- Split View
-
Views
-
Cite
Cite
Hannah Bows, Domestic Homicide of Older People (2010–15): A Comparative Analysis of Intimate-Partner Homicide and Parricide Cases in the UK, The British Journal of Social Work, Volume 49, Issue 5, July 2019, Pages 1234–1253, https://doi.org/10.1093/bjsw/bcy108
- Share Icon Share
Abstract
Despite half a century of research on both domestic violence and elder abuse, homicide of older people by a partner or family member (domestic homicide (DH)) remains largely unexplored. This article presents data drawn from a larger parent study examining homicide of older people (aged sixty and over) in the UK. This analysis is based on a subset of cases that would fall within current definitions of DH (n = 221). Analysis reveals differences in DH of older men and women in relation to the perpetrator gender and relationship and differences between intimate-partner homicides and those perpetrated by other family members. Implications for research, theory and practice are discussed.
Introduction
Domestic homicide (DH) is the most extreme form of domestic violence/abuse (DV) and has been identified as a major global public health challenge (Hanlon et al., 2016). There is currently no statutory definition of DH in England and Wales, although the Home Office (2016) describes DHs as the death of a person aged sixteen and over who has died as a result of violence, abuse or neglect by a person whom he or she was related or had been in an intimate personal relationship, or a member of the same household as himself or herself (Home Office, 2016). This broad umbrella term incorporates various contexts and dynamics, but the majority of studies in the UK and elsewhere have focused on homicide by a spouse or partner (often conceptualised as ‘intimate-partner homicide’ (IPH)). Where the homicide involves a family member, the term ‘familial homicide’ is usually employed and, where the DH involves a child (whether over eighteen or under) as the perpetrator, this tends to be conceptualised as ‘parricide’. Consequently, DH research and theory are currently located across multiple fields of inquiry, which have evolved separately. This narrow focus on specific types/sub-types of homicide is not without its problems. As Benbow et al. (2018) point out, the obvious overlaps between the different categories of homicide complicate research and potentially obscure important findings.
Across Europe, DH is the most frequent cause of violent death of women (Ruuskanen and Kauko, 2008). In the year ending March 2016, there were fifty-seven male and 113 DH victims in England and Wales, representing 14 per cent of all male and 65 per cent of all female homicide victims (ONS, 2018b). These figures are similar internationally; for example, data from the USA show that of all DH committed by a spouse, 81 per cent of victims were female (Durose et al., 2005). DHs are therefore overwhelmingly gendered; women are significantly more at risk of being killed by a partner or family member than men. Importantly, a history of DV is a key feature in the majority of cases. Consequently, DH can be located at the extreme end of a continuum of violence experienced by women across the life course, which is both a cause and a consequence of women’s inequality (Kelly and Westmarland, 2014). The characteristics and risk factors observed in DHs are distinctive from other types of homicide and should be considered a unique category (Taylor and Jasinski, 2011).
DV is a pervasive global problem; the World Health Organisation estimates at least one in three women experience intimate-partner violence at some point in their life. In England and Wales, it is estimated that 1.2 million women and 713,000 men experienced some form of DV in the year ending March 2017 (ONS, 2018a). Historically, the focus of DV research, and inter-personal violence more broadly, has been on young women who have been considered most at risk of victimisation. However, a growing body of evidence has shown older people are at risk of abuse (often labelled elder abuse (EA)), with older women most likely to be abused by spouse/partners or adult sons, prompting the development of a new field of inquiry exploring DV against older women (see Roberto et al., 2013a, 2013b for a useful review). This work, although still relatively limited, has pointed towards the need to examine the intersections of gender and age in explaining and understanding DV against older women (e.g. Roberto et al., 2013a, 2013b). Intersectionality, coined in the 1980s by black feminist writers Hooks (1981) and later Crenshaw (1989), was introduced to describe the multiple oppressions black women faced. Whereas feminism is primarily concerned with the importance of gender and the oppression faced in general by women, intersectionality stresses the importance of the interwoven nature of different categories such as race, class and gender, and how they mutually strengthen or weaken each other. In the context of older women, ‘age as a specific organizing principle of intimate relations intersect with other power relations (e.g. gender)’ (Roberto et al., 2013a, p. 233).
Whilst engaging with intersectionality has been encouraged across social work research and practice (e.g. Murphy et al., 2009; Mattsson, 2014), it remains a relatively new concept and age has been neglected in much of the existing work. Despite the growing research and policy interest in understanding and preventing DHs and an equally growing interest in abuse of older people, the intersecting areas have yet to be explored. There are currently no DH (incorporating intimate-partner and familial/parricide) studies utilising national samples exploring the characteristics and contexts where the victim is over sixty.
In the UK context, adult safeguarding (commonly referred to as adult protection in Scotland) is one of the major strands of social work policy and practice (see Bows and Penhale, 2018 for an introductory background to legislation and policy). In England, the Care Act 2014 introduced an overall framework within which adult safeguarding is situated, whilst the 2012 Domestic Violence, Crime and Victims Act (Amendment) locates identifying and responding to DV as multi-agency responsibility. Combined, the legislation and associated policies and guidance create a broad set of statutory duties for adult social workers to identify, investigate and respond to violence and abuse of adults. Two examples include working with stakeholders to assess risk and develop safeguarding plans through Multi-Agency Risk Assessment Conferences and, where violence or abuse (or, in the case of older adults, neglect) leads to death, through involvement in Domestic Homicide Reviews and Safeguarding Adult Reviews.
However, in practice, there is a disconnect between the legislation, policy and practice approaches and social workers remain confused about their role in relation to both EA and DV (Robbins et al., 2016). The existing guidance on EA and DV is distinct, resulting in older victims falling through the gaps (Wydall et al., 2018) due to different pathways created by adult safeguarding and DV policies and practice. This is exacerbated by the paucity of research, and therefore evidence, regarding the extent, nature and consequences of violence against older people.
Following a review of literature to inform understandings of the challenges in addressing the intersection between older age and DH, this article presents the findings of an empirical study designed to extend knowledge and understanding of DHs of people aged sixty and over before considering implications for future research, policy and practice.
DH: prevalence, characteristics and causes
DH is a major global concern. Internationally, the World Health Organisation (2013) estimates that as many as 38 per cent of all murders of women are committed by intimate partners. The main source of data on homicide in England and Wales is provided by the Homicide Index—a national data-set held by the Home Office that contains detailed record-level information about each homicide recorded by police in England and Wales (ONS, 2018b).
The main sources of data on DHs therefore come from academic research and government/public body inquiries and reports, although much of this work draws on small localised samples or relies on data held in medical files. As Brookman (2015) points out, obtaining large and reliable samples of data on homicide is difficult, namely because of the nature of the data (sensitive) and associated difficulties with obtaining access to it. Nevertheless, the existing research has produced consistent findings. The majority of this work has focused on homicides perpetrated by intimate partners (IPH). It is estimated that around 35 per cent of all murders of women globally are perpetrated by an intimate partner, compared to around 5 per cent of men—findings that are consistent across the academic literature (Frye et al., 2005). Moreover, where women are perpetrators, research suggests this is often a result of their own victimisation and constitutes an act of self-defence (Belknap et al., 2012).
Research suggests there are significant age differences between victims and offenders; some studies have reported that men older than forty-five years of age are at the highest risk of perpetrating an IPH (Shackelford and Mouzos, 2005), whilst women aged sixteen to forty-four are at the highest risk of being victimised (ONS, 2018a, 2018b). Generally, IPH most commonly affects women in cohabiting relationships or those going through a divorce or separation (Campbell et al., 2007; Reckdenwald and Parker, 2010). A previous history of DV is one of the most common risk factors; around 70 per cent of men who kill have a history of using violence against their partners (Campbell et al., 2003) and a recent analysis of Monckton-Smith et al. (2017) found over 90 per cent of cases involved prior stalking and/or DV.
Homicide of older people
Violence against older people is gaining increasing research and policy attention internationally. The majority of existing evidence spans two, currently distinctive, fields of inquiry: EA and DV. Generally, EA definitions incorporate physical and non-physical abuse perpetrated by family members or those in relationships where there is an expectation of trust (WHO, 2002). This is viewed as separate to DV definitions, despite the obvious overlaps in terms of behaviours and contexts (for a discussion, see Penhale, 2003; Policastro and Finn, 2017). Operationally, the main feature separating these two disciplines is age: where the victim is forty-five and experiences physical violence from a partner, this is likely to be labelled DV, but, where the victim is aged seventy, this will probably be called EA. The variety of definitions and approaches that exist have been argued to produce a ‘definitional chasos’ (Mysyuk et al., 2013, p. 50) and an ‘ideological gulf’ between those working in DV services and those in aged care (Scott et al., 2004, p. 7).
A recent global systematic review of EA reports much higher figures, estimating that globally at least one in six older people will experience EA (Yon et al., 2017). Other reviews have reported slightly higher rates of one in four (Cooper et al., 2008). However, DV studies have generally yielded higher prevalence rates compared with EA studies. For example, in a five-country European study, Luoma et al. (2011) report that, overall, 28.1 per cent of women aged sixty to ninety-seven had experienced some form of intimate-partner violence in the previous year.
Data published by the ONS (2018) report that, in the year ending March 2017, there were sixty-four homicides of people aged sixty-five and over. There was a disproportionately high number of female victims aged seventy-five and over compared with the population profile (13 per cent of female homicide victims were aged seventy-five and over, whereas 9 per cent of the female population was in this age group). However, further details on the characteristics of victims, offenders and incidents are not broken down by age group and it is not possible to disaggregate the data provided in the supplementary spreadsheets.
Three recent reports have included analysis of older DHs. Sharp-Jeffs and Kelly (2016) included ten cases involving older victims in their recent review of thirty-two DHs. Six of these were IPHs, whilst four were adult-family homicides. The majority of victims were female and all of the perpetrators were male. Most of the offences occurred in the victim’s home and involved a knife or sharp instrument. The review found ageist assumptions led to missed opportunities, as older people were considered at low risk for victimisation by practitioners.
In a recent analysis of thirty years of parricide data (n = 693 cases), Holt (2017) reports that the mean age across male- and female-victim parricides was 60.5, although, on average, women were older than men. Across the 693 cases, there was an even distribution of male/female victims, although Holt (2017) reports that this does not hold across the lifecycle; most men are killed by their son/daughter in their fifties, whereas more women were killed in their seventies. Overall, 37 per cent of female victims were aged seventy and over compared with 29 per cent of male victims. Further data on the location and method of killing are provided for all cases and are not broken down by the age of the victim.
Finally, a recent study by Benbow et al. (2018) examined thirty-one DHs involving adults aged sixty and over. The majority of victims were female (n = 25) and all but one of the perpetrators were male. In fourteen cases, the perpetrator was an adult child or grandchild. The authors identified four key themes across the thirty-one reviews: alcohol and drug misuse, mental health, finance and history of DA. The findings echo the themes observed in DHs across the life course. Consequently, Benbow et al. (2018) conclude that there is insufficient evidence that age is, per se, a significant factor.
Methodology
This article presents data on DHs of older people drawn from a larger parent study examining homicide of people aged sixty and over in the UK. The overarching aim of this analysis was to examine the nature and characteristics of DHs involving a victim aged sixty and over in the UK. The objectives were: (i) to develop an evidence base to address the current gaps in the existing research, (ii) to examine the gender patterns to DHs involving older victims with a view to developing meaningful explanations and (iii) to examine whether DHs of older victims are qualitatively different from DH involving younger victims.
In England and Wales, data on homicide are provided by the Office for National Statistics, which publishes an annual report drawing on data from the Homicide Index. Whilst these data do provide overall totals of homicides by age group, they are not further disaggregated by other personal characteristics (e.g. gender of victims and offenders, location of homicide and relationship between offender and victim). Furthermore, offender age group data are not provided. Consequently, it is not possible to analyse DHs of older people using this data-set. The parent study therefore used the 2000 Freedom of Information (FoI) Act (England, Wales and Northern Ireland) and the 2002 Freedom of Information Act (Scotland) (‘the FoI Acts’) to gather data from police forces in the UK (Bows, 2017). Ethical approval for the study was obtained from Teesside University ethics committee.
The FoI request was sent by e-mail to all forty-nine forces in the UK. The request comprised two parts: the first part asked for aggregated data on the total number of homicide offences recorded between 1 January 2010 and 31 December 2015, broken down by year, and the proportion involving a victim aged sixty or older. The second part asked for demographic and characteristic data of cases involving a victim aged sixty or older, specifically: the gender of the victim and perpetrator; the age of the victim and perpetrator at the time of the offence; the victim–perpetrator relationship (categories such as stranger, acquaintance, partner, friend, family member, carer), the location of the murder (categories such as victim home, perpetrator home, other residential, public place) and the method of killing (categories such as assault with weapon, stabbing, firearm). All forty-nine forces responded to the request, with forty-five forces indicating they had recorded at least one homicide involving an older victim during the study period (n = 514). In 221 cases, the perpetrator was a spouse or other family member, which can be grouped under the category of DHs. This paper presents the data from those cases.
Data were inputted first into an Excel spreadsheet and then coded and inputted into SPSS v.20 for analysis. The small number of cases in some variables meant chi-square and logistical regression tests were not possible in this analysis. The research is presented through descriptive analytical techniques and the cross-tabulation of the characteristics identified to establish a base understanding of their association with each other.
Following the approach adopted by Sharp-Jeffs and Kelly (2016) in recognition of the different dynamics underpinning DV committed by intimate partners and that by family members (Kelly and Westmarland, 2014), the data in this article are split into two sections: IPH and adult-family homicide (AFH).
Findings
Overall, there were 221 cases of DH, representing 43 per cent of the all homicides of older people recorded between 2010 and 2015 in the UK (n = 514). This equates to an average of forty homicides per year in England and Wales, three in Scotland and less than one in Ireland. To put this into perspective, this means that older victims constitute around one in four DHs in England and Wales. Across these 221 cases, the majority of victims were female (67 per cent), consistent with national data on homicide in England and Wales, which report that female victims accounted for 66 per cent of DHs in 2016 (ONS, 2017).
Overall, there were 102 (46 per cent) cases perpetrated by a partner/ex-partner (IPH) and ninety-seven by a child (44 per cent) (parricide). Consequently, older people are almost as likely to be killed by a partner as they are their child—a significant difference compared with DH in younger age groups.
IPH cases
In the majority of IPH cases (n = 102), the victim was female, accounting for 77 per cent of cases (n = 79). Only twenty-three cases involved a male victim.
Overall, most victims were aged sixty to sixty-nine (n = 49, 48 per cent) and there was a steady decline in recorded DHs with increasing age, although, as Table 1 shows, in female-victim homicides, this decline was less pronounced.
. | 60–69 . | 70–79 . | 80–89 . | Total . |
---|---|---|---|---|
Male | 13 | 7 | 3 | 23 |
Female | 36 | 26 | 17 | 79 |
Total | 49 | 33 | 20 | 102 |
. | 60–69 . | 70–79 . | 80–89 . | Total . |
---|---|---|---|---|
Male | 13 | 7 | 3 | 23 |
Female | 36 | 26 | 17 | 79 |
Total | 49 | 33 | 20 | 102 |
. | 60–69 . | 70–79 . | 80–89 . | Total . |
---|---|---|---|---|
Male | 13 | 7 | 3 | 23 |
Female | 36 | 26 | 17 | 79 |
Total | 49 | 33 | 20 | 102 |
. | 60–69 . | 70–79 . | 80–89 . | Total . |
---|---|---|---|---|
Male | 13 | 7 | 3 | 23 |
Female | 36 | 26 | 17 | 79 |
Total | 49 | 33 | 20 | 102 |
The majority of IPH victims were white (64 per cent), although victim ethnicity data were not collected or refused to provide in 31 per cent of cases.
Perpetrator and offence characteristics
Most IPHs were perpetrated by someone of the opposite sex and can therefore be classed as heterosexual IPHs. As Table 2 shows, where the victim was female (seventy-nine), male perpetrators accounted for seventy-eight (99 per cent) of cases and, where the victim was male, 91 per cent of perpetrators were female. This finding was statistically significant ×2 (1, n = 102) 85.37, p = 0.000.
. | Male perpetrator . | Female perpetrator . | Total . |
---|---|---|---|
Male victim | 2 | 21 | 23 |
Female victim | 78 | 1 | 79 |
Total | 80 | 22 | 102 |
. | Male perpetrator . | Female perpetrator . | Total . |
---|---|---|---|
Male victim | 2 | 21 | 23 |
Female victim | 78 | 1 | 79 |
Total | 80 | 22 | 102 |
. | Male perpetrator . | Female perpetrator . | Total . |
---|---|---|---|
Male victim | 2 | 21 | 23 |
Female victim | 78 | 1 | 79 |
Total | 80 | 22 | 102 |
. | Male perpetrator . | Female perpetrator . | Total . |
---|---|---|---|
Male victim | 2 | 21 | 23 |
Female victim | 78 | 1 | 79 |
Total | 80 | 22 | 102 |
Although the number of same-sex IPHs were therefore higher in male victims, the very small number of cases (two) does not allow any meaningful comparisons or conclusions to be drawn.
Perpetrators were aged between twenty and ninety-nine. As Figure 1 shows, the majority were aged between sixty and sixty-nine (n = 36, 35 per cent). The second most common perpetrator age group was seventy to seventy-nine (25 per cent) and eighty to eighty-nine (22 per cent). Only one perpetrator was aged ninety to ninety-nine. Perpetrators of IPH were therefore similarly aged to their victims.
Overall, the majority of IPH victims were killed in their home (incorporating victim home and victim and perpetrator shared home) (n = 88, 84 per cent). As Figure 2 shows, very few offences were committed outside of the home. However, in six cases, the homicide occurred in a public outside location; although a small number, men were statistically more likely to be killed in a public place (n = 2, 9 per cent) compared with women (n = 4, 6 per cent).
In the majority of cases, the cause of homicide was stabbing (36 per cent) followed by strangulation (16 per cent) and assault with a weapon (11 per cent). However, there were differences in the cause of homicide between male and female victims; only one male victim was strangled (4 per cent of male victims) compared to fifteen female victims (19 per cent). Six cases involved the use of a firearm; in those cases, five of the victims were female.
AFH
Of the 221 cases involving a family member other than a spouse, 119 (54 per cent) were committed by a child, grandchild or other family member. In most cases, the perpetrator was a child or grandchild (n = 97, 81 per cent), which brings these offences under current definitions of ‘parricide’. These are reported separately to the remaining twenty cases involving other adult family members.
Parricide victim characteristics
In the majority of cases (n = 57, 59 per cent), the victim was female. As Table 3 shows, male victims accounted for forty of the cases (41 per cent). In contrast to the IPH cases, there was a more even split in the age groups of victims, with the same number aged sixty to sixty-nine (n = 33, 34 per cent) and seventy to seventy-nine (n = 33, 34 per cent).
. | 60–69 . | 70–79 . | 80–89 . | 90–99 . | 100 and over . | Total . |
---|---|---|---|---|---|---|
Male | 14 | 15 | 8 | 2 | 1 | 40 |
Female | 19 | 18 | 15 | 3 | 2 | 57 |
Total | 33 | 33 | 23 | 5 | 3 | 97 |
. | 60–69 . | 70–79 . | 80–89 . | 90–99 . | 100 and over . | Total . |
---|---|---|---|---|---|---|
Male | 14 | 15 | 8 | 2 | 1 | 40 |
Female | 19 | 18 | 15 | 3 | 2 | 57 |
Total | 33 | 33 | 23 | 5 | 3 | 97 |
. | 60–69 . | 70–79 . | 80–89 . | 90–99 . | 100 and over . | Total . |
---|---|---|---|---|---|---|
Male | 14 | 15 | 8 | 2 | 1 | 40 |
Female | 19 | 18 | 15 | 3 | 2 | 57 |
Total | 33 | 33 | 23 | 5 | 3 | 97 |
. | 60–69 . | 70–79 . | 80–89 . | 90–99 . | 100 and over . | Total . |
---|---|---|---|---|---|---|
Male | 14 | 15 | 8 | 2 | 1 | 40 |
Female | 19 | 18 | 15 | 3 | 2 | 57 |
Total | 33 | 33 | 23 | 5 | 3 | 97 |
Where victim ethnicity was provided (n = 74), most victims were ‘white’ (82 per cent), although a higher proportion of parricides involved an Asian victim compared with IPHs (n = 8, 11 per cent).
Parricide perpetrator and offence characteristics
The majority of parricide perpetrators were sons or grandsons, accounting for eighty cases (82 per cent). Table 4 shows that, even where the victim was male (n = 40, 41 per cent), the majority of perpetrators were also male (n = 36, 90 per cent). In the seventeen cases involving a daughter or granddaughter, the majority of victims were also female (n = 13, 76 per cent).
. | Male perpetrator . | Female perpetrator . | Total . |
---|---|---|---|
Male victim | 36 | 4 | 40 |
Female victim | 44 | 13 | 57 |
Total | 80 | 17 | 97 |
. | Male perpetrator . | Female perpetrator . | Total . |
---|---|---|---|
Male victim | 36 | 4 | 40 |
Female victim | 44 | 13 | 57 |
Total | 80 | 17 | 97 |
. | Male perpetrator . | Female perpetrator . | Total . |
---|---|---|---|
Male victim | 36 | 4 | 40 |
Female victim | 44 | 13 | 57 |
Total | 80 | 17 | 97 |
. | Male perpetrator . | Female perpetrator . | Total . |
---|---|---|---|
Male victim | 36 | 4 | 40 |
Female victim | 44 | 13 | 57 |
Total | 80 | 17 | 97 |
As Figure 3 highlights, parricide perpetrator ages ranged from under sixteen to eighty-nine years. Overall, the majority were aged under sixty (n = 91, 94 per cent); most were between forty and forty-nine (n = 33, 33 per cent), followed by thirty to thirty-nine (n = 24, 26 per cent).
As with IPHs, the majority of AFH occurred in the victim and/or perpetrator’s home (n = 90, 93 per cent). As Figure 4 shows, the most common method of killing was stabbing, accounting for 40 per cent (n = 38) of cases. However, unlike the IPHs, the second most common method was assault without a weapon (beating to death) (n = 20, 21 per cent). In contrast to the IPHs, few cases involved strangulation (n = 3). There were important differences in the gender of victims and method of killing; men were more likely than women to be killed by assault without a weapon (thirteen men compared to seven women) whereas women were more likely to be killed by assault with a weapon (ten women compared to three men), although these results were not statistically significant.
Other family homicides
In twenty-two cases, the perpetrator was an ‘other family member’. In these cases, ten victims were female compared to twelve male victims. The perpetrators in these cases were overwhelmingly male (n = 20, 91 per cent). As with the other categories of homicide, most victims were aged sixty to sixty-nine (n = 11, 50 per cent). Reflecting the other categories, the majority of homicides occurred in the victim’s home (n = 19) and involved the use of a knife or sharp object (n = 16).
Discussion
This paper presents findings from the first study to specifically examine DHs involving older men and women in the UK. Whilst a body of research has examined DH previously, there are very few studies examining gender differences in DHs and there have been no published studies specifically examining, and comparing, DHs of older men and women. Similarly, although a small number of studies have examined homicide of older people more generally, none has incorporated specific analysis of DHs. The data published here begins to address these gaps in evidence.
Overall, there were 221 DHs involving a victim aged sixty and over recorded by the police between 2010 and 2015. This equates to roughly 44 per year, which means that approximately one in four DHs involve an older victim. This finding is consistent with a recent report of DHs by Sharp-Jeffs and Kelly (2016), who reported six of the twenty-four Domestic Homicide Reviews (DHRs) they analysed involved a victim aged sixty or older. Therefore, although older people constitute a smaller proportion of all DHs and the risk of DH decreases with age, the numbers are not insignificant and do not justify the absence of research in this area. To put this into context, rates of DH of older adults are similar to that of child homicide; national data reveal there were fifty-eight homicides involving a child aged from birth to fifteen years in 2017, equating to 9 per cent of all homicide recorded that year (ONS, 2018a, 2018b). Yet, as Benbow et al. (2018) point out, child homicide has received significantly more research and media attention than the homicide of older people (although a recent article in this Journal indicates research on children’s involvement in DH remains limited; see Stanley et al., 2018). Moreover, the figure presented here is unlikely to be wholly accurate, as concerns have been raised previously that homicides of older people may go undetected due to the small proportion of elderly deaths that involve a post-mortem. Elderly deaths may be assumed to be related to older age and less likely to be questioned.
The findings presented here suggest the risk factors for DH in later life are gender-specific and share some similarities and differences with existing understandings of DH against younger age groups. Mirroring international data, most victims in this study were female (67 per cent). Across all the DHs, the majority occurred in the victim’s home (87 per cent) and a sharp instrument or knife was the most common method of killing (41 per cent). Most victims were ‘white’, although the lack of available data on ethnicity is problematic and prevents comprehensive analysis of the specific circumstances of homicide among different ethnic groups.
In terms of differences, the victim age and relationship to the perpetrator differ compared with national profiles of DH. Consistently with previous elder DH research, the ‘young-old’ were the most common age group of victims in this study (Sharp-Jeffs and Kelly, 2016; Benbow et al., 2018) and the numbers decreased with increased age, although this was more pronounced for men than women, who had a higher risk of victimisation in older age groups compared to men.
A similar number of homicides were committed by a partner (n = 102) and adult child (n = 97), supporting a recent study by Benbow et al. (2018) that analysed thirty-one DHRs and reported that AFH was more prominent than IPH in the cases they examined. In the present study, the majority of AFH perpetrators were adult sons (n = 89, 92 per cent). This is a significant finding, as it highlights that the risk factors for younger adults may be different to older adults. However, there were important differences in the relationships between female and male victims and perpetrators. Female victims were generally killed by spouse/partners (53 per cent) followed by sons/daughters (38 per cent), whereas male victims were more likely to be killed by sons/daughters (53 per cent) than by spouses/partners (31 per cent). These findings suggest that men are at more risk of being killed by their children or grandchildren (parricide) than they are by their partners—a finding that is consistent with the existing literature (Holt, 2017; Benbow et al., 2018). However, it is important to note that the overall risk of parricide for older people is much higher for women. Fifty women were killed by a son/daughter compared with thirty-nine men, confirming recent findings by Holt (2017) that older women are at a higher risk of being a victim of parricide than older men. Nevertheless, this finding highlights the need for interventions to prevent homicide in older age must be sensitive to the specific dynamics of DH.
Implications for future research and practice
In terms of research, there is a need to revisit conceptual and theoretical understandings of inter-personal violence towards an integrated, life-course framework that incorporates existing DV/intimate-partner violence, EA and adolescence-to-parent violence. This research has identified that the dynamics of DH in later life are gendered; women remain at a higher risk of DH compared with men, although the almost equal numbers of spouse/partner homicides and adult son/grandson homicides (often termed parricide) bring into sharp relief the need to recognise that age is also critical in understanding violence against older women.
Equally, although older men are at lower risk than women for DH, they are more likely to be murdered by a son/grandson than a spouse/partner, again highlighting the need for an integrated framework that takes these dynamics into account. Scholars should consider whether intersectionality could also be adopted to incorporate age, class, sexualities and other characteristics to develop more comprehensive understandings of violence (including fatal violence) against older people. This theory has already been used to examine the intersections of age and gender in relation to DV (Roberto et al., 2013a) and how race, gender and class interact in female homicides (Parker and Hefner, 2015). Although the findings in the present indicate most victims were ‘white’, the lack of available ethnicity data prevents further analysis on the specific circumstances relating to the intersections of gender, age and ethnicity. Given the importance of examining the matrix of social categories as highlighted by Parker and Hefner (2015), further research that investigates multiple social categories is required.
Similarly, methodologies must extend to incorporate age as well as gender (and other identities). The current dichotomies and disjoints that exist between the different disciplines (e.g. EA, eldercide, DV, DH, adolescent-to-parent violence and parricide) and different forms of violence result in a limited picture of inter-personal violence and homicide, complicating research and hindering the development of a comprehensive evidence base. Both national and international sources of data on violent victimisation must specifically include older victims to enable a life-course analysis of victimisation. In the context of DHs, one way of achieving this is by making all DHRs and Safeguarding Adult Reviews public via a database, as called for by previous researchers (e.g. Sharp-Jeffs and Kelly, 2016; Benbow et al., 2018).
These conceptual and methodological frameworks must extend to policy and practice. DH almost never occurs out of the blue (Monckton-Smith et al., 2014); there is almost always a history of DV and research has consistently shown prior victim engagement with health or statutory services prior to their murder (Adams, 2007; Juodis et al., 2014). As such, these services are uniquely placed to identify women (and men) who may be at risk and provide early intervention and support. However, the majority of the understandings of DH and subsequent service responses have been based on DV perpetrated by intimate partners and the links between this violence and DH. In a recent analysis of DH reviews, Sharp-Jeffs and Kelly (2016) report that older women are often assumed to be at low risk for victimisation and there is a general lack of awareness of the signs of DV and risk factors for DH in later life. This may be partially because the signs are different, such as the abuse is being perpetrated by a child rather than a spouse. Similarly, Benbow et al. (2018) report that few DH reviews of elder homicides identified a history of DV. We concur with Sharp-Jeffs and Kelly (2016, p. 13) that ‘adult social services should receive training on identifying and assessing risk in relation to DV. Training should take an intersectional approach and explore the multiple barriers and increased risk faced by particular groups’.
Furthermore, older people are less likely to be engaging with services that have been trained to recognise signs of DV in younger women (e.g. midwives, health visitors). Ageist stereotypes and narrow understandings of DV mean older victims are often overlooked or the issues are subsumed under ‘elder abuse’ discourses and policies. It is critical that services that are likely to be in contact with older people (including age-related organisations, health and social care services and adult safeguarding social workers) are aware of the signs and risk factors for DH,. We concur with Policastro and Finn (2017) that we need to integrate DV/IPV and EA fields to develop new programmes with older victims specifically in mind.
In England and Wales, legislation (e.g. the 2014 Care Act and the 2012 Domestic Violence, Victims and Crime Act (Amendment)) and associated national guidelines (e.g. Local Government Association and Association Directors of Adult Social Services, 2013; NICE, 2014) place responsibility for preventing and responding to DV (and, by virtue, DH) within a multi-agency framework that incorporates social work, criminal justice, and health and social care. In England and Wales, one of the most important and widely used risk-assessment tools used to identify those at risk of serious physical harm, including homicide, is via the Co-ordinated Action Against Domestic Abuse (CAADA) Domestic Abuse, Stalking and Honour-base violence (DASH) Risk Identification Checklist. However, this risk assessment currently focuses on young victims/offenders; for example, many of the questions in the risk-assessment tool focus on pregnancy and/or young children—issues less likely to affect older adults. In fact, the only time older age is mentioned is in the context of risk to other family members; one of the questions asks whether the perpetrator has been violent or abusive to other family members including elderly parents/relatives. Again, where the victim is elderly themselves, they are less likely to have elderly parents. Moreover, the risk assessments are currently designed around intimate partners as perpetrators and most of the questions would fail to capture adult sons or grandsons as perpetrators, despite current definitions of DV capturing these dynamics. In order to ensure social workers are actively seeking to reduce power imbalances and oppression, it is critically important that risk-assessment tools and safeguarding policies fully incorporate older adults. The Older People’s Commissioner for Wales (2015) has introduced an amended Risk Identification Checklist that includes questions aimed at older victims that may provide a starting point for amending risk assessments in other parts of the UK.
This study is not without limitations. The small number of cases involving older people makes statistical analysis on these cases more difficult; traditional statistical testing for relationships and correlations between variables was not possible using this data-set. Some data, such as victim ethnicity, were not available for a significant proportion of cases. Moreover, as the variables captured in this study were based on victim, perpetrator and incident characteristics, background data on the victim and offender (e.g. economic status and class, health/mental health conditions, alcohol/drug misuse and offending history) are not examined in this study. Further research is now required to examine DHs of older people in more detail. Despite the limitations, the findings of the current research have important implications for theory, policy, practice and future research.
Conclusion
Despite increased research and policy interest in the abuse of older people, available data are scarce and limited by both conceptual and methodological problems. The existing research on different forms of inter-personal violence spans multiple subcategories, making analysis difficult. This is further compounded by varying definitions and fields of study examining these forms of violence involving an older adult. Research on the extent and consequences of violence against older people remains limited, although current estimates suggest risk of experiencing abuse in later life is similar to that in younger age groups and the dynamics are broadly the same.
Given the rapidly ageing (Western) population, the number of older people being abused is likely to increase and, given that the primary risk factor for DH is previous abuse, DH involving older victims is also likely to increase. There is a need to develop a more unified approach to examining violence against older people, including fatal violence. From the data presented here, it would appear that feminist theories (particularly intersectionality) of DV and DH are most appropriate to developing understandings of lethal and non-lethal violence against older women, given the similarities with younger populations reported here. Multi-agency working must also extend to include those working in related disciplines and industries, including age-related organisations, health and social care.
Acknowledgements
The author would like to thank the anonymous reviewers for their constructive and helpful feedback on earlier versions of this manuscript.
Funding
The author received no funding for this research.
Conflict of interest statement. None declared.
References
Home Office (
Local Government Association and Association Directors of Adult Social Service (
NICE (
Older People’s Commissioner for Wales (
ONS (
ONS (
World Health Organisation (
World Health Organisation (