Domestic violence at its extreme results in domestic homicide. Data published by the Australian Institute of Criminology for the period 1 July 2003 to 30 June 2012 reported that intimate partners accounted for 23% of all homicide victims, with females typically being the victims in intimate partner homicides (N = 488; 75%), and males accounting for the majority of offenders (Cussen & Bryant, Reference Cussen and Bryant2015). Subsequently, Bryant and Bricknell (Reference Bryant and Bricknell2017) reported that for the period July 2012 to June 2014, there were 487 recorded homicides, with the most common relationship between offender and victim being a domestic violence relationship (n = 200; 41%).
Supporting these figures, the Domestic Violence Death Review Team (2018) reported that between 2010 and 2014, there were 152 intimate partner homicides which followed an identifiable history of domestic violence. The majority of these homicides involved a male killing a current or former partner (N = 121; 79%). Of the 152 homicides examined, at least 107 children under the age of 18 years survived the death of one or both of their parents. Recent data from European sources provide a similar picture. Stanley et al. (Reference Stanley, Chantler and Robbins2019) undertook an analysis of Domestic Homicide Reviews between 2011 and 2016 in England, revealing a subsample of 55 cases involving 125 child survivors. A similar review in the Netherlands between 2003 and 2012 revealed that 256 children lost a biological parent due to 137 cases of intimate partner homicides (Alisic et al., Reference Alisic, Groot, Snetsalaar, Stroeken and van de Putte2017).
These data highlight a relatively unnoticed and under-researched aspect of domestic homicide – that of the child survivor. Burman and Allen-Meares (Reference Burman and Allen-Meares1994) pointed out that it was primarily when the children were victims themselves, for example, through physical or sexual abuse, that they became the concern of welfare or mental health professionals, but that they tended to be overlooked when the violence occurred between the parents. With attention typically being focused on the deceased and the perpetrator of the crime, the children inadvertently became the neglected victims. In their review of 146 children who experienced the murder of their mother by an intimate partner, Lewandowski and colleagues subsequently noted that ‘We know very little about these children’ (Lewandowski et al., Reference Lewandowski, McFarlane, Campbell, Faye and Barenski2004, p. 212).
Proximal effects
An early indication of the adverse emotional impacts was provided by Malmquist (Reference Malmquist1986), who assessed 16 children between the ages of 5 and 10 years who had witnessed a parental murder. In approximately half of these cases, the children were present when a family member attempted to kill all members of the family, and where the children survived by hiding or running. There is no indication of how long after the events children were seen, or under what circumstances the children were referred. However, the researcher reported that at the time of assessment all children easily met the criteria for post-traumatic stress disorder (PTSD). Anxiety and/or nightmares were present in all the children; restlessness and increased alertness was prominent among many; and varying degrees of school difficulties appeared in the form of trouble concentrating and memory impairment. Signs and symptoms of a major affective disorder were also present, with many children exhibiting persistent disturbances in mood, a general lack of enthusiasm for most things and the emergence of psychosomatic complaints in 10 of the 16 children.
A more detailed account of the range of adverse impacts on these children was reported in a study of 95 children seen at the Traumatic Stress Clinic in London by Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000), following the murder of their mothers. Forty per cent of the children were under 5 years of age at the time of the killing, 41% were between the ages of 6 and 11 years, and 18% were teenagers. In about half of the cases, the children had witnessed previous domestic violence, one-third witnessed the actual killing and a further 10% saw the body of their mother immediately afterwards.
In terms of emotional functioning at the time of referral, the authors reported there was a strong association between witnessing the homicide and the development of PTSD, with moderate-to-severe PTSD affecting some 25% of the children. In addition, approximately 60% of the children had behavioural problems in the moderate-to-severe range, and 40% displayed a range of emotional disorders. Echoing Malmquist (Reference Malmquist1986), these researchers also reported academic deterioration in just over half the school-attending children and a quarter of the children developed health problems that were assessed as being psychosomatic.
Emotional functioning is also affected by a range of multiple losses. Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000) pointed out that in losing a parent, these children actually lose both parents; their mother as a murder victim, and their father arrested and in jail or also dead from a murder-suicide. In addition, children may also lose their home, pets, neighbourhood and school as they are relocated to live with relatives or placed into foster care. These children are, therefore, left with a feeling of their world being suddenly thrown out of control and a profound uncertainty about their future. While the provision of a safe and stable placement would seem a basic necessity for the beginning of recovery, in the experience of Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000), this did not always eventuate.
These researchers pointed out that both sets of relatives may elect to care for the children. Where there were family members (e.g., grandparents or an uncle or aunt willing to take the children), the move to a new family, even if familiar to the children, invariably meant new routines, a new neighbourhood and a new school, with all the attendant uncertainties. This was not always guaranteed, however, and where there were no extended family members immediately available, the children had to be taken into care. While such a placement would ideally need to accommodate all the children, the researchers recounted instances where this was not possible, or where there were half-siblings who were claimed by members of their genetic family, resulting in a further sense of loss and distress for all the children, especially if the children saw their step-siblings as integral to their family, or if older children had taken some of the responsibility of parenting the younger children.
In addition, what the children were told about the homicide depended in part upon the age of the children, what they may have witnessed, and who told them. The researchers recounted that some of the children, particularly young children, had simply been told that their mother had gone away without further explanation, while other children had not been told that their father was in jail but that, for example, he had gone away for work. Other children may not have been given an honest account of the situation for fear of upsetting them or, in the case of the perpetrator’s family who may be having difficulties in coming to terms with the actions of their son/brother, may attempt to rationalise his actions or seek to place the blame elsewhere. The authors added that, should the paternal side have successfully gained residence of the children, which occurred in 15% of the cases, there may be little control over whether or not the children have contact with the father – either while he is still in jail or upon his release – and how he, himself, may explain his actions.
Of the 95 children assessed in the Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000) study, 61 of them were able to be followed-up in order to gain some understanding of their adjustment over time (Kaplan et al., Reference Kaplan, Black, Hyman and Knox2001). Kaplan et al. (Reference Kaplan, Black, Hyman and Knox2001) reported that at follow-up approximately one-third of the children were living with members of the maternal family, 15% with the paternal family and the remainder in some form of care. While children placed with maternal family members had the most stable placements and were doing better in terms of their general emotional functioning, children in care tended to be less settled with approximately 50% of these children having more than one change of placement. The authors also commented that, as a trend, those children living with the paternal family, and especially those children who returned to live with the perpetrator upon his release from prison, were less likely to be showing improvements.
With respect to emotional and behavioural functioning, Kaplan et al. (Reference Kaplan, Black, Hyman and Knox2001) reported that, at follow-up, approximately one-third of the children continued to have significant problems. Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000) considered that at the time of referral, those children who appeared to cope best were those who received longer therapeutic input. It was, therefore, particularly disappointing, in the view of Kaplan et al. (Reference Kaplan, Black, Hyman and Knox2001), that fewer than half of the children had received further therapy after leaving the Traumatic Stress Clinic, with the authors adding that those children living with their father were least likely to receive therapeutic follow-up.
A number of similarities to the above studies were found in the Lewandowski et al. (Reference Lewandowski, McFarlane, Campbell, Faye and Barenski2004) study of 146 children in the USA who had experienced the murder of their mothers. Thirty-five per cent of the children in this study were under four years of age at the time of the killing, 41% were between the ages of 5 and 10 years, and 17% were between 12 and 14 years. Of the 121 mothers killed, the researchers reported that 49% were murder-suicides. In addition, in 67% of the homes, there had been prior physical assault of the mother, and in about one-quarter of the families the perpetrator had threatened the entire family in various ways. The researchers reported that in 35% of the cases a child actually witnessed the murder, and in a further 37% of the cases, a child found the victim’s body.
Although Lewandowski et al. (Reference Lewandowski, McFarlane, Campbell, Faye and Barenski2004) did not report on the emotional functioning of the children at the time of assessment, the authors noted that the children’s lives were significantly disrupted following the murder, with 87% of the children having to move home; 47% to live with the maternal family, 12% with the perpetrator’s family, 10% split between families, and 9% into foster care. Similar to the Kaplan et al. (Reference Kaplan, Black, Hyman and Knox2001) observation about the number of children receiving follow-up counselling, Lewandowski et al. (Reference Lewandowski, McFarlane, Campbell, Faye and Barenski2004) considered that few of the children received appropriate therapy, with almost one-quarter receiving no counselling at all, and approximately one-fifth of the children who actually witnessed the murder or who found the body never seeing a professional counsellor about the event.
Distal effects
While there are relatively little data on child survivors of domestic homicide, the existing evidence is consistent in acknowledging the traumatic impact this experience has on their functioning in all areas of their lives. Evidence from the above studies indicates that over half the children could be living with domestic violence prior to the murder, between 35% and 50% of the children witnessed the killing, and that there were significant levels of emotional and behavioural difficulties among this group of children as a consequence, including symptoms of post-traumatic stress, anxiety, depression and the development of a number of psychosomatic difficulties. In addition, the majority of these children were uprooted from their homes and neighbourhoods and sent to live with relatives or were taken into foster care.
To what extent these difficulties may persist over time, and how children and their caregivers may adjust to the murder of a family member was explored by Hardesty et al. (Reference Hardesty, Campbell, McFarlane and Lewandowski2008). These researchers interviewed 10 caregivers of children who had lost a mother to domestic homicide and, at the time of interview, five of these children were living with the maternal family, one with the paternal family, and the other four with what was described as other family members. The children had a mean age of 11.3 years at the time of the murder, and the caregivers were interviewed at an average of 2.8 years post-event. With respect to the children’s adjustment, all caregivers reported a range of overlapping mental health, behavioural, physical and academic problems. Mental health concerns were reported by seven caregivers, and included symptoms of post-traumatic stress like sleep problems, persistent nightmares and fear of the dark, as well as anxiety, depression and prolonged grief. In addition, physical/psychosomatic complaints were reported by six caregivers, with headaches and stomach aches being the most common.
In an attempt to better understand the stressors acting upon these children and their caregivers, Hardesty et al. (Reference Hardesty, Campbell, McFarlane and Lewandowski2008) examined both the pre- and post-event circumstances in which the children found themselves. Pre-event stressors included the fact that all 10 children had been previously exposed to domestic violence, and that five children had witnessed the murder of their mother (in which two of these were a murder-suicide). In a further two cases, the children were present in the house, did not witness the murder, but found the mother’s body.
With respect to post-event stressors, the authors found co-victims (i.e., adult family members of the homicide victim) suffered a range of their own negative health problems. For example, the authors reported that within seven months of the homicide, two caregivers had suffered heart attacks, and a another caregiver was hospitalised with a heart condition. In addition, the children in one family lost both maternal grandparents within two years of the homicide. Thus, adult family members who suddenly become caregivers must attempt to manage their own trauma and grief reactions in addition to providing for the needs of traumatised children. Frequently, there are additional financial strains on these families, particularly if the children are taken in by grandparents or relatives who already have a number of children to care for.
Finally, echoing Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000), Hardesty et al. (Reference Hardesty, Campbell, McFarlane and Lewandowski2008) noted that recovery for co-victims may be further complicated by estranged family relationships and conflict between the victim’s and perpetrator’s families. Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000) had previously observed that both sets of relatives may claim the children in service of their own emotional needs: it was proposed that for the maternal family, the need involved the resolution of grief and mourning, while for the father’s family, it involved the resolution of shame and guilt – or in some cases the rationalising of the killing. Children can therefore be caught up in unresolved family conflicts; what Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000) referred to as ‘marital conflict by proxy’ (p. 94).
Similar to findings of previous studies, Hardesty et al. (Reference Hardesty, Campbell, McFarlane and Lewandowski2008) reported that children in 9 of the 10 families moved to new homes and communities after the event, and were thus removed from their old support networks and the familiarity of friends and school. The authors commented on the general lack of professional supports available, either because such supports were not readily available, or caregivers were not able to access them, and felt it problematic how little long-term support these families received.
Further longer-term aspects relating to the recovery of child survivors of domestic homicide were provided by Steeves and colleagues in their study of 34 adults interviewed up to 40 years after the event (Steeves et al., Reference Steeves, Parker, Laughton, Knopp and Thompson2011). The participants were aged between 12 and 19 years (mean age = 14 years) at the time of the homicide, and all participants reported that the killings were in the context of ongoing domestic violence. Fourteen participants reported witnessing the killing, while a further three reported finding the body of their mother.
In discussing the event with the researchers, it was evident that the participants were still affected by their experience of the homicide. For many, describing the event brought back vivid and disturbing memories, a major theme being of blood, for example, seeing their mother covered in blood or lying in a pool of blood. For some, there were additional memories of police and media, and of then re-experiencing the homicide on TV.
Most participants reported moving in with extended family members, although two participants who were in their late teens at the time recalled being left alone and unsupported in the family home until they had to leave because of unpaid rent. For those who moved in with relatives, the placement was not necessarily safe or stable, as some participants recalled difficulties with integrating into the new family, while two female participants reported being sexually abused by relatives. Some participants reported ending up homeless and in shelters as a consequence, some reported being suicidal at times, and some began using drugs and alcohol as a means of coping.
As adults, most of the participants considered that the most lasting effect on them was in establishing and maintaining intimate relationships. Many reported having had a number of unsuccessful relationships, although few described being involved in violent relationships. For many participants, parenting their own children was also a major issue, with one participant describing how she lost her children because of her own instability. The authors mentioned that some participants had refused to become parents because they did not want to run the risk of subjecting children to what they had experienced in their families.
Steeves et al. (Reference Steeves, Parker, Laughton, Knopp and Thompson2011) accepted that their findings were limited by the nature of the sample group, speculating that those participants who volunteered to take part may have coped better with their traumatic past. They pointed out that participants in the study spoke of siblings who were incarcerated, had committed suicide or who had otherwise died young, were involved in substance abuse, or who had serious mental health problems. Few such individuals had participated in the study. Nevertheless, the authors considered that given the paucity of information about the long-term consequences of experiencing domestic homicide in childhood, their data could provide a useful description of the lives of some of these survivors and help shape future policy to guide those who may become involved with clients who have experienced domestic homicide.
Recent reviews
In contrast to our understanding of the effects on children of living with domestic violence (e.g., Devaney, Reference Devaney2015; Holt et al., Reference Holt, Buckley and Whelan2008), our understanding of the effects on child survivors of domestic homicide is relatively poor. Alisic et al. (Reference Alisic, Groot, Snetsalaar, Stroeken and van de Putte2015) remarked on how little research had been conducted on this population of children, with the majority of this research coming from either the USA or the UK, and with most data collected before 2000. These authors reviewed 17 articles referencing the area of children and domestic homicide with a view to focusing on children’s mental health and wellbeing, following the loss of a parent through intimate partner homicide.
With respect to childhood trauma, Alisic et al. (Reference Alisic, Krishna, Groot and Frederick2015) pointed out that the literature has traditionally focused on PTSD as an outcome. Nevertheless, and consistent with previous findings in this area (e.g., Hardesty et al., Reference Hardesty, Campbell, McFarlane and Lewandowski2008; Malmquist, Reference Malmquist1986), these authors reported evidence suggesting substantial variation in child mental health and wellbeing outcomes, encompassing the inter-related domains of psychological, social, physical and academic effects. Further, and possibly more importantly, these authors sought to better understand variations in outcome by examining a range of potential risk and protective factors that may serve to mediate effects on a particular child. These risk and protective factors were further grouped into pre-, peri- and post-homicide characteristics.
As examples, Alisic et al. (Reference Alisic, Krishna, Groot and Frederick2015) listed child exposure to pre-homicide domestic violence, parental substance abuse and unstable living conditions as being pre-trauma characteristics that posed risks for poorer outcome. Peri-traumatic risk factors included suicide by the perpetrator, and whether the child was present at the time of the homicide, although early psychological intervention as well as broader services intervention, and the child’s opportunity to say goodbye, for example, by going to the funeral, were listed as being potential protective factors promoting better outcomes. Finally, post-trauma characteristics such as the breakdown of placement and splitting of siblings were seen as risk factors for the child. In addition, supporting Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000), these authors prioritised caregiver distress, problematic contact with the perpetrating parent, conflict between relatives and lack of mental health care, as being particularly important risk factors for the child in the post-homicide period.
While pointing out that there were a large number of factors potentially influencing outcomes in child survivors of domestic homicide, Alisic et al. (Reference Alisic, Krishna, Groot and Frederick2015) cautioned that the evidence base is not strong for most of these factors at this point in time. Nevertheless, in considering care after domestic homicide, Alisic et al. (Reference Alisic, Groot, Snetsalaar, Stroeken and van de Putte2017) pointed to a number of clinical implications that should be addressed by professionals. These included the exploration of previous domestic violence experiences, the child’s exposure to the homicide and crime scene, traumatic grief and the extent of disruption of daily life and its developmental implications.
The diversity in factors and outcomes reported in the literature also underlined the importance of listening to children in considering interventions, as well as strong collaboration among all professionals involved in the care of this population of children. Stanley et al. (Reference Stanley, Chantler and Robbins2019) expressed concern that most Domestic Homicide Reviews gave scant consideration to children’s needs for current or future support, and pointed to instances of professionals involved with families caught up in domestic violence failing to talk to children directly about their experiences. While acknowledging that maintaining a focus on children could be a challenge for practitioners employed in essentially adult services, these authors recommended the need for such practitioners to be provided with training and assessment tools that directed their attention onto children, and that gave them the knowledge of resources that identified need and appropriate referrals. Given that children’s needs for support were likely to be ongoing and may need to be revised over time, these authors also considered that a multi-agency perspective was helpful in illuminating relationships between different family members’ needs, for example, between children and their caregivers, and needs across service divides.
Conclusions
As Burman and Allen-Meares (Reference Burman and Allen-Meares1994) reminded us, ‘Of all the traumatic events that children can experience, none can be more horrific than witnessing the murder of one parent by another’ (p. 28). Given that attention is, initially at least, focused on the victim and the perpetrator of the crime, the surviving children can inadvertently become the neglected victims. The publications documenting this area tend to underline the observation of Lewandowski et al. (Reference Lewandowski, McFarlane, Campbell, Faye and Barenski2004) and supported more recently by Alisic et al. (Reference Alisic, Groot, Snetsalaar, Stroeken and van de Putte2017), that we know very little about these children; in particular their pre-homicide family circumstances, and their life circumstances and recovery in the years following. Regardless of the relative paucity of research on child survivors of domestic homicide, some conclusions can nevertheless be drawn.
First, those studies that made some form of assessment of the child’s functioning at the time of referral were consistent in finding a range of overlapping mental health, behavioural, physical, and academic problems. Perhaps not surprisingly, there was a particularly strong relationship between witnessing the murder or finding the body, and the development of symptoms of post-traumatic stress. Despite the extensive range of difficulties documented in these populations of children, there was also the general comment about the lack of therapeutic services available and/or the number of children who did not receive follow-up therapeutic care.
Second, where children ended up after the event seemed, in many instances, to be somewhat ad hoc. Hardesty et al. (Reference Hardesty, Campbell, McFarlane and Lewandowski2008) regarded a stable relationship with a caring and consistent adult to be a protective factor. However, as Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000) pointed out, this did not always eventuate due to a number of factors, for example, conflict over the children between extended families, or no extended family members available or willing to take the children and the children having to be placed into care. Steeves et al. (Reference Steeves, Parker, Laughton, Knopp and Thompson2011) also reported on instances where there were difficulties with integration into the new family, or child survivors being abused in their new placements. Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000) challenged the assumption that relatives, because they may be familiar with the children, are automatically the best people to provide care and believed instead that a balancing act is needed between the advantages and disadvantages of living with relatives or non-relatives, given that such care needs to be secure and long-lasting.
Third, and directly related to the above, co-victims, that is, adult family members of the homicide victim, may also need their own support. Their capacity to cope with a range of stressors in the immediate aftermath, as well as into the longer term, will also impinge upon child survivors in their care. It was, therefore, imperative, in the view of Hardesty et al. (Reference Hardesty, Campbell, McFarlane and Lewandowski2008) as well as Stanley et al. (Reference Stanley, Chantler and Robbins2019), that such families have access to comprehensive and ongoing services that are relevant to their needs. These authors argued that the complexity of the needs of both the children and the families who are caring for them point to the requirement for long-term supports that aim to foster the best outcomes for both.
Lastly, Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000) reminded us that there is still a prevalent assumption that the troubles of childhood pass, that children are resilient and that they forget. These authors were clear, however, that this is not so, and pointed out that the scars of childhood can last into adulthood and run the risk of affecting both their own welfare as well as the welfare of others. It was therefore their view that, whatever agencies became involved in families following a domestic homicide, they have the knowledge and training, the service structure, and the resources to be able to implement the support necessary for the best long-term outcome for these children and their caregivers.
Domestic violence at its extreme results in domestic homicide. Data published by the Australian Institute of Criminology for the period 1 July 2003 to 30 June 2012 reported that intimate partners accounted for 23% of all homicide victims, with females typically being the victims in intimate partner homicides (N = 488; 75%), and males accounting for the majority of offenders (Cussen & Bryant, Reference Cussen and Bryant2015). Subsequently, Bryant and Bricknell (Reference Bryant and Bricknell2017) reported that for the period July 2012 to June 2014, there were 487 recorded homicides, with the most common relationship between offender and victim being a domestic violence relationship (n = 200; 41%).
Supporting these figures, the Domestic Violence Death Review Team (2018) reported that between 2010 and 2014, there were 152 intimate partner homicides which followed an identifiable history of domestic violence. The majority of these homicides involved a male killing a current or former partner (N = 121; 79%). Of the 152 homicides examined, at least 107 children under the age of 18 years survived the death of one or both of their parents. Recent data from European sources provide a similar picture. Stanley et al. (Reference Stanley, Chantler and Robbins2019) undertook an analysis of Domestic Homicide Reviews between 2011 and 2016 in England, revealing a subsample of 55 cases involving 125 child survivors. A similar review in the Netherlands between 2003 and 2012 revealed that 256 children lost a biological parent due to 137 cases of intimate partner homicides (Alisic et al., Reference Alisic, Groot, Snetsalaar, Stroeken and van de Putte2017).
These data highlight a relatively unnoticed and under-researched aspect of domestic homicide – that of the child survivor. Burman and Allen-Meares (Reference Burman and Allen-Meares1994) pointed out that it was primarily when the children were victims themselves, for example, through physical or sexual abuse, that they became the concern of welfare or mental health professionals, but that they tended to be overlooked when the violence occurred between the parents. With attention typically being focused on the deceased and the perpetrator of the crime, the children inadvertently became the neglected victims. In their review of 146 children who experienced the murder of their mother by an intimate partner, Lewandowski and colleagues subsequently noted that ‘We know very little about these children’ (Lewandowski et al., Reference Lewandowski, McFarlane, Campbell, Faye and Barenski2004, p. 212).
Proximal effects
An early indication of the adverse emotional impacts was provided by Malmquist (Reference Malmquist1986), who assessed 16 children between the ages of 5 and 10 years who had witnessed a parental murder. In approximately half of these cases, the children were present when a family member attempted to kill all members of the family, and where the children survived by hiding or running. There is no indication of how long after the events children were seen, or under what circumstances the children were referred. However, the researcher reported that at the time of assessment all children easily met the criteria for post-traumatic stress disorder (PTSD). Anxiety and/or nightmares were present in all the children; restlessness and increased alertness was prominent among many; and varying degrees of school difficulties appeared in the form of trouble concentrating and memory impairment. Signs and symptoms of a major affective disorder were also present, with many children exhibiting persistent disturbances in mood, a general lack of enthusiasm for most things and the emergence of psychosomatic complaints in 10 of the 16 children.
A more detailed account of the range of adverse impacts on these children was reported in a study of 95 children seen at the Traumatic Stress Clinic in London by Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000), following the murder of their mothers. Forty per cent of the children were under 5 years of age at the time of the killing, 41% were between the ages of 6 and 11 years, and 18% were teenagers. In about half of the cases, the children had witnessed previous domestic violence, one-third witnessed the actual killing and a further 10% saw the body of their mother immediately afterwards.
In terms of emotional functioning at the time of referral, the authors reported there was a strong association between witnessing the homicide and the development of PTSD, with moderate-to-severe PTSD affecting some 25% of the children. In addition, approximately 60% of the children had behavioural problems in the moderate-to-severe range, and 40% displayed a range of emotional disorders. Echoing Malmquist (Reference Malmquist1986), these researchers also reported academic deterioration in just over half the school-attending children and a quarter of the children developed health problems that were assessed as being psychosomatic.
Emotional functioning is also affected by a range of multiple losses. Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000) pointed out that in losing a parent, these children actually lose both parents; their mother as a murder victim, and their father arrested and in jail or also dead from a murder-suicide. In addition, children may also lose their home, pets, neighbourhood and school as they are relocated to live with relatives or placed into foster care. These children are, therefore, left with a feeling of their world being suddenly thrown out of control and a profound uncertainty about their future. While the provision of a safe and stable placement would seem a basic necessity for the beginning of recovery, in the experience of Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000), this did not always eventuate.
These researchers pointed out that both sets of relatives may elect to care for the children. Where there were family members (e.g., grandparents or an uncle or aunt willing to take the children), the move to a new family, even if familiar to the children, invariably meant new routines, a new neighbourhood and a new school, with all the attendant uncertainties. This was not always guaranteed, however, and where there were no extended family members immediately available, the children had to be taken into care. While such a placement would ideally need to accommodate all the children, the researchers recounted instances where this was not possible, or where there were half-siblings who were claimed by members of their genetic family, resulting in a further sense of loss and distress for all the children, especially if the children saw their step-siblings as integral to their family, or if older children had taken some of the responsibility of parenting the younger children.
In addition, what the children were told about the homicide depended in part upon the age of the children, what they may have witnessed, and who told them. The researchers recounted that some of the children, particularly young children, had simply been told that their mother had gone away without further explanation, while other children had not been told that their father was in jail but that, for example, he had gone away for work. Other children may not have been given an honest account of the situation for fear of upsetting them or, in the case of the perpetrator’s family who may be having difficulties in coming to terms with the actions of their son/brother, may attempt to rationalise his actions or seek to place the blame elsewhere. The authors added that, should the paternal side have successfully gained residence of the children, which occurred in 15% of the cases, there may be little control over whether or not the children have contact with the father – either while he is still in jail or upon his release – and how he, himself, may explain his actions.
Of the 95 children assessed in the Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000) study, 61 of them were able to be followed-up in order to gain some understanding of their adjustment over time (Kaplan et al., Reference Kaplan, Black, Hyman and Knox2001). Kaplan et al. (Reference Kaplan, Black, Hyman and Knox2001) reported that at follow-up approximately one-third of the children were living with members of the maternal family, 15% with the paternal family and the remainder in some form of care. While children placed with maternal family members had the most stable placements and were doing better in terms of their general emotional functioning, children in care tended to be less settled with approximately 50% of these children having more than one change of placement. The authors also commented that, as a trend, those children living with the paternal family, and especially those children who returned to live with the perpetrator upon his release from prison, were less likely to be showing improvements.
With respect to emotional and behavioural functioning, Kaplan et al. (Reference Kaplan, Black, Hyman and Knox2001) reported that, at follow-up, approximately one-third of the children continued to have significant problems. Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000) considered that at the time of referral, those children who appeared to cope best were those who received longer therapeutic input. It was, therefore, particularly disappointing, in the view of Kaplan et al. (Reference Kaplan, Black, Hyman and Knox2001), that fewer than half of the children had received further therapy after leaving the Traumatic Stress Clinic, with the authors adding that those children living with their father were least likely to receive therapeutic follow-up.
A number of similarities to the above studies were found in the Lewandowski et al. (Reference Lewandowski, McFarlane, Campbell, Faye and Barenski2004) study of 146 children in the USA who had experienced the murder of their mothers. Thirty-five per cent of the children in this study were under four years of age at the time of the killing, 41% were between the ages of 5 and 10 years, and 17% were between 12 and 14 years. Of the 121 mothers killed, the researchers reported that 49% were murder-suicides. In addition, in 67% of the homes, there had been prior physical assault of the mother, and in about one-quarter of the families the perpetrator had threatened the entire family in various ways. The researchers reported that in 35% of the cases a child actually witnessed the murder, and in a further 37% of the cases, a child found the victim’s body.
Although Lewandowski et al. (Reference Lewandowski, McFarlane, Campbell, Faye and Barenski2004) did not report on the emotional functioning of the children at the time of assessment, the authors noted that the children’s lives were significantly disrupted following the murder, with 87% of the children having to move home; 47% to live with the maternal family, 12% with the perpetrator’s family, 10% split between families, and 9% into foster care. Similar to the Kaplan et al. (Reference Kaplan, Black, Hyman and Knox2001) observation about the number of children receiving follow-up counselling, Lewandowski et al. (Reference Lewandowski, McFarlane, Campbell, Faye and Barenski2004) considered that few of the children received appropriate therapy, with almost one-quarter receiving no counselling at all, and approximately one-fifth of the children who actually witnessed the murder or who found the body never seeing a professional counsellor about the event.
Distal effects
While there are relatively little data on child survivors of domestic homicide, the existing evidence is consistent in acknowledging the traumatic impact this experience has on their functioning in all areas of their lives. Evidence from the above studies indicates that over half the children could be living with domestic violence prior to the murder, between 35% and 50% of the children witnessed the killing, and that there were significant levels of emotional and behavioural difficulties among this group of children as a consequence, including symptoms of post-traumatic stress, anxiety, depression and the development of a number of psychosomatic difficulties. In addition, the majority of these children were uprooted from their homes and neighbourhoods and sent to live with relatives or were taken into foster care.
To what extent these difficulties may persist over time, and how children and their caregivers may adjust to the murder of a family member was explored by Hardesty et al. (Reference Hardesty, Campbell, McFarlane and Lewandowski2008). These researchers interviewed 10 caregivers of children who had lost a mother to domestic homicide and, at the time of interview, five of these children were living with the maternal family, one with the paternal family, and the other four with what was described as other family members. The children had a mean age of 11.3 years at the time of the murder, and the caregivers were interviewed at an average of 2.8 years post-event. With respect to the children’s adjustment, all caregivers reported a range of overlapping mental health, behavioural, physical and academic problems. Mental health concerns were reported by seven caregivers, and included symptoms of post-traumatic stress like sleep problems, persistent nightmares and fear of the dark, as well as anxiety, depression and prolonged grief. In addition, physical/psychosomatic complaints were reported by six caregivers, with headaches and stomach aches being the most common.
In an attempt to better understand the stressors acting upon these children and their caregivers, Hardesty et al. (Reference Hardesty, Campbell, McFarlane and Lewandowski2008) examined both the pre- and post-event circumstances in which the children found themselves. Pre-event stressors included the fact that all 10 children had been previously exposed to domestic violence, and that five children had witnessed the murder of their mother (in which two of these were a murder-suicide). In a further two cases, the children were present in the house, did not witness the murder, but found the mother’s body.
With respect to post-event stressors, the authors found co-victims (i.e., adult family members of the homicide victim) suffered a range of their own negative health problems. For example, the authors reported that within seven months of the homicide, two caregivers had suffered heart attacks, and a another caregiver was hospitalised with a heart condition. In addition, the children in one family lost both maternal grandparents within two years of the homicide. Thus, adult family members who suddenly become caregivers must attempt to manage their own trauma and grief reactions in addition to providing for the needs of traumatised children. Frequently, there are additional financial strains on these families, particularly if the children are taken in by grandparents or relatives who already have a number of children to care for.
Finally, echoing Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000), Hardesty et al. (Reference Hardesty, Campbell, McFarlane and Lewandowski2008) noted that recovery for co-victims may be further complicated by estranged family relationships and conflict between the victim’s and perpetrator’s families. Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000) had previously observed that both sets of relatives may claim the children in service of their own emotional needs: it was proposed that for the maternal family, the need involved the resolution of grief and mourning, while for the father’s family, it involved the resolution of shame and guilt – or in some cases the rationalising of the killing. Children can therefore be caught up in unresolved family conflicts; what Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000) referred to as ‘marital conflict by proxy’ (p. 94).
Similar to findings of previous studies, Hardesty et al. (Reference Hardesty, Campbell, McFarlane and Lewandowski2008) reported that children in 9 of the 10 families moved to new homes and communities after the event, and were thus removed from their old support networks and the familiarity of friends and school. The authors commented on the general lack of professional supports available, either because such supports were not readily available, or caregivers were not able to access them, and felt it problematic how little long-term support these families received.
Further longer-term aspects relating to the recovery of child survivors of domestic homicide were provided by Steeves and colleagues in their study of 34 adults interviewed up to 40 years after the event (Steeves et al., Reference Steeves, Parker, Laughton, Knopp and Thompson2011). The participants were aged between 12 and 19 years (mean age = 14 years) at the time of the homicide, and all participants reported that the killings were in the context of ongoing domestic violence. Fourteen participants reported witnessing the killing, while a further three reported finding the body of their mother.
In discussing the event with the researchers, it was evident that the participants were still affected by their experience of the homicide. For many, describing the event brought back vivid and disturbing memories, a major theme being of blood, for example, seeing their mother covered in blood or lying in a pool of blood. For some, there were additional memories of police and media, and of then re-experiencing the homicide on TV.
Most participants reported moving in with extended family members, although two participants who were in their late teens at the time recalled being left alone and unsupported in the family home until they had to leave because of unpaid rent. For those who moved in with relatives, the placement was not necessarily safe or stable, as some participants recalled difficulties with integrating into the new family, while two female participants reported being sexually abused by relatives. Some participants reported ending up homeless and in shelters as a consequence, some reported being suicidal at times, and some began using drugs and alcohol as a means of coping.
As adults, most of the participants considered that the most lasting effect on them was in establishing and maintaining intimate relationships. Many reported having had a number of unsuccessful relationships, although few described being involved in violent relationships. For many participants, parenting their own children was also a major issue, with one participant describing how she lost her children because of her own instability. The authors mentioned that some participants had refused to become parents because they did not want to run the risk of subjecting children to what they had experienced in their families.
Steeves et al. (Reference Steeves, Parker, Laughton, Knopp and Thompson2011) accepted that their findings were limited by the nature of the sample group, speculating that those participants who volunteered to take part may have coped better with their traumatic past. They pointed out that participants in the study spoke of siblings who were incarcerated, had committed suicide or who had otherwise died young, were involved in substance abuse, or who had serious mental health problems. Few such individuals had participated in the study. Nevertheless, the authors considered that given the paucity of information about the long-term consequences of experiencing domestic homicide in childhood, their data could provide a useful description of the lives of some of these survivors and help shape future policy to guide those who may become involved with clients who have experienced domestic homicide.
Recent reviews
In contrast to our understanding of the effects on children of living with domestic violence (e.g., Devaney, Reference Devaney2015; Holt et al., Reference Holt, Buckley and Whelan2008), our understanding of the effects on child survivors of domestic homicide is relatively poor. Alisic et al. (Reference Alisic, Groot, Snetsalaar, Stroeken and van de Putte2015) remarked on how little research had been conducted on this population of children, with the majority of this research coming from either the USA or the UK, and with most data collected before 2000. These authors reviewed 17 articles referencing the area of children and domestic homicide with a view to focusing on children’s mental health and wellbeing, following the loss of a parent through intimate partner homicide.
With respect to childhood trauma, Alisic et al. (Reference Alisic, Krishna, Groot and Frederick2015) pointed out that the literature has traditionally focused on PTSD as an outcome. Nevertheless, and consistent with previous findings in this area (e.g., Hardesty et al., Reference Hardesty, Campbell, McFarlane and Lewandowski2008; Malmquist, Reference Malmquist1986), these authors reported evidence suggesting substantial variation in child mental health and wellbeing outcomes, encompassing the inter-related domains of psychological, social, physical and academic effects. Further, and possibly more importantly, these authors sought to better understand variations in outcome by examining a range of potential risk and protective factors that may serve to mediate effects on a particular child. These risk and protective factors were further grouped into pre-, peri- and post-homicide characteristics.
As examples, Alisic et al. (Reference Alisic, Krishna, Groot and Frederick2015) listed child exposure to pre-homicide domestic violence, parental substance abuse and unstable living conditions as being pre-trauma characteristics that posed risks for poorer outcome. Peri-traumatic risk factors included suicide by the perpetrator, and whether the child was present at the time of the homicide, although early psychological intervention as well as broader services intervention, and the child’s opportunity to say goodbye, for example, by going to the funeral, were listed as being potential protective factors promoting better outcomes. Finally, post-trauma characteristics such as the breakdown of placement and splitting of siblings were seen as risk factors for the child. In addition, supporting Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000), these authors prioritised caregiver distress, problematic contact with the perpetrating parent, conflict between relatives and lack of mental health care, as being particularly important risk factors for the child in the post-homicide period.
While pointing out that there were a large number of factors potentially influencing outcomes in child survivors of domestic homicide, Alisic et al. (Reference Alisic, Krishna, Groot and Frederick2015) cautioned that the evidence base is not strong for most of these factors at this point in time. Nevertheless, in considering care after domestic homicide, Alisic et al. (Reference Alisic, Groot, Snetsalaar, Stroeken and van de Putte2017) pointed to a number of clinical implications that should be addressed by professionals. These included the exploration of previous domestic violence experiences, the child’s exposure to the homicide and crime scene, traumatic grief and the extent of disruption of daily life and its developmental implications.
The diversity in factors and outcomes reported in the literature also underlined the importance of listening to children in considering interventions, as well as strong collaboration among all professionals involved in the care of this population of children. Stanley et al. (Reference Stanley, Chantler and Robbins2019) expressed concern that most Domestic Homicide Reviews gave scant consideration to children’s needs for current or future support, and pointed to instances of professionals involved with families caught up in domestic violence failing to talk to children directly about their experiences. While acknowledging that maintaining a focus on children could be a challenge for practitioners employed in essentially adult services, these authors recommended the need for such practitioners to be provided with training and assessment tools that directed their attention onto children, and that gave them the knowledge of resources that identified need and appropriate referrals. Given that children’s needs for support were likely to be ongoing and may need to be revised over time, these authors also considered that a multi-agency perspective was helpful in illuminating relationships between different family members’ needs, for example, between children and their caregivers, and needs across service divides.
Conclusions
As Burman and Allen-Meares (Reference Burman and Allen-Meares1994) reminded us, ‘Of all the traumatic events that children can experience, none can be more horrific than witnessing the murder of one parent by another’ (p. 28). Given that attention is, initially at least, focused on the victim and the perpetrator of the crime, the surviving children can inadvertently become the neglected victims. The publications documenting this area tend to underline the observation of Lewandowski et al. (Reference Lewandowski, McFarlane, Campbell, Faye and Barenski2004) and supported more recently by Alisic et al. (Reference Alisic, Groot, Snetsalaar, Stroeken and van de Putte2017), that we know very little about these children; in particular their pre-homicide family circumstances, and their life circumstances and recovery in the years following. Regardless of the relative paucity of research on child survivors of domestic homicide, some conclusions can nevertheless be drawn.
First, those studies that made some form of assessment of the child’s functioning at the time of referral were consistent in finding a range of overlapping mental health, behavioural, physical, and academic problems. Perhaps not surprisingly, there was a particularly strong relationship between witnessing the murder or finding the body, and the development of symptoms of post-traumatic stress. Despite the extensive range of difficulties documented in these populations of children, there was also the general comment about the lack of therapeutic services available and/or the number of children who did not receive follow-up therapeutic care.
Second, where children ended up after the event seemed, in many instances, to be somewhat ad hoc. Hardesty et al. (Reference Hardesty, Campbell, McFarlane and Lewandowski2008) regarded a stable relationship with a caring and consistent adult to be a protective factor. However, as Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000) pointed out, this did not always eventuate due to a number of factors, for example, conflict over the children between extended families, or no extended family members available or willing to take the children and the children having to be placed into care. Steeves et al. (Reference Steeves, Parker, Laughton, Knopp and Thompson2011) also reported on instances where there were difficulties with integration into the new family, or child survivors being abused in their new placements. Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000) challenged the assumption that relatives, because they may be familiar with the children, are automatically the best people to provide care and believed instead that a balancing act is needed between the advantages and disadvantages of living with relatives or non-relatives, given that such care needs to be secure and long-lasting.
Third, and directly related to the above, co-victims, that is, adult family members of the homicide victim, may also need their own support. Their capacity to cope with a range of stressors in the immediate aftermath, as well as into the longer term, will also impinge upon child survivors in their care. It was, therefore, imperative, in the view of Hardesty et al. (Reference Hardesty, Campbell, McFarlane and Lewandowski2008) as well as Stanley et al. (Reference Stanley, Chantler and Robbins2019), that such families have access to comprehensive and ongoing services that are relevant to their needs. These authors argued that the complexity of the needs of both the children and the families who are caring for them point to the requirement for long-term supports that aim to foster the best outcomes for both.
Lastly, Harris-Hendriks et al. (Reference Harris-Hendriks, Black and Kaplan2000) reminded us that there is still a prevalent assumption that the troubles of childhood pass, that children are resilient and that they forget. These authors were clear, however, that this is not so, and pointed out that the scars of childhood can last into adulthood and run the risk of affecting both their own welfare as well as the welfare of others. It was therefore their view that, whatever agencies became involved in families following a domestic homicide, they have the knowledge and training, the service structure, and the resources to be able to implement the support necessary for the best long-term outcome for these children and their caregivers.