

Repercussion for Committing Crime

Fifty-one percent of the incidents went to trial. Of the 49% that did not, 34% of the offenders committed suicide.

Summary and Conclusions

The purpose of this exploratory study was to determine if the state of a marital or cohabitation relationship influenced or precipitated the killing of a child by a parent or guardian. Factors such as separation, divorce and custody concerns surfaced. Other profile variables were identified and analysed as well. Eighty-eight (88) questionnaires were analysed to gather information on the circumstances and factors related to the deaths of a child(ren) by a parent or guardian

Findings indicate that from 1990 to 1993, 99 children died in 88 incidents. Eleven of the 88 incidents were multiple killings accounting for 22 child deaths.

A provincial analysis of the 132 incidents revealed that Ontario and Quebec recorded the highest incidents of homicide of children by a parent or guardian.

In terms of absolute risk, age was the clearest factor. The majority of children killed were less than one year of age. Both mothers and fathers killed infant children at the same frequency. Preschool children up to 5 years of age (excluding infants) were the second most likely age group to be killed by their parents. In other words, very young children were definitely more at-risk of victimization.

The average age of all children killed was 4 years. Mothers killed more children than fathers. In both single and multiple killings, the average age of a child killed by the father was older than by the mother. The gender of child homicide victims was evenly distributed.

Twenty-five percent (25%) of children were killed using a weapon (shooting 13% and stabbing 12%), physically abused 21%, strangled 21%, Shaken Baby Syndrome 11% and drowned 10%. In multiple killings, the use of a firearm was more prevalent, closely followed by carbon monoxide poisoning. Mothers strangled or asphyxiated children until death. In comparison, fathers were more physically abusive. Fathers chose to use a firearm to kill older children (over 10 years of age). Mothers used a variety of methods to kill older children.

After killing the children, 59% of the offenders did not contact the police or request emergency assistance, with one quarter of these offenders committing suicide. Forty percent reported the incident or requested emergency assistance.

The offender was most often described as depressed and/or mentally ill. An open- ended question revealed that over half the offenders were viewed by the investigator, as emotionally unstable. This finding is consistent with other research studies.

Investigators scored factors which they determined had an effect on the killings, These were as follows: angry with spouse or mate (43%); family discord (38%); frustrated living in an extended family situation (28%); and separation adjustment problems (23%). Custody dispute was cited as a factor in 5% of the child deaths.

An analysis of an open-ended question on the details of each case showed that in 52% of the incidents, the offender was emotionally unstable and in 47% of the incidents, marital discord and cohabitation relationship problems existed. Other stressors identified were lacking parenting skills (19%) and seeking revenge on the other spouse by the killings (19%).

In summary, this research did find that factors such as family discord, relationship problems and angry with a spouse or mate surfaced often as reasons for the killings. Since all the responses were based upon a file fact review and the investigators interpretation of these facts, some bias may enter the results. However, the researcher noted that these findings were consistent with other research findings in Canada and Australia on the issue.

Based upon the homicidal investigation reports, the findings of this study as well as others cited in the literature review, revealed that children were at-risk when these factors were present: emotional instability; mental illness; angry with a spouse or mate; family discord; relationship problems; and viable stress arising from the problems associated with separation, divorce and occasionally, custody battles. Support was also provided for the correlation between family discord and the killing of children. To lessen the risk to children, new ways must be devised to share information between agencies, including the electronic accessibility to national data collection systems that house such vital information.

Findings from this study support the premise that emotional and relationship instability were factors present in many child deaths by a parent(s) or guardian(s). However, more research was required to determine the extent of the problem, better define the relationship characteristics, develop indicators to assess at-risk family situations and gather facts that will support new intervention and prevention programs. Additionally, to determine if there were circumstances in the law dealing with this issue that may place children more at-risk, for example infanticide and the risk of death of children under 1 year of age.

In conclusion, the results of this study and other studies have shown a need for the collection and analysis of more in-depth data on this issue. This step will enable law makers and enforcers to more fully understand the complexities of the issue, assess the risks to children, review the circumstances in the law that may place children more at-risk and gather facts to support new intervention and prevention programs.

Recommendations

It is recommended that:

more data be collected on the circumstances of the incident, trial and the case disposition;
a standardized questionnaire be completed by police investigators (and/or other sources) when an incident of this nature occurs, especially as it relates to missing children;
a "flag" appear on the Canadian Police Information Centre system database when a child death incident is entered, requesting investigators to locate and complete a questionnaire on this issue and forward it to a predetermined site for analysis;
a review of the laws on child homicide be conducted to determine if there are circumstances in the law that may place children at-risk, especially as the law relates to infants and very young children; and
new and enhanced ways be devised and implemented to exchange information between partners, including the electronic accessibility to national and provincial data collection systems that house relevant and vital information to child safety and well being.

References

Bourget, D. & Bradford J. (1990). Homicidal Parents, Canadian Journal of Psychiatry, Vol. 35, No. 3

Canadian Centre for Justice Statistics, 1995. Juristat, Highlights, Vol. 15, pp. 1, 7, 14, & 15

Canadian Centre for Justice Statistics, 1997. Juristat, Highlights, Vol. 17, No. 11, pp. 1 & 9

Department of Justice Canada, (1992). Domestic Homicides involving the use of Firearms. Dancey Consultants Inc., pp. 39-41

Goetting, A. (1995). Homicide in Families and Other Special Populations. Springer Publishing Company, New York, NY

Holmes, R. & Holmes, S. (1994). Murder in America. Sage Publishing, Thousand Oaks, California

Kiedrowski, J., Jayewardene, C.H.S. and Dalley, M.L. (1994). Parental Abduction of Children: An Overview and Profile of the Abductor. Royal Canadian Mounted Police, Canadian Police Library, Ottawa, Ontario.

Kiedrowski, J., Jayewardene, C.H.S. and Dalley, M.L. (1994). The Police and Parental Abduction: An Overview for Police Direction. Royal Canadian Mounted Police, Canadian Police Library, Ottawa, Ontario

Logan, M. (1995). Mothers Who Murder: A Comparative Study of Filicide and Neonaticide. RCMP Gazette, 57 (7), pp. 2-10

National Center for Missing and Exploited Children (1994). Deceased Child Project, Arlington, Virginia

Rascovsky, A. (1995). Filicide: The Murder, Humiliation, Mutilation, Denigration and Abandonment of Children by Parents. Jason Aronson Inc., Northvale, New Jersey

Silverman, R., Kennedy, L. (1988). Women Who Kill Their Children. Violence and Victims, Vol. 3, No. 2, pp. 113-125

Strang, H. (1996). Children Victims of Homicide. Australian Institute of Criminology, Vol. 53, Canberra, ACT 2601, Australia

Update: National Center for Prosecution of Child Abuse (1996). The Buck Stops Nowhere. American Prosecutors Research Institute's National Center for the Prosecution of Child Abuse, Alexandria, Virginia

Wilson, A.V. (1993). Homicide: The Victim/Offender Connection. Anderson Publishing Co., Cincinnati, Ohio

Appendix A

Historical Perspective and The Laws of Canada

Filicide, the killing of one's child, is deep rooted in the history of mankind. As part of tribal customs or rights of passage, children were victimized and subjected to terrifying experiences at birth.

In early biblical times, the sacrifice of one's children was often viewed as a test of faith. As well as having religious roots, the phenomenon of child sacrifices was a reflection of cultural origins. Earlier societies routinely sacrificed the first born child. In some cultures, female babies were left to die. In other societies, certain ceremonies were held as part of a newborn's acceptance into life. In ancient Greece, for example, a father could reject an unwanted child by disposing of it during the naming and acceptance ceremony (Rascovsky, 1995).

In the 1700's, in the countries of Sweden, Norway and Denmark, a child's life depended on a ceremony similar to baptism. The baby was placed in front of the father (on his lap or the floor). If the baby was to live, the father picked it up and gave it a name and a gift. If it was to be rejected, it was killed immediately (Rascovsky,1995).

When filicide is studied in-depth, tolerance of the act varies from culture to culture in its degree of acceptance. Fortunately, in Western societies, the protection of children is embedded in the laws of the country. Laws that have been enacted to protect children's rights and to punish parents and guardians who kill their children.

The Laws of Canada

Section 233 of the Criminal Code of Canada, Infanticide, 2000, states that a female person commits infanticide when by wilful act or omission, she causes the death of her newly born child, if at the time of the act or omission, she is not fully recovered from the effects of giving birth to the child and by reason thereof or of the effect of lactation consequent on the birth of the new child her mind is disturbed. Every female person who commits infanticide is guilty of an indictable offence and liable to imprisonment for a term not exceeding five years (Martins Annual Criminal Code, 2000).

The Criminal Code of Canada also refers to the law and punishment for the crimes of Abandoning Child, Section 218.; Killing Unborn Child in the Act of Birth, Section 238.; Neglect to Obtain Assistance in Childbirth, Section 242.; Duty of the Persons to Provide the Necessaries, Section 215.; and Concealing the Body of the Child, Section 243. Homicide is explained by the Code as a person who commits homicide when, directly or indirectly, by any means, he causes the death of a human being (Section 222.). Homicide is culpable or not culpable. Homicide that is not culpable is not an offence. Culpable homicide is murder or manslaughter or infanticide. A person commits culpable homicide when he causes the death of a human being, by means of an unlawful act, by criminal negligence, by causing that human being, by threat or fear of violence or by deception, to do anything that causes his death, or, by wilfully frightening that human being, in the case of a child or a sick person (Martins Annual Criminal Code, 2000).

Research Study Definitions

Incident: the actual case file, which recorded investigative information on the circumstances of the child homicide.

Child deaths: the number of children killed.

Response: the answer recorded for a question.

Multiple killings: the killing of more than one child by the offender(s) in the same incident.

Acknowledgements

The National Police Services and the Missing Children's Registry, RCMP would like to acknowledge the assistance and cooperation of the following persons, who provided not only feedback in their area of expertise, but also provided overall comments on the final draft, Dr. Angus Dalley, PhD., Bonita Coole-Stevenson, B.A., M.P.A., and Jessie-Lynn MacDonald, B.A., M.A. (Soc./Psyc). Appreciation is also extended to Robert Morris, B.A., LLB. and Ron Stevenson, M.Phil., LLB., for reviewing the report at various stages and supplying valuable advice.

Appreciation is extended to Victor Stewart, RCMP, Staff Sergeant (retired) who volunteered to help locate the cases and disseminate the questionnaires for the Missing Children's Registry. As well, Francine Piva for typing the final document.

The RCMP would like to acknowledge Statistics Canada for supplying the initial information, making it possible to locate the cases.

Also, a grateful thanks is extended to the dedicated police officers across Canada who took time from their duties to fill out the questionnaire sharing the insights they gained during the investigations.

Marlene Dalley

Royal Canadian Mounted Police

The Killing of Canadian Children by a Parent(s) or Guardian(s):
Characteristics and Trends 1990-1993

by Marlene L. Dalley, Ph.D.

Ottawa, Ontario
CANADA

January 1997 & 2000
Missing Children's Registry
& National Police Services