The “Pathways to intimate partner homicide” project: Key stages and events in male-perpetrated intimate partner homicide in Australia

Dr Hayley Boxall

Research Manager, Australian Institute of Criminology

Laura Doherty

Research Analyst, Australian Institute of Criminology

Dr Siobhan Lawler

Senior Research Analyst, Australian Institute of Criminology

Christie Franks

Former Research Analyst, Australian Institute of Criminology

Dr Samantha Bricknell

Research Manager, Australian Institute of Criminology

This report addresses work covered in the ANROWS research project “The ‘Pathways to intimate partner homicide’ project: Key stages and events in male-perpetrated intimate partner homicide in Australia”. Please consult the ANROWS website for more information on this project.

ANROWS research contributes to the six National Outcomes of the National Plan to Reduce Violence against Women and their Children 2010–2022. This research addresses National Plan Outcome 6 – Perpetrators stop their violence and are held to account.

Suggested citation:

Boxall, H., Doherty, L., Lawler, S., Franks, C., & Bricknell, S. (2022). The “Pathways to intimate partner homicide” project: Key stages and events in male-perpetrated intimate partner homicide in Australia (Research report, 04/2022). ANROWS.

Australian Institute of Criminology

GPO Box 1936
Canberra ACT 2601

Author acknowledgement

We are grateful for the support and insights provided by the PIPH Advisory Group, comprised of Sarah Burr, Dr Li Erikkson, Associate Professor Kate Fitz-Gibbon, Dr Rae Kaspiew, Professor David Keatley, Dr Manjula O’Connor, Kuljeeta Singh, Dr Brian Sullivan and Tahlia-Rose Vanissum. Your amazing insights and support for this project, as well as your willingness to contribute your time and energy, helped us to understand the huge dataset that we created. We also acknowledge the valuable support provided by the J.V. Barry Library who conducted literature review searches and sourced the material for us. In particular, we would like to thank Megan Whittle for assisting in identifying sentencing remarks for cases included in the final sample from various public databases. The authors acknowledge the Victorian Department of Justice and Community Safety as the source organisation of the National Coronial Information System data, and the National Coronial Information System as the source of coronial data used in this report. The authors further acknowledge the Supreme Court Library Queensland, the Supreme Court of Western Australia, the Courts Administration Authority of South Australia and the Courts Library of the Northern Territory for providing access to sentencing remarks for the homicides referred to throughout this report. Finally, we acknowledge the women who have died in Australia, and the loved ones that were left behind.

Acknowledgement of lived experiences of violence

ANROWS acknowledges the lives and experiences of the women and children affected by domestic, family and sexual violence who are represented in this report. We recognise the individual stories of courage, hope and resilience that form the basis of ANROWS research.
Caution: Some people may find parts of this content confronting or distressing. Recommended support services include 1800RESPECT (1800 737 732), Lifeline (13 11 14) and Benestar’s Aboriginal and Torres Strait Islander Support Line (1800 816 152).

Contents

Acronyms

Definitions and concepts

Executive summary

Methods

Results

Discussion and implications for policy and practice

Conclusion

Introduction

Methods

Incident selection

Data coding

Analysis

Limitations

Final sample

Results

Pathway 1: Fixated threat

Pathway 2: Persistent and disorderly

Pathway 3: Deterioration/acute stressor

Crossover cases

Themes associated with intimate partner homicide

Key theme 1: The emotional, mental and physical health of offenders

Key theme 2: Offender experiences of trauma

Key theme 3: Separation

Key theme 4: Hegemonic masculinities and traditional gender norms

Key theme 5: Pre- and post-migration experiences

Discussion

Before the relationship started

After the start of the relationship

Changes in patterns of violence and abuse within the relationship

Key events that increased the likelihood of lethal violence occurring

Implications for policy and practice

Conclusion

References

Appendix A: Literature Review

Trends in intimate partner homicide

Theoretical frameworks for understanding intimate partner homicide

Risk factors for intimate partner homicide

Typologies of intimate partner homicide

The pathway to intimate partner violence

Research gap

Acronyms

Acronym	Definition
ABI	Acquired brain injury
AIC	Australian Institute of Criminology
ANROWS	Australia’s National Research Organisation for Women’s Safety
AOD	Alcohol and other drug-related disorders
BSA	Behaviour sequence analysis
CALD	Culturally and linguistically diverse
CSA	Crime script analysis
DAS	Deterioration/acute stressors
FT	Fixated threat
FTAC	Fixated threat assessment centre
IPH	Intimate partner homicide
IPV	Intimate partner violence
MI	Mental illness
NCIS	National Coronial Information System
NHMP	National Homicide Monitoring Program
PD	Persistent and disorderly
PTSD	Post-traumatic stress disorder

Definitions and concepts

Concept	Definition
Coercive control	For the purpose of this research, “coercive control” refers to the micro-regulation of women’s lives by an intimate partner (usually) in order to maintain dominance or control (Stark, 2007). This can involve a range of behaviours, including frequent belittling and derogatory comments, monitoring of their whereabouts, interfering with their relationships and financial abuse.
Lethal incident/violence	Throughout this report, the term “lethal incident” or “lethal violence” is used to refer to the incident in which the victim was killed by the offender and could not be saved through medical intervention.
Man/men	The term “man” or “men” is used throughout this report to refer to the reported biological sex of the offender. This is primarily attributable to the absence of gender information in the dataset.
Protection order	For the purpose of this study, “protection order” refers to the various legal orders that placed restrictions on the nature and frequency of the respondent’s (offender) contact with the applicant. In the different Australian jurisdictions, these orders are also known as apprehended violence orders, domestic violence orders and family violence intervention orders. This definition also includes interim orders that may be made by the police on behalf of the applicant (victim).
Intimate partner homicide	Intimate partner homicide is defined as an incident where a male offender was charged with killing their female current or former intimate partner by state and territory police agencies.
Intimate partner violence	For the purpose of this research, intimate partner violence is defined as physical violence, sexual violence or emotionally abusive, harassing and controlling behaviours that occur between current or former intimate partners.
Relationship	Relationship is, for the purpose of this research, broadly defined. It includes going on a date, regular dating partners, serious or casual sexual relationships, and emotionally committed relationships, such as long-term, cohabiting, engaged or married partners.
Woman/women	The term “woman” or “women” is used throughout this report to refer to the reported biological sex of the victim. This is primarily attributable to the absence of gender information in the dataset.

Executive summary

Since 1989–90 there have been an average 68 intimate partner homicides (IPH) per year in Australia, the majority of which were perpetrated by a male offender against a female intimate partner (Bricknell & Doherty, 2021).

Intimate partner homicide is the most common form of homicide in Australia and most victims are women. In 2018–19, 15 per cent of all homicide incidents involved the murder (or manslaughter) of a female by a male intimate partner (Bricknell & Doherty, 2021).

Understanding of IPH in Australia – particularly the nature and course of relationships between victims and offenders – is limited. In particular, there is a critical need to examine the sequence of events, interactions and relationship dynamics preceding and coinciding with the male-perpetrated homicide of a female intimate partner, in order to inform prevention and intervention initiatives.

To address this knowledge gap, the Australian Institute of Criminology (AIC) developed the “Pathways to intimate partner homicide” project (PIPH). The study aimed to answer the following research questions:

• Is there a distinct progression of events/phases that lead up to IPH?
• Do these phases follow a similar sequential pathway? Are antecedent sequences identifiable and what do they look like?
• What proportion of incidents are identified as outliers? How do outliers differ in form and context?
• If a common IPH sequence(s) is identifiable, where do recognisable intervention points exist?

Methods

The PIPH sample comprised 199 incidents of male-perpetrated IPH of a female partner that occurred in Australia between 1 July 2007 and 30 June 2018. For the purpose of the study, IPH was defined as an incident where a male offender was charged with killing their female current or former intimate partner by state and territory police agencies.

The three main data sources analysed for the PIPH project were:

• judges’ sentencing remarks for IPH matters considered by the Supreme Court
• coronial findings from the National Coronial Information System
• information sourced from the AIC’s National Homicide Monitoring Program.

Although this project involved the collection, analysis and interpretation of significant amounts of information, and identified patterns to map diverse pathways to IPH, some limitations must be acknowledged. These include the absence of information about victims and the lack of consistency between magistrates and coroners regarding the level of detail provided about cases.

Key characteristics of the PIPH sample (n)

Perpetrators by State

State	Number of perpetrators
Western Australia	30
Northern Territory	15
South Australia	13
Queensland	42
New South Wales	56
Australian Capital Territory	1
Victoria	39
Tasmania	3

Offenders and Victims

	Offenders	Victims
Mean age	41 years (SD=11.8)	39 years (SD=11.8)
Born in Australia	55%	68%
Born Aboriginal and/or Torres Strait Islander	26%	24%

Results

Analysis of the cases included in the PIPH dataset identified three primary offender cohorts and pathways: fixated threat (FT), persistent and disorderly (PD) and deterioration/acute stressor (DAS). There was sufficient information to classify 181 of the 199 cases into one or more of the pathways. There were also crossover cases that involved elements of two or all three of the pathways (n=27, 15%), and three cases that could not be categorised (outliers; not removed from the dataset).

Fixated threat

FT offenders constituted one third of all IPH offenders in the study (33%, n=59). Despite being jealous, controlling and abusive in their relationships, FT offenders were relatively functional in other domains of their life. In many cases they were typically middle-class men who were well respected in their communities and had low levels of contact with the criminal justice system. Their abusive behaviour often took the form of controlling, stalking and monitoring behaviours which escalated in the context of the victim’s perceived withdrawal from the relationship (e.g. separation). Among FT offenders, IPH was used as a means to re-establish control over the victim or in other domains of his life that he blamed her for his loss of control over (e.g. his access to their children). FT offenders were likely to try and conceal what they had done and plead not guilty.

Persistent and disorderly

PD was the most common pathway identified in the analysis (40%, n=73). PD offenders were often Aboriginal and/or Torres Strait Islander peoples; had complex histories of trauma and abuse; had co-occurring mental, emotional and physical health problems; and had significant histories of violence towards intimate partners and others. The relationships that the PD offenders were involved in were characterised by persistent IPV and frequent criminal justice system contact (including protection orders). Despite this, separation was relatively rare for these relationships. In many cases the lethal violence incident was similar in nature to previously reported instances of abuse he had perpetrated on the victim. However, the risk of lethal violence was significantly heightened by co-occurring individual-level and situational-specific vulnerabilities, such as heavy alcohol use or the absence of capable guardians to stop the offender.

Deterioration/acute stressor

One in 10 offenders (11%, n=19) were classified as DAS. These men tended to be non-Indigenous, older, and to have significant emotional, mental and physical health problems. They also demonstrated low levels or an absence of aggression and violent behaviours or tendencies. As such, criminal justice system contact was rare. DAS offenders were in long-term, “happy” and non-abusive relationships with the victim until the onset or exacerbation of a significant life stressor (or stressors) triggered a deterioration in their health and wellbeing. This shift in their trajectory had a negative impact on the offender’s attitudes towards the victim (for example, he may have begun to perpetrate IPV). At the time of the lethal incident, there was no obvious intent to harm the victim. Instead, an argument would occur which, coupled with the offender’s impaired emotional regulation skills, resulted in a nearly instantaneous decision to harm the victim. The DAS offenders were likely to seek help for the victim, demonstrate remorse and plead guilty.

Themes associated with IPH

The pathways identified highlighted key differences in trajectories to IPH. However, there were several themes that appeared to be important in explaining the occurrence of IPH, regardless of the offender classification. These included the emotional, mental and physical health (including alcohol and drug-related disorders; AOD) of offenders; experiences of trauma; pre- and post-migration experiences; separation; and hegemonic masculinities and traditional gender norms.

Theme 1: Offender emotional, mental and physical health problems

The majority of IPH offenders in the sample (73%, n=145) were described as having experienced at least one emotional, mental or physical health condition during their lifetime. These included:

• long-term health conditions: 19 per cent of offenders (n=38)
• cognitive impairment (which limited offender’s agency and management of day-to-day life): 17 per cent of offenders (n=9)
• mental health problems: 43 per cent of offenders (n=86)
• alcohol and other drug problems: 54 per cent of offenders (n=107).

Nearly half (42%, n=84) of IPH offenders had a history of two or more co-occurring physical, intellectual or mental health or AOD problems.

Mental and physical health conditions, including AOD, may have contributed to offender pathways to IPH in a small number of ways. In particular, chronic and persistent consumption of alcohol and illicit substances from early adolescence has been shown to have a negative impact on the neurological development of individuals, which may in turn increase the likelihood of violent offending. Finally, AOD and mental health issues and IPV and IPH may be underpinned by common risk factors, such as experiences of childhood trauma.

Theme 2: Experiences of trauma

More than half (55%, n=109) of offenders had experienced traumatic life events, including war and conflict, homelessness, incarceration, abuse and neglect, and the death of significant family members (including carers). For example, 32 per cent (n=63) of offenders had experienced abuse and neglect during their childhood and adolescence. Experiences of childhood abuse and neglect have been shown to increase risk of IPV and IPH perpetration later in life. These relationships have been explained by examining the links between childhood abuse and the development of dysfunctional attachment among individuals, who then use violence as a means of mitigating the stress they experience within intimate relationships. Alternatively, social learning theory suggests that children who witness IPV between carers and are targets of abuse may believe that violence is a legitimate method of resolving conflict within intimate relationships.

Among Aboriginal and/or Torres Strait Islander offenders, experiences of trauma were framed by judges and coroners through an intergenerational and community disadvantage lens. Some experiences of trauma were more common among Aboriginal and/or Torres Strait Islander offenders than non-Indigenous offenders, such as the death of family members (28%, n=14 vs. 7%, n=10).

Theme 3: Separation

At the time of the lethal violence, the victim and offender were separated in a third (32%, n=64) of cases. Of incidents in which the length of time between separation and the lethal violence was known, one in two occurred within three months (56%, n=23) and 80 per cent (n=33) occurred within a year. During the post-separation period, offenders often began or increased their perpetration of IPV (68%, n=27), often resulting in victims obtaining a protection order (25%, n=16). Conflict during separation centred around care arrangements for shared children, re-partnering and concerns about “love rivals”, and financial disputes.

Theme 4: Hegemonic masculinity and traditional gender norms

There was consistent evidence across the pathways that the offender’s motivation to kill was associated with a perceived violation of gendered norms associated with femininity, which challenged the offender’s masculinity. Such challenges included:

• the victim returning to work or dedicating herself to a career
• the victim refusing to submit to the offender’s demands or expectations
• the victim “fighting back” against the offender during an incident of violence
• the victim having an affair
• the victim re-partnering.

Adherence to traditional gender norms among offenders may been associated with exposure to hegemonic masculine communities (as was the case in 24%, n=48 of cases). Such communities included sport teams and groups, male-dominated work industries and military service. Further, research suggests that adherence to traditional gender norms is more likely when the individual has witnessed and been the target of family violence during their childhood.

Theme 5: Pre- and post-migration experiences

Of the 61 offenders born overseas, 47 were born in locations where English is not the primary language. Some of these offenders (13%, n=6) were forcibly displaced from their birth country by war, political conflict, or risk of ethnic, racial or religious persecution. In many cases involving a CALD offender, changes in the power dynamics within the relationship post-migration was a key component of the pathway to IPH. An important factor that influenced a shift in power dynamics was disparity between the victim and offender regarding their acculturation in Australia, and one partner being dependent on the other for their residency in Australia.

Discussion and implications for policy and practice

Disrupting the fixated threat offender trajectory

The challenge with responding to IPV and IPH perpetrated by FT offenders is that these men are often not visible to law enforcement. A major barrier to identifying FT offenders is that their abuse involves primarily non-physical forms of coercive controlling behaviours, including stalking and emotional and verbal abuse. Disrupting FT offending pathways requires investment in continuing to educate and train frontline staff to identify when coercive control is present, and to treat it seriously when it is detected. Developing innovative responses such as intelligence-led policing, delivered in partnership with representatives from the family law, mental health and domestic violence sectors, is a promising avenue for stopping this type of offender and in turn protecting high-risk victims and their families.

Disrupting the persistent and disorderly offender trajectory

The PD offenders were the most “visible” cohort of offenders within the PIPH sample. By visible, we mean that PD offenders often had high levels of contact with statutory services, including the criminal justice system. They were also arguably the most complex, demonstrating the most criminogenic needs of all pathways identified. In other words, there were multiple and interwoven risk factors for criminal offending present in the lives of PD offenders.

Encouragingly, the evidence base regarding “what works” in responding to this cohort of offenders is relatively strong. The most important strategy for preventing IPH among PD offenders is early intervention and the provision of targeted, integrated and timely support. Prevention responses most relevant to PD offenders must be multipronged and tailored across family, community and school settings. While preventing early adversity is key, it is also important to provide support to address ongoing contextual and situational risk factors for violence such as comorbidity and alcohol use. This can be achieved through the provision of perpetrator interventions that are integrated with AOD and mental health services.

Disrupting the deterioration/acute stressor trajectory

In contrast to the other two pathways, many DAS offenders do not demonstrate established risk factors for IPH. As such, obvious early intervention points for preventing IPH among DAS offenders are lacking. A defining characteristic of DAS offenders was their experiences of acute mental, emotional and physical health problems, which often resulted in engagement with health services. This indicates a (missed) opportunity for intervention. Mental health professionals may be well placed to conduct risk assessments and identify those whose risk of IPH is increasing in the context of deteriorating mental health. Bystander intervention programs that target family and friends may also be a relevant prevention avenue.

Conclusion

The findings from this work provide significant insights into the nature and course of male-perpetrated IPH in Australia. This research provides important evidence for policymakers around the diversity across the pathways that can lead to IPH. However, there were a number of key themes across all pathways. These included experiences of emotional, mental and physical health (including AOD); trauma; pre-and post-migration stress; separation; and hegemonic masculinity. Unfortunately, due to a lack of publicly available information about victims and their families, the current analysis could provide little insight into the lives of victims, and the factors or events in their own pathways that may have presented opportunities for disruption and intervention. This is a clear gap in the research that urgently needs addressing.

Overall, the research found there is not one single, universal pathway to IPH. Indeed, the journey to IPH is instead a series of pathways that branch, weave and intertwine, depending on circumstance. Despite the potential complexity and diversity of the pathways to IPH, it is possible to identify intervention points for most cases. This body of work provides evidence and guidance to better respond to prevention of men’s lethal violence against women.

Summary of key characteristics of the three primary pathways

Prior to the relationship between the victim and offender starting

	Fixated threat (n=59)	Persistent and disorderly (n=73)	Deterioration/acute stressors (n=19)
Histories of traumatic experiences	49% of all FT offenders had experienced trauma in their lifetime	70% of all PD offenders had experienced trauma in their lifetime	42% of all DAS offenders had experienced trauma in their lifetime
Histories of violence towards former partners	21% of FT offenders had been abusive towards former partners	40% of PD offenders had been abusive towards former partners	11% of DAS offenders had been abusive towards former partners
Histories of violence towards others, and non-violent offending	17% of FT offenders had been violent towards other people. 37% of FT offenders had been involved in non-violent offending (described as relatively minor)	37% of PD offenders had been violent towards other people. 55% of PD offenders had been involved in non-violent offending	21% of DAS offenders had been violent towards other people. 53% of DAS offenders had been involved in non-violent offending (described as relatively very minor)
Histories of mental illness, physical health conditions and AOD	MI=42% LTHC=12% AOD=29% Cognitive impairment=8% Rates of comorbidity were low – 25% of FT offenders had two or more co-occurring mental, physical and cognitive health issues	MI=36% LTHC=16% AOD=74% Cognitive impairment=30% Rates of comorbidity were high – 49% of PD offenders had two or more co-occurring mental, physical and cognitive health issues	MI=68% LTHC=32% AOD=53% Cognitive impairment=32% Rates of comorbidity were very high – 68% of DAS offenders had two or more co-occurring mental, physical and cognitive health issues
Perceived functionality in other domains of their life (e.g. education)	Offenders were described as typically functional or successful in public-facing domains of their life. Many offenders were employed in well-respected industries, or owned their own businesses. Offenders were also described as being upstanding members of the community. While a significant proportion of FT offenders had been involved in other forms of offending, these offences were characterised as relatively minor	Overall, PD offenders were described as being dysfunctional in various domains of their life. Many offenders had been unemployed for significant periods of time, and had consistent and regular contact with the criminal justice system for different forms of offending. This cohort had low levels of educational achievement, and had started using alcohol and illicit substances from a very young age	Overall, DAS offenders were perceived as relatively functional in some domains of their life, in particular their social relationships. Although a significant proportion had been involved in other forms of offending, this was characterised as generally very minor. The significant histories of physical and mental health issues meant that some offenders had been unemployed for extended periods of time and/or were on disability pensions

Relationship between the victim and offender

	Fixated threat (n=59)	Persistent and disorderly (n=73)	Deterioration/acute stressors (n=19)
Characteristics of the relationship	81% were either married or in a de facto relationship median relationship length=9 years 51% had at least one child together	66% were married or in a de facto relationship median relationship length=2 years 22% had at least one child together	89% were either married or in a de facto relationship median relationship length=23 years 47% had at least one child together
Nature of IPV within the relationship	There was evidence that 85% of FT offenders were abusive towards the victim during the relationship. Primarily, the violence was described as coercive controlling behaviours, and non-physical forms of abuse (e.g. stalking and emotional and verbal abuse)	There was evidence that 79% of PD offenders were abusive towards the victim during the relationship. Primarily, the abuse was described as physical violence, including slapping, hitting, punching, assaults with a weapon and non-fatal strangulation. Non-physical forms of abuse, including emotional and verbal abuse, were also common	Only two DAS offenders were detected for perpetrating abusive behaviours towards the victim during the relationship. These behaviours were described as relatively minor
Manifestations of coercive control	FT offenders were described as being jealous of the victim’s friends and family members and would interfere with her relationships. They were also possessive and controlling, and would stalk the victim	PD offenders were described as being very jealous of their partner’s relationships with other men, and would use physical violence as a means of control	Very little evidence of DAS offenders engaging in controlling behaviours

In the weeks and months leading up to the lethal incident

	Fixated threat (n=59)	Persistent and disorderly (n=73)	Deterioration/acute stressors (n=19)
Separation status at time of lethal incident	In 61% of cases the offender and victim had separated. In all of these cases, the instigator of the separation was the victim	Separation was rare among PD offenders; in 12% of cases the offender and the victim had separated	Separation was rare among DAS offenders; only 10% of cases involved a victim and offender who were separated at time of lethal incident
Offender’s experiences of significant stressors or changes in their physical and mental health	In many cases it was described that the separation from the victim led to the offender’s mental health deteriorating. Some offenders were described as displaying symptoms consistent with mental health conditions, including anxiety and paranoid thinking, reduced attention to personal hygiene and emotional distress	Despite the presence of chronic AOD and other mental health-related issues, there was very little evidence that the offender’s mental and physical health deteriorated in the lead-up to the lethal incident	A crucial stage of the DAS offender trajectory was the offender experiencing escalating symptoms associated with historical health conditions, and the onset of new stressors. For example, 68% of DAS offenders experienced escalating symptoms associated with underlying mental health conditions in the lead-up to the lethal incident
Changes in patterns of violence and abuse perpetrated against the victim	In the context of the separation, the offender’s controlling behaviours increased and changed significantly. In particular, they started stalking the victim and any new partners they had, including “text bombing” them, and repeatedly attempting to reconcile. They may have also started to try and punish victims by spreading rumours about them to their friends and work colleagues	There was little evidence that patterns of violence and abuse changed within relationships involving a PD offender. However, the level of contact that the offender had with the criminal justice system may have increased. At time of the lethal incident, 26% of PD offenders were the subject of court orders that placed restrictions on the nature and frequency of their contact with the victim	In four cases, there was evidence that the offender started to become abusive and violent towards the victim during the later stages of their relationship. This was described as involving low levels of emotional abuse and minor forms of physical violence

Lethal incidents

	Fixated threat (n=59)	Persistent and disorderly (n=73)	Deterioration/acute stressors (n=19)
Intentionality	There were several indicators that FT offenders entered the same space as the victim with an intent to kill her: 34% made threats to kill the victim in the lead-up to the lethal incident 36% engaged in planning activities prior to entering the space with the victim (e.g. obtaining weapons). Also, many of the offenders used subterfuge or forced their access to the victim	There was very little evidence that PD offenders entered the same space as the victim with an intent to kill her. Rather, the decision to seriously harm the victim appeared to be instantaneous. In particular, only two offenders engaged in planning activities. Further, it was rare that offenders used subterfuge or force to gain access to the victim	There was very little evidence that DAS offenders entered the same space as the victim with an intent to kill her. Rather, the decision to seriously harm the victim appeared to be instantaneous. In particular, no offenders engaged in planning activities. Further, it was rare that offenders used subterfuge or force to gain access to the victim
Key characteristics of incidents	Most common weapon: knife (51%) offender intoxication: 31% victim intoxication: 15% presence of bystanders: 31% concealment activities: 49%	Most common weapon: hands and feet (52%) offender intoxication: 84% victim intoxication: 88% presence of bystanders: 34% concealment activities: 45%	Most common weapon: knife (53%) offender intoxication: 58% victim intoxication: 32% presence of bystanders: 32% concealment activities: 58%
Offender willingness to be held accountable for their actions	Most FT offenders appeared to be unwilling to be held accountable for their actions: 49% pled not guilty 63% were viewed as not being remorseful 22% appealed their conviction	PD offenders overall appeared to be willing to be held accountable for their actions: 40% pled not guilty 30% were viewed as not being remorseful 14% appealed their conviction	DAS offenders overall appeared to be willing to be held accountable for their actions: 26% pled not guilty 11% were viewed as not being remorseful 11% appealed their conviction

Introduction

Since 1989–90 there have been an average 68 intimate partner homicides (IPH) per year in Australia, the majority of which were perpetrated by a male offender against a female intimate partner (Bricknell & Doherty, 2021). While the rate of IPH has decreased over the last 30 years (Bricknell & Doherty, 2021), it remains the most common form of homicide in Australia. In 2018–19, 15 per cent of all homicide incidents involved the murder (or manslaughter) of a female by a male intimate partner (Bricknell & Doherty, 2021); over half of female homicide victims in Australia are killed by a male intimate partner (Bricknell & Doherty, 2021; Cussen & Bryant, 2015).

Australian male-perpetrated IPH trends and figures are consistent with international jurisdictions as well. A systematic review of 227 studies of IPH across 66 countries estimated 38.6 per cent of female homicide incidents (i.e. female victimisation from homicide) and 13.5 per cent of all homicide incidents were IPH (Stöckl et al., 2013). Among the 18 high-income country studies included in the review, the female IPH rate was 41.2 per cent of all female homicide incidents and 14.9 per cent of all homicide incidents (Stöckl et al., 2013).

A large body of research has attempted to identify risk factors associated with IPH, with the aim of preventing the occurrence of these devastating and high-impact events within the community. This research has identified a number of individual- and relationship-level risk factors for IPH, including:

• offender experiences of childhood trauma, including witnessing intimate partner violence (IPV) between carers and being the target of family violence (Aldridge & Browne, 2003; Dobash et al., 2007; Kivisto, 2015)
• offender mental health, particularly depression and suicidal ideation (Bridger et al., 2017; Lysell et al., 2016; Matias et al., 2019)
• the presence of non-fatal strangulation events within the relationship between the victim and offender (Campbell et al., 2003; Glass et al., 2008; Spencer & Stith, 2018)
• sexual violence perpetrated against the victim by the offender (Campbell et al., 2007; Harden et al., 2019; Spencer & Stith, 2018)
• separation between the victim and offender (Aldridge & Browne, 2003; Campbell et al., 2007; Dobash et al., 2004; Dobash & Dobash, 2011; Garcia et al., 2007; Kivisto, 2015; Spencer & Stith, 2018; Wilson & Daly, 1993)
• offender jealousy (Aldridge & Browne, 2003; Caman et al., 2016b; Campbell et al., 2003; Kivisto, 2015; Matias et al., 2019).

Although the above cited studies have provided valuable information about the contexts within which IPH may be more likely to occur, they only provide static representations of IPH – they capture the critical elements but not the patterning or convergence of these elements in the lead-up to and commission of IPH. Moving beyond the examination of static risk factors, more recent IPH research has emphasised the fluidity of violence or the “influential and continuous interaction between individuals and the various situations they encounter” (Vatnar et al., 2017, p. 395). It is the specific sequence of encounters and events that can direct the fatal course of an intimate partner relationship (Dixon & Graham-Kevan, 2011; Keatley et al., 2021; Vatnar & Bjørkly, 2008).

Such sequencing of IPH was recently undertaken by Monckton Smith (2019). The study involved two stages of data collection and analysis: first a review of 372 cases of male-perpetrated IPH included in the United Kingdom’s Counting Dead Women database for the period 2012-2015, and then a more detailed analysis of 25 cases IPH from the author’s previous research. Monckton Smith (2019) identified eight stages that led to and culminated in IPH. These eight stages were associated with three relationship phases – pre-relationship, relationship, and post-relationship – and the known risk factors associated with each phase. In summary:

• Stage 1 or pre-relationship assumes the homicide perpetrator exhibited behavioural risk markers in previous relationships, specifically coercive control, IPV and/or stalking.
• Stages 2 and 3 capture the formation of the relationship and the establishment of roles and obligations. Initial, apparently benign gestures from the male partner about commitment are supplanted with obliged fidelity from the female partner and increasing displays of coercive control.
• Stages 4 and 5 describe the post-relationship phase. Stage 4 is identified as the trigger event – the female partner’s emotional or physical severance from the relationship. Stage 5 represents the male partner’s response where he attempts to re-establish control by escalating and/or diversifying the pattern of abuse.
• Stage 6 is the juncture where the male partner interprets his situation as lost and a “considered decision” to murder his partner emerges. Monckton Smith (2019) notes that this decision is not irreversible and social and environmental factors may dissuade this course of action.
• Stage 7 is the preparation and perpetration of the homicide. Preparation comes in the form of “creating opportunities to kill” as well as conventional patterns of planning.

Monckton Smith’s (2019) research provides a blueprint for IPH, but it is preliminary. What is needed is a complementary examination of this blueprint (alongside the IPH literature) to establish if IPH follows a scheduled pathway and how and when the interactions of pathway factors contribute to the lethal outcome. Of equal relevance is the generalisability of the documented research findings to the Australian context, particularly in regard to Aboriginal and/or Torres Strait Islander women who are absent from these analyses.

To fill this gap, the “Pathways to intimate partner homicide” (PIPH) project was established. The PIPH project aimed to build on and complement studies from state and territory domestic and family violence death reviews, the Australian Institute of Criminology’s (AIC) Homicide in Australia series, and the Australian and international academic literature to describe the nature and course of intimate partner relationships that end in the male-perpetrated homicide of a female partner, and the events and behaviours that could foreshadow fatal outcomes within relationships. More specifically, the current study aimed to answer the following research questions:

• Is there a distinct progression of events/phases that lead up to IPH?
• Do these phases follow a similar sequential pathway? Are antecedent sequences identifiable and what do they look like?
• What proportion of incidents are identified as outliers?
• If a common IPH sequence(s) is identifiable, where do recognisable intervention points exist?

Methods

The PIPH project involved the analysis of 199 incidents of male-perpetrated IPH of a female partner that occurred within Australia during the period 1 July 2007 to 30 June 2018. For the purpose of the study, IPH was defined as an incident where a male offender was charged with killing their female current or former intimate partner by state and territory police agencies. Intimate partner was broadly defined to include any form of relationship, including date/casual, boyfriend/girlfriend, de facto and registered marriage. This definition of IPH is consistent with that of the National Homicide Monitoring Program (NHMP) which is maintained by the AIC.

The study used three main data sources:

• sentencing remarks for IPH matters considered by the Supreme Court (that provide the details of each judge’s reasons for the sentence they deliver), drawn from public access sites, free-of-charge online databases or by formal request for identified matters
• coronial findings accessed through the National Coronial Information System (NCIS)
• the AIC’s NHMP (described below).

The NHMP was established by the AIC in 1990 and contains data on all homicide incidents, victims and offenders recorded by Australian state and territory police agencies from 1989–90 onwards. Homicide incidents include all cases resulting in a person or persons being charged with murder, manslaughter or equivalent offences; all murder-suicides classed as murder by police; and all other deaths classed as a homicide by police whether or not an offender was apprehended.

The NHMP draws on two key sources of data: offence records obtained from each Australian state and territory police agency, and coronial records from the NCIS. Offence records are provided on a financial year basis for each homicide that occurred during the reference period. These data are cross-referenced and supplemented with information from the NCIS, and with additional material from court documents and media reports where available.

Incident selection

As shown in Figure 1, the first stage in selecting the PIPH sample involved identifying all cases in the NHMP that met our definition of IPH. During the 2007–08 to 2017–18 period, there were 498 male-perpetrated IPHs of a female partner. Incidents were removed from the sample if the offender had suicided or was killed by legal intervention (i.e. was shot by law enforcement) prior to the incident being reported to the NHMP (n=89). The decision to exclude these matters was for pragmatic reasons; there was less publicly available information about these cases due to the absence of court proceedings.

Of the remaining 409 cases, an oversample of 300 cases was selected. The oversample was stratified by jurisdiction and the Indigenous status of the offender to ensure the final sample reflected the spread of IPH incidents across Australia. An oversample was deemed necessary to ensure that if cases had to be removed due to erroneous coding or missing information, another case could be substituted.

From the oversample of 300 cases, a case was included in the analysis if it had been adjudicated and the offender was found guilty of murder, manslaughter or other offences in relation to the death of the victim. This meant that cases that had not been finalised at time of data extraction were not included in the final sample. Cases were also excluded from the final sample if the offender was acquitted or found not guilty by reason of mental illness, or the offender died prior to finalisation of criminal justice proceedings. The incident selection process is summarised in Figure 1.

Figure 1: Incident selection process

Data coding

Data were extracted from the previously identified data sources in accordance with a coding framework (and associated guidelines) developed by the research team in consultation with the PIPH Advisory Group. The PIPH Advisory Group was comprised of representatives from academia, government, and non-government organisations that provide support services for victims and perpetrators of IPV, as well as representatives from culturally and linguistically diverse and First Nations communities. The coding framework was informed by a literature review of IPH risk factors, offender and victim characteristics (including the eight-stage homicide trajectory developed by Monckton Smith; 2019), and the crime script for IPV developed by Boxall and colleagues (2018; see Appendix A).

Coding of cases was undertaken by a team of five AIC researchers, and all incidents were double coded. The use of multiple coders was necessary not only to protect the safety and wellbeing of researchers who were reading highly sensitive and potentially distressing information, but also to mitigate the impact of researcher bias. In particular, the use of multiple coders facilitated opportunities for discussion between team members about how to deal with information that was not easily classifiable, or that was contradictory. Contradictory information was very rare within the dataset. Where present, we prioritised some sources of data over others, depending on the information that was in dispute. For example, where there was a disagreement about a victim-related variable, the team would typically prioritise information extracted from coronial records. However, where there was a disagreement about offender information, we prioritised sentencing records.

Multiple coders, however, raised potential concerns about coding consistency. This was addressed through coder “checks” which involved the coders discussing cases with the broader research team. Because most of the information extracted from the datasets was qualitative in nature, this limited opportunities for inter-coder reliability testing.

The research team piloted the coding framework with a small number of randomly selected incidents during the early stages of the project period. This allowed for the coding framework to be revised in accordance with the findings from this process. The coding framework was also revised throughout the project period, as new themes and codes were identified by the research team and the PIPH Advisory Group. This reflexive process of updating and revising the coding framework as data extraction and analysis was occurring reflects the largely exploratory nature of the research.

An important note about “positive” coding

While piloting the coding framework, the research team became aware that the data sources typically only provided information about victims and offenders where it was deemed pertinent to the case. Information not deemed as such was rarely discussed. For example, if the offender did not perpetrate coercive controlling behaviours, there would be no discussion of coercive control in the data source. As such, it became difficult to distinguish between cases in which such behaviours were absent, and cases in which there was no information available to the judge or coroner about these behaviours. To overcome this difficulty, we only coded information if the judge or coroner explicitly stated the presence or absence of specific behaviours, events or factors. If this was not stated, the data item was coded as missing. As such, there is a high proportion of cases for which data is missing.

Crucially, the prevalence estimates that are described throughout this report are inclusive of missing data. This is for two reasons. First, it is important to highlight where the major gaps in data are when researching IPH. Transparency in data gaps may lead to further work being done to address them, which may include data linkage, but also the establishment of new data collection mechanisms. Second, we are mindful of the importance of this research in addressing a high-harm form of violence against women, and as such are cautious about overinflating or misrepresenting the importance of specific factors of events in the IPH pathway. As such, our findings focused on data veracity and reliability, with an understanding that any additional work done in this space will likely see some of these figures increase significantly.

Analysis

Extracted data were analysed by the research team to identify patterns of events and behaviours that preceded IPH incidents (i.e. trajectories). Two primary frameworks informed the analysis: crime scripts analysis (CSA) and behavioural sequence analysis (BSA). CSA provides a framework for identifying the sequential stages of the crime commission process, from start to end, and for mapping the interactions between an offender, a victim and their immediate environment (Boxall et al., 2018; Chiu et al., 2011; Cornish, 1994). BSA involves the analysis of data to map and identify transitions between behaviours and events at the micro (e.g. discrete events) and macro levels (e.g. life histories; Keatley, 2018). The two were combined to provide an “aerial” view of the lives of victims and offenders and to examine the homicide event within the pattern of behaviour exhibited during the course of the relationship, and the offender’s prior behaviour.

CSA is applied to understand the procedural aspects of crime events – that is, how particular crimes occur and take place, and in particular the temporal events and behaviours that occur as the crime is enacted. However, because of its focus on the crime event itself, CSA typically does not capture the sequence of events preceding the crime and, for the purposes of this study, the nature of the relationship between the victim and offender. BSA complements this gap by being sufficiently flexible to examine crime events in detail as well as the sequence of events and behaviours involving the victim and offender that precede and follow the crime event. The unit of analyses for BSA are pairs of events/observations, rather than aggregate-level trajectories, which help identify behaviours that frequently occur together and those that are outliers. The combined approach enables the identification and description of common IPH trajectories, the homicide event and the nature of IPH incidents that do not conform to the conventional pathway.

Limitations

This work has collected, analysed and interpreted significant amounts of information and identified patterns to map diverse pathways to IPH. However, some limitations must be acknowledged and are described below.

Factors impacting availability of data

Jurisdictional and time differences

The level of detail available in the different data sources varied between jurisdictions and over time. For example, sentencing remarks from New South Wales were noticeably longer and more detailed than those from other jurisdictions. This may be attributed to the harsher sentences applied in New South Wales (Brignell & Donnelly, 2015), which may require a greater level of detail to justify. As such, there is comparatively more information about offenders from some states and territories than others. Further, coronial findings for IPHs towards the end of the sample extraction period provided substantial information about the victim’s and offender’s relationship, such as the presence (or absence) of IPV. In particular, coercive control and the associated risk of lethal violence was discussed in great detail in coronial findings from later years, in line with increasing community awareness about the nature and prevalence of coercive control in domestically violent relationships over the past decade, and particularly IPHs (Boxall & Lawler, 2021; Boxall & Morgan, 2021b).

Guilty pleas

The information presented in court during a criminal trial for homicide is focused on determining whether or not the offender intended to kill the victim. This means that the nature of the data available was affected by whether or not the defendant pled guilty. In cases where the offender pled not guilty, the defence and prosecution were required to provide more information to determine the facts that occurred, which were summarised in the sentencing remarks. In contrast, if the offender accepted guilt for the crime and cooperated with police in their investigation of the homicide, certain information was presented and relayed by judges in their remarks.

Similarly, more detailed information about the offender, their circumstances and the facts of the incident was available when the offender appealed the decision, compared to cases that were not appealed, as during the appeal judges were required to again summarise the facts of the case and available evidence. As such, our findings about offenders may reflect the characteristics of those men who plead not guilty and are found guilty (of murder or manslaughter) through the court process.

Lack of information about victims

One of the key limitations of this project was that there was very little information about victims of IPH in any of the data sources. This means that the findings lack insight into a victim’s life history and experiences, both within and outside of her relationship with the offender.

In part, the lack of information about victims in sentencing remarks was expected due to the focus being on sentencing and assessing the perpetrator’s risk of reoffending. At times, judges would refer to victims but they would usually provide no more than a brief summary statement. Consistent information that was provided included the victim’s age, whether they were Aboriginal and/or Torres Strait Islander and if they had children. Victim impact statements were sometimes mentioned but rarely discussed, with judges noting that they were unable to consider these statements when determining an appropriate sentence. Occasionally, information about the victim’s drug and alcohol use was discussed when perceived by the judge to be relevant to the circumstances of the homicide incident. Overall, sentencing remarks were not a suitable source of data for understanding victims’ experiences.

However, the dearth of information about victims in the coronial data was unexpected. The role of the coroner is to determine the identity of the deceased, and the circumstances of their death (including date, place and cause). As the identity of the deceased was never disputed, the coronial findings were typically focused on the lethal incident itself (in order to make the required findings, as referenced above). In some cases, coroners discussed the victim’s background, including where they were born, family and support networks, work history, previous relationships, mental and physical health, criminal justice involvement and service engagement. However, the consistency of this information was very low, meaning it could not be included in the analysis.

Taken together, this means that the findings from the PIPH project, despite being about pathways to IPH, can say very little about victims of IPH. Unfortunately, this is a limitation broader than our study, with little known about victims of IPH more broadly (see Appendix A). Further work is needed to understand IPH victims’ life histories to inform prevention approaches.

Purpose of sentencing remarks

At a jurisdictional level, there was some consistency in structure and style across sentencing remarks, as judges’ opinions were tailored to be consistent with the relevant states’ and territories’ sentencing guidelines. These guidelines exist to reduce unnecessary inconsistency in sentencing and provide judges with a benchmark for particular offences and offenders. When sentencing an offender for a crime, judges will often refer to the aims or purpose of sentencing which are set out by each state and territory in Australia. These include retribution (or punishment), general and specific deterrence, rehabilitation, incapacitation (or community protection), denunciation and restoration. The information provided by judges can differ in length and content based on the individual characteristics of the judge. Most judges will include an overview of the accepted facts of the case, aggravating and mitigating circumstances, information about the offender, reference to the impact on the victim and reference to the aims of sentencing outlined above. The information contained in these decisions is important as remarks are referred to in future cases and can influence decisions in future criminal trials. They also reflect the wider society over which the court rules and the values of that community.

However, the information contained in sentencing remarks is not ideal for measuring real-world trends, but rather indicates how judges weigh facts and arguments (Hall & Wright, 2008, p. 65). That is, the information and reasons for sentencing provided by a judge do not necessarily reflect the breadth of evidence that was considered or available. Fortunately, the level of consistency between judgements is such to allow deep analysis of certain themes (albeit some more than others) and provide important insights into this significant social challenge. Further, the triangulation of these data with other sources such as coronial findings strengthens the validity and reliability of the methodological design.

Biases in the data

Data interrogation and reality testing

Judges often engaged in data interrogation and reality testing of the information presented to them. In some cases the judge explained their process of determining fact from fiction whereas in others this was either implied through other comments or not discussed. This process of determination is foundational to sentencing and discretionary judicial practice. Judges weigh the facts and the evidence presented by both sides and assess the risk and rehabilitation potential of an offender. Process and style are unique for each judge, and are influenced by precedent, past practice, intuition, experience and expertise (Potas, 1991). In some cases, judges would challenge offenders’ accounts of their conduct and state of mind at the time of the incident. Notably, in a number of cases the offender reported not having any memory of the incident but other evidence (such as police interview) contradicted this. For example, in Case 160 the judge outlined their process for determining the facts based on the information presented:

Some aspects of the offender’s account to others have an air of unreality that must lead to a degree of circumspection about his claims. His purported memory of inflicting no more than one stab wound upon [the victim], for example, seems unlikely given the multiplicity of blows. The difference in the offender’s claims to [the psychiatrist] as to his level of intoxication and those made contemporaneously to the events to police points to unreliability.

In this example, the judge weighed different evidentiary sources to determine that it was unlikely the offender represented themselves truthfully. Evidentiary sources in this example included police evidence about the incident itself and the clinical assessment of the offender in the pre-sentence report.

By the time such evidence is presented and discussed by judges in their remarks it has gone through a sorting process where erroneous information has been identified and therefore may be determined to be false or omitted altogether. While this lends a level of reliability to the evidence analysed, it is still selective and inconsistently reported in sentencing remarks. Further research analysing how judges engage in reality testing and data interrogation for domestic violence and IPH criminal trials is needed.

Aboriginal and/or Torres Strait Islander offenders

The data sources used for this study had a number of limitations specific to Aboriginal and/or Torres Strait Islander offenders:

• There was a pre-existing narrative of disadvantage identified in the sentencing remarks, likely attributable to sentencing guidelines.
• The data sources did not allow for a nuanced discussion of the different ways in which IPH (and IPV more broadly) may manifest among Aboriginal and/or Torres Strait Islander peoples and non-Indigenous victims and offenders.
• The data sources did not provide sufficient information to comment on the impacts of intergenerational trauma or systemic racism.

State and territory legislation recognises that Aboriginal and/or Torres Strait Islander peoples uniquely experience ongoing negative social and health consequences as a result of previous policies and practices implemented by federal and state and justice agencies, as well as broader processes of colonisation which may have informed the development of these policies and practices. Aboriginal and/or Torres Strait Islander offenders may therefore be sentenced through such a lens, consistent with the Fernando principles. The Fernando principles are a set of mitigating sentencing considerations that allow the courts to take into account the relationship between social deprivation and crime. This acknowledges that some social groups (particularly Aboriginal and/or Torres Strait Islander peoples) have historically and disproportionately experienced significant systemic disadvantage which impacts on their offending behaviour and overrepresentation in the criminal justice system (ALRC, 2018). Where it is submitted that the Fernando principles apply, it must be established that the offender has suffered significant disadvantage due to their Indigenous status. For example, while being intoxicated is not usually a mitigating factor, it can be considered as such when alcohol use problems have been developed within or as a response to the socio-economic environment in which a person has been raised. Even in cases where the Fernando principles did not apply, pre-existing narratives of disadvantage were observed. That is, more detailed information was provided about the life histories of Aboriginal and/or Torres Strait Islander offenders than of non-Indigenous offenders. While it is important to acknowledge the inequity and historical context that influences criminal justice system engagement among Aboriginal and/or Torres Strait Islander peoples, the Australian Law Reform Commission (ALRC) argues that to broadly accept that Aboriginal and/or Torres Strait Islander offenders are less responsible for their actions is to deny them and their victims their full human dignity (ALRC, 2018). Here we recognise the implications for our findings are that the data are skewed towards understanding IPH offending among Aboriginal and/or Torres Strait Islander peoples through the lens of disadvantage.

Final sample

The largest proportion of cases included in the analysis occurred in New South Wales (28%, n=56), followed by Queensland (21%, n=42) and Victoria (20%, n=39). This is generally reflective of the distribution of Australia’s population (Australian Bureau of Statistics [ABS], 2020) however, as noted earlier, the number of incidents from each jurisdiction included in the sample is proportionate to the number of incidents of male-perpetrated homicide of a female intimate partner occurring in each jurisdiction over the reference period. Aboriginal and/or Torres Strait Islander women were overrepresented as victims of IPH, representing 24 per cent (n=48) of the final sample, but three per cent of the general population of Australia (ABS, 2020). The final sample by jurisdiction and Indigenous status is detailed in Table 1.

Table 1: Final sample by jurisdiction and Indigenous status of victim (n)

	Aboriginal and/or Torres Strait Islander	Non-Indigenous	Not stated/unknown	Total
NSW	6	49	1	56
Vic	1	35	3	39
Qld	11	31	0	42
WA	14	16	0	30
SA	2	11	0	13
Tas	0	2	1	3
ACT	0	1	0	1
NT	14	1	0	15
Total	48	146	5	199

Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

Victims

The mean age of victims at time of the lethal incident was 39 years (range=16–81 years, SD=11.81). Almost one quarter (24%, n=48) of victims were Aboriginal and/or Torres Strait Islander (information missing for five cases). Just over two thirds of victims were born in Australia (68%, n=136), and almost one third were born overseas (31%, n=61; information missing for two cases).

The highest level of education completed by the victim was only known in eight per cent of cases (n=16). Of these victims, most had completed tertiary education at university (n=5) or another tertiary education institution (e.g. technical and further education; n=4), followed by secondary education (Year 10 or above; n=3). Thirty-six per cent (n=72) of victims were employed and 29 per cent (n=57) were unemployed (information missing for 26 cases; see Table 2).

Offenders

The mean age of offenders at time of the lethal incident was 41 years (range=18–74 years, SD=11.77; information missing for seven cases). Just over one quarter of offenders (26%, n=51) were Aboriginal and/or Torres Strait Islander (information missing for 24 cases). Fifty-five per cent of offenders were born in Australia (n=110) and 31 per cent (n=61) were born overseas (information missing for 28 cases).

The highest level of education completed by offenders was available for 52 per cent (n=104) of cases. Of these, the greatest proportion had completed Year 10 and above (52%, n=54), followed by those who had completed Year 9 and below (36%, n=37). Thirteen per cent (n=13) of offenders had no educational attainment. Thirty-seven per cent (n=73) of offenders were employed at the time of the homicide, and 23 per cent (n=45) were unemployed (information missing for 52 cases; see Table 2).

Table 2: Victim and offender characteristics

	Victims		Offenders
	n	%	n	%
Mean age in years (median; range)a	39 (38; 16–81)		41 (40; 18–74)a

Indigenous status

	Victims		Offenders
	n	%	n	%
Aboriginal and/or Torres Strait Islander	48	24	51	26
Non-Indigenous	146	73	124	62
Not stated/unknown	5	3	24	12

Country of birth

	Victims		Offenders
	n	%	n	%
Australia	135	68	110	55
Overseas	61	31	61	31
Not stated/unknown	2	1	28	14

Educational attainment

	Victims		Offenders
	n	%	n	%
Tertiary (university)	5	3	9	5
Tertiary (other institution)	4	2	7	4
Secondary (Years 10–12)	3	2	38	19
Secondary (Year 9 and below)	2	1	35	18
Primary school	1	1	2	1
No educational attainment	1	1	13	7
Not stated/unknown	183	92	95	48

Employment status

	Victims		Offenders
	n	%	n	%
Employed	72	36	73	37
Unemployed	57	29	45	23
Not in the work force	39	20	27	14
Student	5	3	2	1
Not stated/unknown	26	13	52	26

Note: Percentage totals may not equal 100 due to rounding.
^a Excludes seven cases where this information was missing.
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

Relationships

Victims and offenders were most often in de facto relationships (38%, n=76) or married (36%, n=72). In 20 per cent of cases (n=40) the victim and offender were boyfriend/girlfriend, and in four per cent (n=8) they were in a dating/casual relationship. The average length of relationships was 10 years (median=5 years; range=2 days to 52 years).

Most victims and offenders lived together at some point during their relationship, either on a full- or part-time basis (70%, n=139; information missing for 25 cases). Eighteen per cent (n=35) of victims and offenders never lived together. In 39 per cent of cases (n=77), victims and offenders had at least one child together (information missing for 39 cases). The mean number of children was two (median=2; range=1–13). The victim’s and/or offender’s children from previous relationships lived with them in 24 cases (14%; information missing in 107 cases).

The offender was older than the victim in 64 per cent of cases (n=128; information missing for seven cases). Of these, the mean age difference was seven years (range=1–42 years). The victim was older than the offender in one quarter (25%, n=49) of cases, and was, on average, older by six years (range=1–19 years). There was no age difference in 15 cases (see Table 3).

Table 3: Relationship characteristics

Relationship type

	n	%
De facto	76	38
Married	72	36
Boyfriend/girlfriend	40	20
Date/casual	8	4
Not stated/unknown	3	2
Average relationship length in years (median)a	10 (5)
At least one childb	77	39
Median number of children (range)c	2 (1–13)
Cohabitation on a full- or part-time basisd	139	70

Age difference between victim and offender

	n	%
Victim older	49	25
Offender older	128	64
Victim and offender the same age	15	8
Not stated/unknown	7	4

^a Excludes 46 cases where relationship length was not stated or unknown.
^b Denominator includes 39 cases where this information was not available.
^c Limited to cases where the victim and offender had at least one child together.
^d Denominator includes 25 cases where this information was not available.
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

Results

There was enough information about 181 of the 199 cases included in the study to facilitate classification into an offender cohort and pathway. Although the other 18 cases are still included in the higher level analysis (described in later sections of this report), they are not included in the description of the primary pathways. The main reason for excluding these cases was the lack of information about the relationship between the victim and offender prior to the lethal incident taking place.

Overall, three key trajectories were identified within the PIPH sample: fixated threat (FT), persistent and disorderly (PD), and deterioration/acute stressor (DAS). However, as demonstrated in Figure 2, there were also cases that involved elements of two or three of these pathways, and these were classified as “crossover” cases. Finally, three cases were determined to be outliers because they could not be classified into any of the three pathways, or even as crossover cases. This is primarily due to the presence of unique factors that were not consistent with any of the pathways. Specifically, in Case 195 the offender appeared to be acting in self-defence; the victim had been abusive towards the offender throughout their relationship and he lived with disability which impacted the perceived threat associated with the victim. Meanwhile, Cases 186 and 150 appeared to be and were described by various individuals within the dataset as involving euthanasia, whereby the offender had killed their partner at their request, in the context of the victim’s deteriorating mental and physical health.

The following sections of this report describe the three primary pathways in detail and then the crossover cases.

Figure 2: Visual representation of the spread of the final sample across the three primary pathways (n)

	n
Fixated threat (FT)	59
Persistent and Disorderly (PD)	73
Deterioration/Acute Stressors (DAS)	19
FT/PD	13
FT/DAS	8
PD/DAS	4
PD/DAS/FT	2

Pathway 1:
Fixated threat

One in three cases (33%, n=59) included in the sample were classified as FT. The trajectory underpinning this category of offender is described in Figure 3 and in detail below.

Briefly, FT cases involved a male abuser who was described as being highly functional in many public-facing domains in their life. In particular, they were likely to be employed in well-respected industries or to have their own businesses, have low levels of AOD and be described as well-respected members of their communities. However, within their relationship with the victim they were characterised as highly controlling and jealous. FT offenders’ controlling behaviours appeared to escalate in the context of their perceived loss of control over the victim, typically because the victim decided to end the relationship.

In situations where the relationship ended, FT offenders became increasingly focused and determined to reconcile, and also to harm the victim’s reputation and social relationships. This period often co-occurred with the onset or increase in monitoring and stalking behaviours, including online.

After repeated frustrated attempts to reconcile with the victim, FT offenders typically experienced a perceived shift in thinking, where they became more motivated to use extreme forms of violence as a means of re-establishing control over the victim. This may have involved the offender planning to use lethal violence, including purchasing or obtaining the necessary weapons and other materials to kill the victim, and also to obscure their involvement. It was at this point that offenders entered the same space as the victim with an intent to use physical violence as a means of control. They typically gained access to the victim through the use of subterfuge or force. After the incident, most FT offenders attempted to conceal what they had done, and to plead not guilty when charged with the murder.

Figure 3: Fixated threat pathway

Offender characteristics (before relationship start): Offenders were typically middle class and had low levels of contact with the criminal justice system

Overall, the majority of FT offenders were non-Indigenous (81%, n=48; information missing for nine cases), and 36 per cent were born in Australia (n=21; information missing for 12 cases). Almost one in two FT offenders were born overseas (n=26, 44%), of which 20 offenders (77%) were from CALD backgrounds (i.e. English was not the primary language in their country of origin). At time of the lethal incident, the median age of FT offenders was 43 years old (information missing for four cases).

A key characteristic of many FT offenders is that they were perceived as functional and successful in public-facing domains of their life, both prior to and after starting their relationship with the victim. In particular, many FT offenders were consistently employed in well-regarded industries. For example, the offender in Case 187 was employed as a real estate agent, and the offender in Case 129 was a geologist working in the mining industry. Also, a number of FT offenders, including those in Cases 190, 90, 162, 200 and 47, were described as self-employed and/or running their own businesses. Further, some FT offenders were described as “upstanding” members of their local community. For example, the offender in Case 30 volunteered with local agencies to support migrants and refugees to settle within the community. Crucially, this finding is inconsistent with other homicide research which has argued that most IPH offenders are socially and economically marginalised (Polk, 1994).

The overall “veneer” of respectability associated with FT offenders was also influenced by the low levels of AOD (29%, n=17; information missing for 27 cases), cognitive impairment (8%, n=5; information missing for 41 cases), and long-term health conditions (LTHC; 12%, n=7; information missing for 37 cases), particularly relative to DAS and PD offenders. This said, approximately two in five FT offenders experienced symptoms associated with mental illness prior to starting their relationship with the victim (42%, n=25; information missing for 26 cases). The most common diagnoses were anxiety and depression. The co-occurrence of two or more mental and physical health conditions was also relatively rare among FT offenders: 25 per cent of FT offenders had experienced symptoms associated with multiple health conditions prior to starting their relationship with the victim (n=15; information missing for 35 cases).

One in three FT offenders had had contact with the criminal justice system prior to the start of their relationship with the victim for their involvement in criminal behaviours (n=20; information missing for 15 cases). Further, one in five FT offenders (21%, n=12; information missing for 41 cases) had contact with the criminal justice system for abuse and violence towards prior partners. For example, the offender in Case 31 had been convicted of assaulting their former partner and in Case 62 the offender had been charged with breaching a protection order applied for by their former partner.

Although this information was missing for many cases, that sentencing judges did not refer to the prior criminal histories of FT offenders may be because the IPH had been their first reported offence, or because their prior offending was perceived to be “minor” and was therefore irrelevant for informing the sentencing process. Taken together, what this suggests is that FT offenders overall, and the risk they posed to their partners in particular, was not visible to criminal justice and other statutory agencies. This is likely to have been influenced by their perceived middle-class status, their consistent employment, and their limited levels of contact with the criminal justice system (Voce & Boxall, 2018).

The offender and victim start their relationship

Relationship characteristics

As shown in Table 4, the majority of FT offenders were married to (53%, n=31) or in a de facto relationship (29%, n=17) their victims. The average length of relationships was approximately 11.4 years, with a median length of nine years. Half of FT offenders had at least one child with the victim (51%, n=30; median=2.8, range=1–13; information missing for seven cases), and were living with the victim on a full- or part-time basis (68%, n=40; information missing for four cases). Taken together, this indicates that FT offenders’ relationships with victims were long-term and committed.

Table 4: Characteristics of offender relationships with victims (cases involving an FT offender only; n=59)

Relationship type

	n	%
Married	31	53
De facto	17	29
Boyfriend/girlfriend	11	19
Date/casual	0	0
Average relationship length (years; median)a	11.4 (9)
At least one childb	30	51
Median number of children (range)c	2.7 (1–13)
Cohabitation on a full- or part-time basisd	40	68

^a Excludes 10 cases where this information was missing.
^b Denominator includes seven cases where this information was missing.
^c Limited to cases where the victim and offender had at least one child together.
^d Denominator includes four cases where this information was missing.
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

Power imbalances within the relationship

In many cases involving an FT offender, the early stages of the relationship with the victim were characterised by significant power imbalances; specifically, the offender had more power than the victim. These power imbalances were in many cases at least partially attributable to the offender being significantly older than the victim at time of starting their relationship. For example, in Case 175 the victim was only 16 years old when she married the offender who was then 45. Further, in Case 31 the victim was only 17 years old and still in high school when she met the offender who was 23 at the time. In 75 per cent of cases involving an FT offender, the offender was older than the victim (n=44; information missing for four cases). In these relationships, the average age difference was 8.4 years (median=7 years, range=1–30 years).

Other factors that may have contributed to power imbalances within relationships between FT offenders and victims included the offender being employed while the victim was not, and social isolation experienced by victims. Social isolation experienced by women was often caused in part by the violence and abuse perpetrated by the offender (see below), as well as language and cultural barriers experienced by some victims who had migrated to Australia during the early stages of their relationship with the offender (see section on the role of pre- and post-migration experiences for more on this). For example, in Case 178 the victim and offender migrated to Australia after marrying in southeast Europe when the victim was 19 and the offender was 30. This was the victim’s first time in Australia whereas the offender had family members living in Australia and had travelled there extensively. As such, when they immigrated to Australia, the victim was completely dependent on the offender due to language and employment barriers.

Some of the victims were also financially dependent on their partners, primarily because of their role as primary caregivers for their children. This is consistent with the broader IPV literature which has found that having small children is a risk factor for experiencing abuse within relationships (Bowen et al., 2005; Boxall & Morgan, 2021a).

History of controlling behaviours and non-physical abuse within the relationship

Overall, there was consistent evidence that FT offenders were abusive towards their partners throughout the relationship (85%, n=50). In particular, FT offenders were described as jealous and controlling (68%, n=40; information missing for 12 cases) and emotionally abusive (63%, n=37; information missing for 15 cases). Specific behaviours that were detected through the analysis included:

• the offender making demeaning and belittling comments to the victim and calling them names
• the offender attempting to restrict the victim’s contact with other people
• the offender interfering with the victim’s relationships with other family members and friends
• the offender being suspicious of their partner’s friends and family members, and accusing them of having affairs.

In other cases, the offender was described more generically as being “controlling”. For example, it was noted in Case 80 that the offender always “had to have his own way”.

Crucially, it was much less common that FT offenders were described as being physically abusive towards victims (36%, n=21; information missing for 15 cases; see Figure 4). Because the abuse was primarily non-physical, and perhaps in combination with the offender’s perceived functionality or success in other domains of their life (see above), victims often had very little contact with or support from domestic violence services and the police during the life of their relationships. Relatedly, many of these men may not have been physically violent towards their partners as this may have increased their risk of detection by statutory agencies. It has been noted in the IPV literature more broadly that offenders who have “more to lose” from their contact with criminal justice agencies because of their standing within the community and employment are more likely to engage in non-physical abusive behaviours (see for example Voce & Boxall, 2018).

Figure 4: Characteristics of violence and abuse within the relationship (cases involving a FT offender; %)

Characteristics	Cases involving a FT offender (%)
Coercive controlling behavioursa	68
Emotional abuseb	63
Threats to killc	37
Physical violenceb	36
Stalkingd	29

^a Denominator includes 12 cases where this information was missing.
^b Denominator includes 15 cases where this information was missing.
^c Denominator includes 21 cases where this information was missing.
^d Denominator includes 25 cases where this information was missing.
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

Overall, only 12 FT offenders had any contact with the police prior to the lethal incident taking place (20%; information missing for 27 cases), and only 12 per cent had been arrested for IPV (n=7; information missing for 22 cases). As described below, when FT offenders did come to the attention of police, this was likely to have occurred in the context of escalating threatening and controlling behaviours post-separation. As such, it appeared that these victims and offenders were largely unknown or invisible to the police prior to this point, as too was the risk posed by these men.

Offender feels his control over the victim start to slip

Consistent with the findings from other IPH research (Monckton Smith, 2019; Sheehan et al., 2015), a key stage in the FT offender pathway was the offender beginning to believe that they were losing control over the victim. This was primarily attributed to the victim “withdrawing” from the relationship, which may have involved:

• the victim establishing boundaries within their relationship that may not have been there previously (e.g. in Case 171 the victim reportedly attempted to limit how much contact they were having with the offender)
• the victim refusing to make a desired commitment to the relationship (e.g. in Case 164 the victim refused to marry the offender despite his repeated proposals and claims that she had promised to do so)
• the victim refusing to acquiesce to the offender’s demands and attempts to control her behaviours (e.g. in Case 22 the offender did not like the victim’s family and so would try to limit her contact with them, however she continued to maintain her relationship with them).

A significant factor that often contributed to the offender’s perceived loss of control within the relationship was the victim seeking to separate from the offender. Critically, as shown in Figure 5, in 61 per cent of FT cases (n=36; information missing for seven cases) the offender was separated from the victim at time of the lethal incident. This includes one case where the victim and offender were negotiating reconciliation. In the majority of FT cases where the victim and offender were separated at time of the lethal violence, the victim was identified as the instigator of the end of the relationship (78%, n=28; information missing for seven cases).

Figure 5: Relationship status at time of lethal incident (cases involving an FT offender only; n=59; %)

Relationship status	Cases involving a FT offender (n)	Cases involving a FT offender (%)
Separated	35	59
Never separated	22	22
Reconciled	3	5
Negotiating reconciliation	1	2

Note: Denominator includes seven cases where this information was missing.
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

In FT cases where the victim and offender had never been separated (22%, n=13), the loss of control experienced by the offender may have been attributable to more gradual changes in the power dynamics within the relationship. For example, there were a small number of cases involving an FT offender where the victim became financially independent during the later stages of the relationship because of their re-entry into the workforce after having children, or because of their flourishing careers. The increased financial independence and autonomy of these victims coincided with their withdrawing from the relationship, which appeared to make the offender very angry and upset (e.g. Cases 31 and 178). The role of changes in power dynamics in IPH is discussed in more depth in later sections of this report.

The offender blames the victim for his loss of control in other domains of his life

For most FT offenders, the focus of their need for control was their partner. As such, their partner’s perceived withdrawal from the relationship was a significant stage in the pathway that ended with the victim’s murder. However, in other FT cases it appeared that it was the offender’s loss of control in other domains of their life which may have led to the lethal incident taking place. Key domains included their relationship with their children, financial status and employment.

Critically, in all of these cases, the offender blamed the victim for their loss of control in other domains of life, and the hardships they had experienced as a result. For example, as shown in Case study 1, in some cases where the victim and offender had separated and the victim was the primary carer for shared children, the victim’s decision to relocate elsewhere and take the children with them appeared to be a significant factor that contributed to FT offenders’ increasing levels of anger and hostility towards the victim. In these cases, the offender became enraged that the victim had made a decision that they had little control over and/or that they believed was unfair or negatively impacted them.

In other cases, the offender blamed the victim for their loss of control over businesses jointly owned by the victim and offender, or financial stress or distress related to the division of joint assets. For example, in Case 189 the victim was the sole shareholder in a successful business that they jointly ran with the offender. The offender was the director of the business, and he was described as being obsessed with the day-to-day operation of the business. In the context of the relationship breaking down, the victim served the offender with legal papers notifying them that the victim had taken over as director for the business and that he was legally obliged not to sabotage the business. This made the offender incredibly angry as he viewed the business as his own, and resented the victim for her perceived role in removing his control over it. Shortly after being served with these legal papers, the offender killed the victim in her office.

The offender’s controlling and stalking/monitoring behaviours escalate, often in the context of the victim refusing to reconcile or acquiesce to his demands

Regardless of the source the FT offender’s belief that the victim was withdrawing from the relationship, by way of response the offender became more controlling of them as a means of re-establishing their dominance over her and punishing her. The desire of offenders to re-establish control over their partners manifested in a small number of key ways. First, the offender engaged in emotionally abusive and threatening behaviours. For example, in Case 200, after the victim separated from the offender, she began to receive persistent and harassing phone calls and texts from the offender, and he would interrogate her, her family members and their children about her romantic status. Further, in Case 29 the offender’s erroneous belief that the victim was withdrawing from the relationship coincided with the offender becoming verbally abusive towards the victim and their children, particularly while intoxicated.

Stalking and monitoring of victims’ movements was another way in which FT offenders attempted to re-establish control over victims. Stalking in the lead-up to the lethal incident was detected in 25 per cent of cases involving an FT offender (n=15; information missing for 38 cases).

As described in Case study 1, stalking and monitoring behaviours included following victims or frequenting locations where the victim was expected to be. Occasionally the offender interrogated the victim or their friends and family members or even their children about their whereabouts and schedule, and whether they had re-partnered or were having an affair. In a small number of cases, the offender attended the victim’s residence when they were not there and/or even broke into their residences. This was demonstrated in Case 80; in the period following the relationship ending, the offender would call her home phone and peer into her windows to see if she was home.

The escalation of controlling behaviours displayed by FT offenders was often facilitated through the use of technology. For example, in Case 187, in the post-separation period the victim’s online dating profiles were accessed by the offender without her consent, and in Case 175 the offender recorded the victim’s phone calls. Further, there was clear evidence in Case 179 that the offender had been stalking and monitoring the victim’s movements for an extended period of time leading up to the lethal incident, through GPS tracking software as well as by installing CCTV cameras in their shared apartment and outside the front door. Offenders calling and texting the victim repeatedly were also commonly reported behaviours for FT offenders (e.g. Cases 94, 80, 58, 187 and 147).

Critically, in cases where the victim and offender had separated and the victim had re-partnered, the offender often began to abuse and stalk the new partner as well. This was observed in Case 47: after the victim re-partnered the offender was described as becoming very angry and started to abuse the victim and her new partner via text and phone calls.

Another manifestation of FT offenders’ desire to re-establish their control over victims was repeated and persistent attempts to reconcile. This was often despite victims remaining resolute in their decision to separate. For example, in Case 190 the victim separated from the offender after his long-term affair with another woman was discovered. In the following two-month period, he made repeated attempts to reconcile with her which she constantly rebuffed. Further, in Case 171, the victim and offender worked for the same company. Despite the victim seeking support from the company’s human resources department (resulting in actions being taken to protect the victim), the offender continued to make repeated attempts to reconcile with her both within the workplace and outside of it.

The escalation in behaviours and threat posed by offenders during this phase of the pathway appears to have been understood by victims. In some cases, after separating from their partner and in the context of changes in the patterns of abuse, the victim may have for the first time had contact with the police or legal services seeking advice about applying for a protection order. For example, in Case 171, after an incident where the offender had threatened to commit suicide, the victim sought advice from the Family Court about applying for a protection order. However, she chose not to progress with her application for unknown reasons. In Case 30 the victim first applied for a protection order against the offender after they had separated and he began to make explicit threats to harm her. This was the first time the victim had had any contact with the police about his behaviours, despite his persistent physical abuse of her throughout their 25-year marriage. However, again consistent with the findings described above, it was relatively rare for victims to have contact with the police. That these victims started to reach out to support services after often extended periods of service avoidance and non-engagement may provide some insight into the cognitive state of women, and that they recognised the level of threat posed by their partners at this point.

Of critical importance is that in most cases where the victim and offender had separated (or were negotiating reconciliation), the period of escalation which preceded the lethal violence was often very short. The median length of time between the separation and the lethal incident was only 51 days (mean=97 days; information missing for 12 cases). In 53 per cent of cases involving an FT offender, the lethal incident occurred within six months of the separation (n=19; see Figure 6). This indicates that for women in relationships with FT offenders, risk of lethal violence occurring increased significantly in the context of relationship dissolution. This is supported by the broader IPV literature (Aldridge & Browne, 2003; Campbell et al., 2007; Dobash et al., 2004; Dobash & Dobash, 2011; Garcia et al., 2007; Kivisto, 2015; Spencer & Stith, 2018; Wilson & Daly, 1993).

Figure 6: Length of time between separation and lethal incident (months; cases involving FT offenders only; n=36; %)

Length of time	Cases involved FT offenders (%)
Less than a month	8
1-2 months	8
3-6 months	19
7-12 months	8
13 months or longer	6

Note: Limited to cases involving an FT offender where the victim and offender were separated or negotiating reconciliation at time of the lethal incident. Denominator includes 12 cases for which this information was missing.
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

The offender becomes motivated to harm the victim

In the context of escalating controlling behaviours, including stalking and monitoring of the victim online and in the real world, which often co-occurred with repeated failed attempts to reconcile with the victim, the offender appeared to become more motivated to use lethal violence against their partner as a means of controlling her. The offender’s change in thinking was indicated in some cases by the offender verbalising their intent to kill the victim. These threats may have been communicated to the victim herself or other family members or friends. The offender threatening to kill the victim in the lead-up to the lethal incident was detected for one in three cases involving an FT offender (34%, n=20; information missing for 32 cases). For example, in Case 94, during the weeks leading up to the lethal incident the offender left numerous messages on the victim’s phone saying that he was going to cut her throat, and even offered to pay his brother money to “bash” the victim. Further, in Case 90 the offender told his friends that if the victim tried to leave him again he would bash her head in with a golf club. In this case, the offender subsequently murdered the victim using a golf club.

Further, one in three FT offenders (36%, n=21; information missing for 15 cases) began to plan how they would murder the victim and conceal their involvement. This commonly involved the offender:

• purchasing or obtaining accoutrement to undertake the act (e.g. gaining access to the murder weapon; Cases 20 and 140)
• learning how to murder their partner (e.g. Cases 127 and 47)
• laying the groundwork for establishing an alibi or to conceal their involvement in the murder (e.g. Cases 173 and 121).

The offender enters the same space as the victim, often through subterfuge, with an intent to control her using lethal violence

The change in thinking demonstrated by many FT offenders meant that at time of the lethal incident occurring, they entered the space with the victim with the clear intent of controlling her, primarily through the use of lethal violence. For many offenders, this may have involved entering the same space with the victim under false pretences, as a means of getting the victim to lower their guard and so become more vulnerable to lethal violence. As described in Case study 1, in other cases the offender entered the same space as the victim without their knowledge or consent. For example, in Case 190 the offender broke into the victim’s house through a back door and assaulted her with a weapon while her back was turned. Also, in Case 55 the offender threw a rock through the victim’s window to gain entry and killed her while she slept.

FT offenders often came armed with the weapons they used to murder the victim; the most common weapon used by FT offenders to murder the victim was a knife or other sharp implement (51%, n=30). In one in 10 cases the offender used a firearm (n=7, 12%).

That offenders entered the same space as the victim with an intent to control her with lethal force is also demonstrable by the lack of conflict within the incident itself – in other words, the offender and victim were often not arguing about anything at time of the incident occurring. Further, it was unlikely that FT offenders were intoxicated at time of the incident (30%, n=18; information missing for 15 cases).

Further evidence of FT offenders being highly motivated to kill the victim upon entering the same space as them could be elicited from the finding that approximately half attempted to conceal their role in the homicide (49%, n=29; information missing for five cases). Concealment activities included:

• disposing of the weapon
• fleeing from the scene
• concealing the body
• implicating other people in the homicide
• denying their guilt when questioned by law enforcement and others.

Finally, almost half of FT offenders pled not guilty when charged with murder (49%, n=29; information missing for four cases), and maintained their innocence even when confronted with overwhelming evidence of their role in the crime. Demonstrations of remorse were also rare among FT offenders; only one in four were determined by the court to have been sincerely remorseful for their actions (24%, n=14; information missing for eight cases; see Table 5).

Table 5: Key characteristics of lethal incidents (cases involving an FT offender only; n=59)

Location of incident^a

	n	%
Victim’s/offender’s home	41	69
Public place	9	15
Other person’s home	5	8
Other	3	5
Presence of childrenb	23	39
Presence of adult bystandersb	18	31

Offender intoxicated^c

	n	%
Alcohol	16	27
Drugs	2	3
Both	0	0

Victim intoxicated

	n	%
Alcohol	5	8
Drugs	4	7
Both	0	0

Weapon used

	n	%
Knife or other sharp instrument	30	51
Hands and feet	14	24
Blunt instrument	13	22
Firearm	7	12
Rope	3	5
Fire	2	3
Drugs	0	0
Other	7	12
Presence of overkille	35	59
Offender called for assistance for the victimf	10	17
Offender attempted to self-harmg	12	20
Offender pled guiltyg	26	44

Offender conviction

	n	%
Murder	53	90
Manslaughter	5	8
Malicious infliction of grievous bodily harm	1	2
Offender appealed convictiona	13	22

^a Denominator includes one case where this information was missing.
^b Denominator includes six cases where this information was missing.
^c Denominator includes 15 cases where this information was missing.
^d Denominator includes five cases where this information was missing.
^e Denominator includes 10 cases where this information was missing.
^f Denominator includes two cases where this information was missing.
^g Denominator includes three cases where this information was missing.
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

Pathway 2:
Persistent and disorderly

The most common trajectory identified within the sample was persistent and disorderly (PD). Overall, 40 per cent of offenders (n=73) in the sample were classified as PD. The trajectory underpinning this category of offender is described in Figure 7 and in detail below.

A key characteristic of the PD trajectory is that, unlike the DAS and FT pathways, there was not a clear set of stages that led to the offender killing the victim. Instead, the trajectory involved the presence of persistent and ongoing violence and abuse within the relationship perpetrated by the offender, that started during the early stages of the relationship, and continued until the victim’s death. Although in many cases the violence and abuse may have escalated in severity or frequency at different points in the relationship, this was not characteristic of all matters involving a PD offender. Further, although the relationship may have been described as “volatile”, “chaotic” or “tumultuous”, another defining characteristic of these relationships was their persistence: the victim and offender were rarely separated at time of the lethal incident occurring. In general, these relationships could be characterised as committed and chaotic. This said, the absence of a clear trajectory in PD cases is potentially also attributable to the briefness of the relationships between some offenders and the victims.

For some PD offenders, there was evidence that their abusive behaviours eventually led to punitive criminal justice responses, including court-ordered restrictions on the nature and frequency of their contact with victims (e.g. protection orders and good behaviour bonds that banned the victim and offender from seeing each other). However, even when these orders were in place, the relationship between the victim and offender appeared to continue. To avoid detection by law enforcement, the victim and offender may have started to avoid seeing their friends and family members, or being seen together in public more generally. This meant that the victim and offender were continuing to spend time together and were potentially more socially isolated.

In the context of the ongoing and persistent nature of the violence and abuse within the relationship, the offender would enter the same space as the victim often with the intent to harm her, as he had done so previously on numerous occasions. That the victim and offender were still in a relationship together at time of the lethal incident occurring meant that the offender had easy access to the victim. After coming together in the same space, the offender started to physically abuse the victim. This appeared to coincide with the offender becoming very emotionally heightened as a result of an argument with the victim, or because of their sexual jealousy. However, while in many cases the abuse targeted at the victim may have been consistent with previous incidents of violence within the relationship, this time her risk of being killed may have been heightened due to situation-specific vulnerabilities. These may have included the absence of capable guardians to stop the offender, the intoxication levels of the victim, and pre-existing injuries or health issues of the victim. However, in some cases the violence used by the offender during this incident was much more severe, involving the use of weapons, which resulted in the death of the victim.

Figure 7: Persistent and disorderly offender pathway

Offender characteristics (before relationship start): Offender had complex histories of trauma, mental illness and alcohol and other drug-related disorders, and had significant histories of contact with the criminal justice system

Overall, half of PD offenders were Aboriginal and/or Torres Strait Islander (55%, n=40; information missing for eight cases). Three in four PD offenders were born in Australia (73%, n=53; information missing for 12 cases), and 11 per cent (n=8) were born overseas. At time of the lethal incident, the median age of PD offenders was 39 years old (SD=9.19; information missing for one case).

Experiences of trauma

As noted in the limitations section of this report, information about the lives of offenders prior to starting their relationship with the victim was inconsistently available within the PIPH dataset. However, where this information was available, experiences of significant trauma during childhood and adolescence – prior to starting their relationship with the victim – appeared to be common among PD offenders. For example, one in four PD offenders were reported to have witnessed IPV between carers in their household (23%, n=17; information missing for 54 cases) while 16 per cent were targets of abuse and violence from carers (n=12; information missing for 58 cases). One in 10 PD offenders had both witnessed family violence in their family of origin and been a target of abuse themselves (12%, n=9; information missing for 61 cases).

Further, one in five PD offenders was reported to have experienced the death of a family member during their childhood or adolescence (18%, n=12; information missing for 56 cases). This family member was frequently a primary carer, or a sibling. For example, in Case 40 the offender was very close to his grandparents while he was growing up, and he often stayed with them because his father was an abusive alcoholic. The deaths of his grandparents, first his grandmother when he was an adolescent and then his grandfather a few years prior to starting the relationship with the victim, were described as having a significant and negative impact on his life, including his emotional development and schooling.

Mental illness, alcohol and other drug use disorders and physical health conditions

Prior to starting their relationship with the victim, there was evidence that many PD offenders had significant and co-occurring mental and physical health issues, including alcohol and other drug use disorders. One in three offenders had experienced symptoms associated with mental health conditions (36%, n=26; information missing for 43 cases), and one in six had an LTHC (16%, n=12; information missing for 55 cases). Health conditions included diabetes, hypertension, renal disease, liver disease, tachycardia, prostate problems, blood diseases, paralysis, hemiplegia and leprosy.

Also, one in three PD offenders had a cognitive disability (30%, n=22; information missing for 46 cases), including a suspected or diagnosed acquired brain injury (ABI; 23%, n=17; information missing for 50 cases).

Overall, 74 per cent of PD offenders had a suspected or diagnosed AOD (n=54; information missing for 18 cases). Crucially, in many PD cases the offender started using alcohol and other drugs in their early adolescence (e.g. Cases 182, 118 and 191). This is consistent with the evidence on the aetiology of substance use disorders showing that adolescence is peak time of initiation to use (Tomlinson et al., 2016; White et al., 2019).

The co-occurrence of two or more physical and mental health issues (comorbidity) was very high among the PD cohort. One in two PD offenders (49%, n=36; information missing for 35 cases) had multiple physical and mental health issues at time of starting their relationship with the victim. For example, in Case 59, the offender used methamphetamines three to four times a week in the lead-up to the homicide and they had an AOD that originated in their early teens; they would regularly binge drink to the point of blacking out.

Criminal offending and violence and abuse towards others

There was consistent evidence that many PD offenders had significant histories of violence towards others, including intimate partners. Forty per cent of PD offenders were reported to have been abusive towards former partners (n=29; information missing for 41 cases), and 33 per cent had been charged with an offence against a former partner, including assault, sexual assault, stalking and breach of a protection order (n=24; information missing for 47 cases). For example, in Case 23, the offender had a significant history of violence towards former partners. This included being charged with rape when he was only an adolescent (for which he was sentenced to three years in a juvenile justice facility). He had also been charged on two separate occasions with aggravated assault and rape against a former partner, for which he was convicted and sentenced to a term of imprisonment.

What this means is at time of entering their relationship with the victim, many PD offenders had already developed patterns of abusive behaviours that they had perpetrated against former partners, but were also likely to have been reported to law enforcement for them.

Violence towards non-intimate partners, including family members, friends, acquaintances and police officers, was also common among PD offenders (e.g. Case 35; see Case study 2). Overall, one in three PD offenders had been involved in violent offending towards non-intimate partners (37%, n=27; information missing for 40 cases). Finally, 55 per cent of PD offenders had been involved in non-violent offences, including break and enter and other forms of property crime, as well as drug-related offending and public disorder (n=40; information missing for 25 cases). Taken together, what this suggests is that, where this information was available, many PD offenders were involved in non-IPV-related offending.

The victim and offender start their relationship

Half of PD offenders were in a de facto relationship with the victim (53%, n=39; information missing for three cases), while one in four were in a boyfriend/girlfriend relationship (25%, n=18), and five per cent were in a casual/dating relationship (n=4). The median length of relationships was two years, but one in five PD offenders had been in a relationship with the victim for less than 12 months prior to the lethal incident occurring (22%, n=16; information missing for 19 cases).

As shown in Table 6, 22 per cent of PD offenders (n=16; information missing for 27 cases) had at least one child with the victim (median=2, range=1–8), and 66 per cent were living with the victim on a full- or part-time basis (n=48; information missing for 14 cases).

Taken together this indicates that many PD offenders were in relatively new relationships with their victims, and had not been with them for significant periods of time prior to the lethal violence taking place.

Table 6: Relationship characteristics (cases involving a PD offender only)

Relationship type

	n	%
De facto	39	53
Boyfriend/girlfriend	18	25
Marriage	9	12
Date/casual	4	5
Unknown	3	4
Average relationship length, in years (median)a	5 (2)
At least one childb	16	22
Median number of children (range)c	2 (1–8)
Cohabitation on a full- or part-time basisd	48	66

^a Excludes 19 cases where this information was missing.
^b Denominator includes 27 cases where this information was missing.
^c Limited to cases where the victim and offender had at least one child together.
^d Denominator includes 14 cases where this information was missing.
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

Violence and abuse within the relationship

The relationships between PD offenders and victims was characterised by persistent abusive behaviours perpetrated by the offender against the victim. Overall, 79 per cent of PD offenders (n=58) were described as being violent and abusive towards the victim. As shown in Figure 8, 71 per cent of PD offenders were described as physically violent towards the victim, including pushing, shoving, punching and slapping them (n=52; information missing for 16 cases). A small number of PD offenders also perpetrated more severe forms of abuse like non-fatal strangulation (5%, n=4; information missing for 61 cases). As a result of the physical violence perpetrated against them, one in four victims had experienced some form of injury (23%, n=17; information missing for 42 cases). Commonly described injuries included bruises, abrasions, cuts, swelling and broken bones.

Non-physical forms of abuse were also commonly perpetrated by PD offenders. One in three PD offenders were described as emotionally abusive towards the victim, including behaviours such as verbal abuse, name-calling and put-downs (32%, n=23; information missing for 48 cases). Further, 36 per cent of PD offenders had also been controlling towards their partner (n=26; information missing for 45 cases). Interestingly, unlike the FT offenders, PD offenders were unlikely to have been described as stalking the victim (4%, n=3; information missing for 62 cases), or threatening to kill them (12%, n=9; information missing for 58 cases).

Figure 8: Characteristics of violence and abuse within the relationship (cases involving a PD offender; n=73; %)

Characteristics	Cases involving a PD offender (n=73; %)
Physical violencea	71
Coercive controlling behavioursb	36
Emotional abusec	32
Stalkingd	4
Threats to kille	3

^a Denominator includes 16 cases where this information was missing.
^b Denominator includes 45 cases where this information was missing.
^c Denominator includes 48 cases where this information was missing.
^d Denominator includes 62 cases where this information was missing.
^e Denominator includes 65 cases where this information was missing.
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

Thirty-eight per cent of PD offenders had been charged for abusing their partner prior to the lethal incident (n=28; information missing for 30 cases). In some cases, the offender had been charged on multiple occasions (e.g. Cases 50, 48 and 25). Further, 40 per cent of PD offenders had been the subject of a protection order because of their behaviours towards the victim during their relationship (n=29; information missing for 29 cases).

However, in other cases where the offender was described as being violent and abusive towards the victim, the offender had not been charged with assaulting their partner and/or were not the subject of a protection order. This may have been because the violence, which was present, had not been reported to the police (e.g. Cases 23 and 40). For example, in Case 40 it was noted that the victim had been seen by others with black eyes prior to the lethal incident occurring, but no report was made to the police. The barriers to reporting incidents of IPV are well documented in the literature (Felson et al., 2002; Robinson et al., 2020; Voce & Boxall, 2018).

Volatility within the relationship

Relationships between PD offenders and victims were described by various individuals within the dataset as “turbulent” (e.g. Case 43), “volatile” (e.g. Case 125) and “toxic” (e.g. 32). Noting that these terms are highly problematic in that they downplay the presence of potentially abusive behaviours within the relationship, they were used by coroners and judges in particular to highlight the significant levels of conflict within the relationship. For example, in Case 32, the offender’s relationship with the victim was described as being marked by repeated arguments because of his disapproval about her involvement in sex work.

In many cases, the conflict within the relationship appeared to be exacerbated when the offender and/or the victim were intoxicated (e.g. Case 48). Overall, in 45 per cent of PD cases both the offender and the victim had AOD-related issues (n=33; information missing for 40 cases). Alternatively, 61 per cent of PD offenders who had an AOD were in a relationship with a victim who also had an AOD (n=33; information missing for 21 cases).

In addition to high levels of violence and abuse, conflict and mutual AOD issues, relationships involving PD offenders were also characterised by high levels of persistence. While one in four offenders had separated from the victim at some point during the relationship (25%, n=18; information missing for 18 cases), at time of the lethal incident, only 12 per cent (n=9) were separated from the victim, while 60 per cent were not separated (n=44; information missing for 20 cases; see Figure 9).

Figure 9: Relationship status at time of lethal incident (cases involving a PD offender only; n=73; %)

Relationship status	Cases involving a PD offender (n)	Cases involving a PD offender (%)
Never separated	37	51
Separated	9	12
Reconciled	7	10
Negotiating reconciliation	0	0

Note: Denominator includes 20 cases where this information was missing.
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

Lifetime persistence of abusive behaviours

While the majority of PD offenders were reported to have been abusive towards the victim during the relationship, in other cases there was a notable absence of detected abusive behaviours. However, in these cases, many of the offenders had been violent towards a previous partner, and their relationship with the current partner had only been ongoing for a short period of time. For example, in Case 125, the offender had only been with the victim for one month prior to killing her. Although the relationship had been described as “volatile”, there was no evidence of abuse and violence specifically. However, this offender had been convicted on nine previous occasions for assault-related offences, and had been charged with threatening to kill his former partner after commencing his relationship with the victim.

Of cases involving a PD offender who had been violent towards a previous partner (n=29), 21 per cent were reported to not have been violent towards the current partner (n=6). Interestingly, the longest relationship within this sub-group was 14 months, and the shortest was less than a month (information missing for one case). As such, it is possible in these cases that the violence and abuse had not been detected or had not occurred because it had not started yet, and the lethal incident was the both the onset and end of the abuse within the relationship. These cases may not have involved persistent patterns of abusive behaviours within the relationship, but the offender’s prior history of violence towards other partners demonstrated that they were high risk for perpetrating abusive behaviours against the victim as well. This is supported by a volume of research which has demonstrated that the strongest predictor of future IPV reoffending is prior reoffending (Hulme et al., 2019), and that persistence of IPV across relationships is common (Aldarondo & Sugarman, 1996; Capaldi et al., 2003; Vickerman & Margolin, 2008).

Overall, 88 per cent of PD offenders had been abusive towards a former partner or the victim (n=64; information missing for nine cases). Further, 32 per cent of PD offenders had been violent towards a former partner and the current victim (n=23; information missing for 41 cases).

The offender’s level of contact with the criminal justice system increases, and they become the subject of court orders limiting the nature and frequency of contact with the victim

In many cases involving a PD offender, the offender’s level of contact with the criminal justice system increased in the months and weeks leading up to the lethal incident. In particular, some offenders became the subject of bail or parole conditions, or protection orders, which placed restrictions on the nature and frequency of their contact with the victim. Conditions varied across cases, including:

• restrictions on alcohol consumption by the offender (e.g. in Case 163 a condition of the protection order against the offender was that he was not supposed to consume alcohol with the victim)
• their living arrangements (e.g. in Case 169 the offender was bailed after assaulting the victim on the condition that he not live with the victim)
• their contact with the victim (i.e. non-contact provisions; e.g. in Case 32 a condition of the offender’s bail for assaulting the victim was that he not see her anymore).

At time of the lethal incident occurring, 26 per cent of PD offenders were the subject of some form of court order which placed restrictions on the nature and frequency of their contact with the victim (n=19; information missing for 39 cases). In a small number of other cases, the offender was the subject of other orders that protected former partners from their violence and abuse. For example, in Case 125, the offender was actually on a good behaviour bond for threatening to kill his former partner, after commencing the relationship with the current victim. Also, in Case 23 the offender had been charged and convicted of rape and aggravated assault against his former partner in absentia, because he had fled the jurisdiction when released on bail. While he was evading a warrant for his arrest, he met and started a relationship with the victim in another jurisdiction where he ultimately killed her.

However, despite these legal restrictions, in most cases the offender continued to live with and see the victim. Of those matters where there were current court orders in place at time of the lethal incident, half of offenders were still in a relationship with the victim (53%, n=10; information missing for four cases). Breaches of protection orders that came about as a result of the persistence of the relationship were referred to in sentencing remarks and coronial records as “breaches by consent” – suggesting that the victim had voluntarily chosen to continue their relationship with the offender despite the presence of ongoing abuse. This is a contentious issue as the ability for women to “consent” to participate in an abusive relationship is questionable due to the ongoing pressure and manipulation that may be exerted upon them by their partners.

Putting aside the question of consent, in the context of the relationship continuing and the presence of legal restrictions, some offenders and victims began to engage in “service avoidance”. Service avoidance broadly describes purposeful actions taken by victims and offenders to hide their relationship from services that may be monitoring their status. This includes formal services like law enforcement, as well as family members and friends. For example, in Case 28, after the victim was assaulted by the offender, a protection order was put in place forbidding the offender from seeing the victim. However, the relationship continued, and the victim and offender went to great efforts to hide their relationship from police as well as the victim’s family who did not approve of the offender. This resulted in the victim and offender going on extended camping trips to remote areas of the bush where they could be together without fear of detection. It was during one of these camping trips that the offender beat the victim to death over a number of days.

The offender enters the same space as the victim

At time of the lethal incident occurring, the offender entered the same space as the victim as part of the victim’s and offender’s usual routines. Because the relationship was ongoing, unlike in FT cases the victim was not required to use subterfuge, or force, to gain access to the victim.

In many cases, the lethal incident was indistinguishable from other forms of violence and abuse that had previously occurred within the relationship. However, in these cases, the risk of lethal violence may have been heightened due to the presence of situation-specific vulnerabilities. For example, as noted above, in cases where the offender became the subject of court orders limiting their contact with the victim, the victim and offender may have attempted to avoid detection by law enforcement and others by spending more time together in isolation. In these situations, the ability of bystanders to intervene and stop the offender from abusing his partner and/or engaging in significant acts of violence was more limited. For example, in Case 28 the victim died of internal bleeding after being beaten by the offender over a period of a few days on a camping trip together.

Alternatively, in some cases the lethal incident occurred when either or both the victim and offender were highly intoxicated. Overall, 84 per cent of offenders were intoxicated at time of the lethal incident (n=61; information missing for nine cases). While this is not an excuse for the violence, as is discussed in later sections of this report, the offender’s intoxication levels may have inhibited their emotional regulation skills and made it less likely that they would stop themselves from inflicting serious harm on their partners. In 88 per cent of cases the victim was intoxicated (n=64; information missing for two cases), which not only may have impaired their ability to defend themselves, but also increased the likelihood of serious harm occurring. For example, in Case 40 the victim’s cause of death was recorded as methamphetamine toxicity and asphyxiation; after the victim collapsed due to the effect of the methamphetamines she ingested, the offender tied a ligature around her neck.

In other cases, the situation-specific risk factor that increased the likelihood of serious violence taking place was the presence of weapons. For example, in Case 32 the offender had brought a gun in from the shed prior to the lethal incident because he was going to sell it to a friend. However, during the evening he and the victim consumed a significant amount of methamphetamines and he had an argument with the victim. When she left the room he grabbed the gun and shot her. It is unclear whether he would have had the means to kill the victim if not for the presence of the gun. Further, it was noted in a small number of cases that the victim was still recovering from prior assaults and resulting injuries at time of the lethal incident occurring, which may have increased her risk of experiencing lethal violence.

Although the lethal incident appeared to be largely consistent with the patterns of violence that had occurred previously within the relationship, it is important to note that the offender appeared to at some point make an instantaneous decision to, if not murder their partner, inflict serious and significant harm on them. This was primarily due to the offender becoming very emotionally heightened for various reasons. In some cases the victim and the offender had an argument, primarily about the offender’s sexual jealousy and suspicions about the victim’s relationships with other men. This was particularly notable among cases involving First Nations offenders; in many of these cases the presence of “jealousing” was noted as a factor that contributed to the offender’s heightened emotional state (e.g. Cases 5, 7, 12). The role of “jealousing” in lethal incidents involving a PD offender is described in Case study 3.

In other cases, the source of the offender’s emotional distress appeared to be unrelated to the victim, but was attributable to external causes. In these situations, the offender appeared to use violence against the victim as a means of expressing this emotional distress and regaining their sense of control lost in other domains. For example, in Case 193, the offender, who was Aboriginal and/or Torres Strait Islander, was visited by the police on the night of the lethal incident because his niece had been picked up for fighting. The offender was warned to keep his niece there for the night or she would be arrested. This led to a noticeable change in the offender’s mood and he became very aggressive towards the victim. Over the course of the next 12 hours, he beat the victim repeatedly with his fists and weapons. She eventually died from her injuries.

Despite the sustained beatings perpetrated by offenders against victims, in many cases it appeared that the offender did not appear to realise the level of harm they had inflicted on their victims, or the victim’s risk of dying. For example, it was common that offenders would beat the victim and then go to sleep (e.g. Cases 29, 34 and 56). In Case 159, after drinking together for a few days the victim and offender had an argument which led to the offender beating the victim. After he had stopped, the victim appeared to recover slightly and even smoked a cigarette. In the morning the offender was awoken by the victim’s brother who had found the victim dead in the hallway.

That offenders may not have realised the level of harm they had perpetrated is perhaps also supported by the finding that in 35 per cent of cases, other adults witnessed the violence and abuse being perpetrated against the victim (n=25; information missing for 12 cases). Although in some cases bystanders attempted to intervene, it was also common that they did not. For example, in Case 66 the offender’s brother was present in the home when the victim was killed and heard banging, shouting and the victim screaming. He told police that he remained in his bedroom throughout the night because he was afraid of the offender’s violent tendencies when drinking. A friend of the family also heard the violence when he was next door, but did not intervene.

Consistent with the international literature, reasons for bystanders not intervening when witnessing IPV included fear of being assaulted by the offender themselves (e.g. Case 41), being physically incapable of doing so because of their own high levels of intoxication (see Case study 3), and concerns about involving law enforcement due to their own illegal behaviours and/or fears of criminalisation (Banyard et al., 2020; McMahon & Dick, 2011). The latter were particularly common among First Nations communities.

However, that lethal incidents involving a PD offender typically occurred as part of ongoing and persistent patterns of abuse, that their friends and family members may have witnessed previously, may have de-sensitised these bystanders to the level of risk of serious and lethal harm being experienced by the victim. In other words, like the offender, bystanders themselves may have underestimated the victim’s risk of being killed.

While many PD offenders made instantaneous decisions to harm their partner, and the violence may have been consistent with previous non-lethal assaults, their responses to the violence diverged between remorseful and horrified by their actions, to cavalier and callous. Upon awaking and finding that the victim had died from her injuries, the offender would frequently express surprise and attempt to find assistance for them (40%, n=29; information missing for four cases). However, in 45 per cent of cases involving a PD offender, the offender did engage in concealment activities after discovering the victim had died (n=33; information missing for 11 cases; see Table 7). This typically included behaviours such as blaming other people for the murder (see for example Case study 2), hiding the body (e.g. Cases 167, 155 and 32), lying to law enforcement and others about their knowledge of the victim’s death and her whereabouts (e.g. Cases 169 and 43), and destroying evidence linking him to the crime (e.g. Cases 182, 50 and 51). Further, only two in five PD offenders expressed remorse for killing the victim (41%, n=30; information missing for 21 cases).

Further, as shown in Table 7, less than half of PD offenders overall pled guilty to murdering the victim (48%, n=35; information missing for nine cases). All 73 PD offenders were convicted of an offence, the most common being murder (66%, n=48). This means that one in three PD offenders were convicted of lesser charges, primarily manslaughter (33%, n=24). One in 10 PD offenders appealed their conviction (14%, n=10; information missing for two cases).

Table 7: Key characteristics of lethal incidents involving PD offenders (n=73)

Location of incident^a

	n	%
Victim’s/offender’s home	39	53
Public place	18	25
Other person’s home	12	16
Other	3	4
Presence of childrenb	14	19
Presence of adult bystandersc	25	34

Offender intoxicated^d

	n	%
Alcohol	43	59
Drugs	5	7
Both	13	18

Victim intoxicated^e

	n	%
Alcohol	35	48
Drugs	13	18
Both	16	22

Weapon used^f

	n	%
Hands and feet	38	52
Knife or other sharp instrument	26	36
Blunt instrument	18	25
Firearm	4	5
Fire	1	1
Rope	1	1
Drugs	0	0
Other	6	7
Presence of overkillg	39	53
Offender called for assistance for the victimh	29	40
Offender attempted to self-harmi	3	4
Offender pled guiltyj	35	48
Offender conviction	73	100
Murder	48	66
Manslaughter	24	33
Dangerous driving	1	1
Offender appealed convictione	10	14

^a Denominator includes one case where this information was missing.
^b Denominator includes 20 cases where this information was missing.
^c Denominator includes 12 cases where this information was missing.
^d Denominator includes nine cases where this information was missing.
^e Denominator includes two cases where this information was missing.
^f Denominator includes three cases where this information was missing.
^g Denominator includes eight cases where this information was missing.
^h Denominator includes four cases where this information was missing.
ⁱ Denominator includes 15 cases where this information was missing.
^j Denominator includes nine cases where this information was missing.
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

Pathway 3:
Deterioration/acute stressor

Approximately one in 10 offenders (11%, n=19) included in the sample were classified as following the deterioration/acute stressor (DAS) pathway. The trajectory underpinning these cases is described in Figure 10 and in detail below.

DAS cases typically involved a male offender who was described as historically having very low levels of aggressive and violent behaviours (and tendencies), including towards former intimate partners. These men were also unlikely to have been in contact with the criminal justice system for other forms of offending. However, the majority of offenders in this cohort had significant and co-occurring mental and physical health conditions at time of starting their relationship with the victim.

The relationships between victims and offenders, which were often longstanding, were described as overall positive and non-abusive. However, this started to change when the offender experienced the onset of a new stressor or problem in their lives, or the exacerbation of a pre-existing stressor or issue. Regardless of the source of this stressor, it had a significant and negative impact on the offender’s health and wellbeing, and their behaviours and emotional state more generally. This may have led to the onset of violence and abuse within the relationship (perpetrated by the offender against the victim), although this was relatively rare. Either because of the violence and abuse that had started within the relationship, or the overall decline in functioning of the offender, or both, conflict within the relationship increased.

Despite the increasing levels of conflict within the relationship, the offender’s deteriorating health and wellbeing and the onset of abuse perpetrated against the victim, when the offender entered the same space as the victim he did so without an obvious intent to harm her. Rather, the victim’s and offender’s occupation of the same space was consistent with their ongoing and committed relationships, often occurring in their shared residence. However, lethal violence risk for the victim was heightened due to the acute nature of the offender’s symptoms, in particular their intoxication levels, and the presence of an argument. In these circumstances, the offender appeared to become very emotionally distressed and murdered the victim using weapons immediately at their disposal. Afterwards, they were likely to demonstrate remorse for their actions, seek help for the victim and plead guilty to the offence.

Figure 10: Deterioration/acute stressor offender pathway

Offender characteristics (before relationship start)

Overall, the majority of DAS offenders were non-Indigenous (79%, n=15; information missing for two cases), and 53 per cent were born in Australia (n=10; information missing for two cases). One in three DAS offenders were born overseas (37%, n=7). At time of the lethal incident, the median age of DAS offenders was 49 years old (SD=14.6; see Table 10).

Similar to PD offenders, prior to meeting the victim it appeared that many DAS offenders had experienced significant physical and mental health issues. For example, two in three DAS offenders (68%, n=13; information missing for five cases) had previously been diagnosed with or had a suspected mental illness at time of starting their relationship with the victim. The most common forms of mental health conditions experienced by DAS offenders were depression, psychosis and adjustment disorders, personality disorders and post-traumatic stress disorder (PTSD). Further, approximately half of DAS offenders had an AOD (53%, n=10; information missing for six cases).

One in three DAS offenders had an LTHC at time of entering into a relationship with the victim (32%, n=6; information missing for 10 cases). Physical health conditions included arthritis (e.g. Case 183) and chronic pain (e.g. Case 10). In a small number of cases the physical health issues experienced by offenders was attributable to prior traumatic experiences. For example, in Case 161 the offender had been a prisoner of war and as a result of his treatment and torture during this period sustained long-term physical injuries as well as PTSD. Further, six DAS offenders had a suspected or diagnosed cognitive disability (32%; information missing for 11 cases).

Comorbidity among DAS offenders was also very common; 68 per cent of DAS offenders had two or more co-occurring mental and/or physical health issues (n=13; information missing for five cases). For example, at time of starting their relationship with the victim:

• 42 per cent of DAS offenders had both a mental illness (MI) and AOD issue (n=8; information missing for eight cases)
• 21 per cent of DAS offenders had both an MI and an LTHC (n=4; information missing for 13 cases)
• 16 per cent of DAS offenders had both an AOD and an LTHC (n=3; information missing for 12 cases).

Considering the presence of some of these significant and co-occurring physical and mental health conditions, it is not surprising that many DAS offenders were unemployed for extended periods throughout their lives. Further, it was reported that a small number were on a disability pension at time of the lethal incident occurring (e.g. Cases 149 and 102).

Although the histories of mental and physical health issues and AOD among DAS offenders are consistent with the PD cohort, a key difference is the absence of detected violence towards others. Overall, only two DAS offenders were described as being violent towards a previous intimate partner (11%; information missing for 12 cases) and four had been violent towards non-intimate partners (21%; information missing for 10 cases). Approximately half of DAS offenders had been involved in non-violent offending, which was described as relatively minor in nature (53%, n=10; information missing for three cases; e.g. traffic offences and theft). Reflecting on the relatively minor and primarily non-violent criminal offending histories of DAS offenders prior to the murder of their partners, judges noted that some offenders were “unremarkable” or had led relatively “normal” lives.

The victim and offender start their relationship

The majority of DAS offenders were married (53%, n=10) or in a de facto relationship (37%, n=7) with the victim. The median length of relationships was approximately 23 years (information missing for two cases; see Table 8). Half of DAS offenders (47%, n=9; information missing for one case) had at least one child with the victim (median=2, range=1–3), and were living with the victim on a full-time basis (89%, n=17). Taken together this indicates that DAS offenders were in long-term and committed relationships with their victims, and had been with them for significant periods of time prior to the lethal violence taking place.

Table 8: Characteristics of relationships (cases involving a DAS offender only; n=19)

Relationship types

	n	%
Married	10	53
De facto	7	37
Boyfriend/girlfriend	1	5
Date/casual	1	5
Unknown	0	0
Average relationship length (years; median)a	20.8 (23)
At least one childb	9	47
Median number of children (range)c	2.8 (1–13)
Cohabitation on a full- or part-time basis	17	89

^a Excludes two cases where this information was missing.
^b Denominator includes one case where this information was missing.
^c Limited to cases where the victim and offender had at least one child together.
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

In general, relationships between DAS offenders and victims were described as being overall “happy” and positive (e.g. Cases 149, 72, 161, 102 and 86). For example, in Case 86 the offender’s stepdaughter said that from her experience the victim and offender were happy in each other’s company and there was no conflict. Further, only two DAS offenders were reported to have been abusive towards the victim during the early stages of the relationship (11%). In Case 161 the offender was reported to have physically pushed the victim during an argument on one occasion, although no other incidents of violence or abuse were reported. In Case 117 the offender was described by the victim as “bullying” her throughout their relationship. However, in neither case was the DAS offender reported to the police or the subject of a protection order.

The offender’s mental and physical health begins to decline significantly

During the period preceding the lethal incident, DAS offenders’ physical and mental health appeared to decline significantly. This was primarily attributed to the symptoms associated with offenders’ pre-existing physical and mental health conditions getting worse. For example, in some cases DAS offenders who already had an AOD began consuming alcohol more frequently (e.g. Cases 152, 149 and 10). In other cases, the offender began consuming illicit substances again after a period of abstinence, increased their use of illicit substances significantly, or tried different drugs for the first time (e.g. Cases 154 and 44). Overall, 70 per cent of offenders who had a pre-existing AOD experienced the escalation of symptoms (n=5; information missing for three cases). Meanwhile, it was reported in a small number of cases that the offenders’ physical health had continued to deteriorate as they aged (Cases 86 and 183).

Further, almost all DAS offenders who had a pre-existing MI experienced the escalation of symptoms (92%, n=12; information missing for one case). In some cases, the escalation of the offender’s mental health issues was partially attributed to the offender’s increased alcohol and/or drug use (e.g. Case 72), while in another case the offender was reported to have stopped taking his medication for his diagnosed depression (e.g. Case 198).

Crucially, it appeared that for some DAS offenders the onset or escalation of a pre-existing health condition contributed to the onset or escalation of other problems or stressors in their lives. In other words, in some cases the offender was dealing with multiple and interrelated stressors and issues. For example, in Case 44 the offender began using steroids more frequently during the lead-up to the lethal incident, which led to his self-diagnosed testicular atrophy and sexual impotence. This in turn appeared to contribute to his use of methamphetamines. Further, as described below, in Case 183 the offender’s increased consumption of alcohol led to the loss of his job (see Case study 4).

While for the majority of DAS offenders the decline in their physical and mental health was due to the exacerbation of symptoms associated with historical conditions, in other circumstances it was attributable to the offender’s experience of a new and unique stressor. For example, in Case 169 the offender’s and the victim’s relationship was described as being positive and loving until the offender was imprisoned for visiting a brothel in his home country of China. As a result of his imprisonment, the offender lost his high-status job, and experienced deep shame. He was also diagnosed with depression and PTSD as a result of his experiences while imprisoned. Further, in Case 117 the offender developed PTSD after his experiences in various conflict zones, which later developed into a delusional disorder.

The offender’s behaviours and emotional state change

The deterioration in the health and wellbeing of DAS offenders, as well as the onset of new acute stressors, appeared to have a negative impact on their behaviours and attitudes. For example, in Case 152 the offender was described by the sentencing judge as having led a “blameless life” until the age of 37 when his alcohol consumption increased and it started to affect his behaviour in negative ways. In particular, he became more erratic and less committed to his job and family.

The stressors experienced by DAS offenders also appeared to have a negative impact on the offenders’ behaviours towards the victims specifically. For example, in Case 72 the offender’s escalating symptoms associated with his diagnosed schizophrenia led to the offender becoming paranoid that the victim was having an affair. Because of this, he argued with her constantly and accused her of infidelity. Meanwhile, in Case 198, after the offender started smoking cannabis and stopped taking his antidepressants, he started to become verbally aggressive towards the victim.

Violence and abuse within the relationship starts for the first time

As demonstrated in some of the above described cases, occasionally the change in behaviours towards the victim involved the onset of violent and abusive behaviours perpetrated by the offender (21%, n=4). In all four cases, the offender was emotionally abusive towards the victim, and in two cases the offender planned or attempted to self-harm. For example, in Case 152, in the weeks leading up to the lethal incident the offender was reported to the police and the subject of a protection order after an incident where the offender verbally abused the victim, and attempted to strangle their son. Only one other DAS offender in the entire cohort had been reported to the police for abusive behaviours toward the victim during this stage of the trajectory. This indicates the very low levels of visibility that law enforcement had of these relationships and the risk posed by these offenders to their partners.

Conflict between the victim and offender increases

In the context of the offender’s deteriorating health and wellbeing, and the onset of violence and abuse within the relationship, it is perhaps unsurprising that conflict between the victim and offender began to increase. In most cases, this conflict appeared to centre on the impact of the offender’s mental and physical health issues and other associated stressors on the offender’s behaviours, and the victim and their relationship more generally. For example, in Case 152, as the offender started to drink more, the victim encouraged him to participate in treatment services which he did not appear to be motivated to do. Further, in Case 198, in response to the offender becoming increasingly aggressive towards her, the victim encouraged him to change his behaviours, and even asked a family friend who was also a faith leader to talk to the offender about his behaviours.

As shown in Figure 8, the increase in conflict within the relationship and changes in the behaviour of the offender were interrelated, and the order in which these stages of the trajectory occurred differed across cases. For example, in some cases it appeared that as the victim encouraged the offender to seek support for their issues (conflict) the offender’s behaviour towards them changed and they may have become more hostile and aggressive. In other cases the changes in the offender’s behaviour towards the victim led in turn to increased conflict within the relationship. Alternatively, it could be that these stages in the trajectory co-occurred.

Despite the increased conflict within the relationship, separation between the victim and offender was relatively rare. At time of the lethal incident, only two cases involved a DAS offender who was no longer in a relationship with the victim. As shown in Figure 11, in 68 per cent of cases (n=13; information missing for one case) the victim and offender had not separated at any point in their relationship, while in 16 per cent of cases (n=3) the victim and offender had reconciled after a period of separation at time of the lethal incident. This further contributes to the overall impression of relationships between DAS offenders and victims as being long-term and committed until time of the lethal violence occurring.

Figure 11: Relationship status at time of lethal incident (cases involving a DAS offender only; n=19; %)

Relationship status	Cases involving a DAS offender only (n)	Cases involving a DAS offender only (%)
Never separated	13	68
Reconciled	3	16
Separated	2	11
Negotiating reconciliation	0	0

Note: Denominator includes one case where this information was missing.
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

The offender enters the same space as the victim without a clear motivation to perpetrate significant harm against her

Against a backdrop of the offender’s declining mental and physical health and increasing levels of conflict within the relationship, because the victim and offender typically remained together they continued to occupy the same spaces on a regular basis. In particular, at time of the lethal incident, the victim and offender were typically together in their shared residence and had been cohabiting without issue for a number of hours prior to the victim being killed. The offender’s access to the victim did not require subterfuge on the offender’s part, or forced entry.

Although they had typically been spending time together peacefully, it was commonly reported that at some point the offender’s emotional regulation skills became significantly impaired. For example, half of DAS offenders were highly intoxicated at time of the lethal incident occurring (58%, n=11; information missing in two cases), which increased to 90 per cent among men who had previously been identified as having an AOD (n=9). In some cases, it was also reported that at time of entering the same space as the victim, the offender was experiencing acute symptoms associated with underlying mental health issues. For example, in Case 72 the offender was reported to have been psychotic at the time of the lethal violence occurring, which was exacerbated by his use of amphetamines.

Frequently, after entering the same space together, the victim and offender would then start to argue about something. Although the offender was occasionally unclear on what the argument was about, in some situations the focus was on the offender’s intoxication. In other cases the argument focused on the victim’s dissatisfaction with the relationship and her telling the offender that she was intending to leave him. For example, in Case 169 the offender had flown to visit the victim in Australia and they had spent some time together shopping, eating and having sex. Later in the evening she left the apartment to visit her other lover (who the offender was aware of), ostensibly to break up with him. However, when she returned she told the offender that she had decided to remain with her lover and that she was instead leaving him. This resulted in a long and protracted argument, and her eventual murder. Further, in Case 117, the victim accused the offender of bullying her throughout their relationship and as a result she was going to leave him for the sake of her own sanity. In a reaction of “acute psychological distress and anger”, the offender killed the person who he later described to police as his “soulmate”. The Crown argued the offender’s attachment to his wife was connected to a need to control her.

Taken together, in the presence of an argument, the offender’s high levels of intoxication and other acute symptoms associated with their underlying conditions, the offender became very emotionally heightened and decided to harm the victim. Perhaps reflecting the relative instantaneous nature of their decision to murder the victim, DAS offenders often used weapons immediately at their disposal. The most commonly identified weapon for incidents involving a DAS offender was a knife (typically taken from the kitchen; 53%, n=10), followed by a blunt instrument (26%, n=5; see Table 10).

In further evidence of the potentially spontaneous decision of the offender to kill the victim, 42 per cent of DAS offenders sought assistance for the victim, including calling the police or ambulance services, after the lethal violence (n=8; information missing for one case). Two in three DAS offenders were assessed by sentencing judges as demonstrating genuine remorse for their actions (63%, n=12; information missing for five cases). As described in Case study 4, evidence of the offenders’ remorse included demonstrations of emotional distress related to their actions, apologising for their actions and also holding themselves accountable for their actions. Sixty-three per cent of DAS offenders pled guilty to murder at the first court hearing (n=12; information missing for two cases) and appeals against conviction were only applied for in two cases.

Table 10: Key characteristics of lethal incidents involving DAS offenders (n=19)

Location of incident

	n	%
Victim’s/offender’s home	17	89
Other person’s home	1	5
Public place	1	5
Presence of childrena	5	26
Presence of adult bystanders	6	32

Offender intoxicated^b

	n	%
Alcohol	6	32
Drugs	3	16
Both	2	11

Victim intoxicated

	n	%
Alcohol	3	16
Drugs	2	11
Both	1	5

Weapon used^c

	n	%
Knife or other sharp instrument	10	53
Blunt instrument	5	26
Hands and feet	3	16
Rope	3	16
Firearm	0	0
Fire	0	0
Drugs	0	0
Other	2	11
Presence of overkill	14	74
Offender called for assistance for the victima	8	42
Offender attempted to self-harmb	5	26
Offender pled guiltyb	12	63
Offender conviction	19	100
Murder	16	84
Manslaughter	2	11
Recklessly endangering life	1	5
Offender appealed conviction	2	11

^a Information missing for one case.
^b Information missing for two cases.
^c Percentage totals do not equal 100 because offenders may have used more than one weapon.
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

Crossover cases

Fifteen per cent of cases (n=27) involved elements of two or three of the above described primary pathways. As shown in Figure 12, the most common “type” of crossover cases were those that involved elements of the FT and PD pathways (48%, n=13), followed by cases involving elements of the FT and DAS pathways (30%, n=8), and then cases involving elements of the PD and DAS pathways (15%, n=4). Only two cases involved elements of all three pathways (7%).

Figure 12: Breakdown of crossover cases (n=27; %)

Crossover cases	n	%
FT/PD	13	48
DAS/FT	8	30
PD/DAS	4	15
FT/PD/DAS	2	7

Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

Fixated threat/persistent and disorderly

Consistent with the PD pathway, most FT/PD offenders had significant histories of trauma, AOD and mental health conditions. Overall, 74 per cent had an AOD (n=14; information missing for three cases), 68 per cent had an MI (n=13; information missing for five cases), and 26 per cent had witnessed or been the target of family violence in their family of origin (n=5; information missing for 13 cases).

Further, prior to their relationship with the victim, it was likely that they had been:

• violent towards former partners (47%, n=9; information missing for 10 cases)
• violent towards non-intimate partners (e.g. family members and acquaintances; 32%, n=6; information missing for 11 cases)
• involved in non-violent offending (32%, n=6; information missing for four cases).

As such, like many PD offenders, FT/PD offenders were highly visible to the criminal justice system, and were perceived as dysfunctional in different domains of their life.

The nature of relationships between FT/PD offenders and victims were relatively diverse within this cohort. In some cases, the offender and victim had been in a relationship for an extended period of time and had children together (e.g. Cases 151 and 180), and in other cases the victim and offender had only been together for a relatively short period of time. For example, in Case 110 the victim and offender had only been together for 5 to 6 weeks prior to the lethal incident. However, consistent with both the FT and PD trajectories, the relationships were characterised as abusive (95%, n=18). Overall, 74 per cent of FT/PD offenders had been physically abusive towards the victim during their relationship (n=14; information missing for three cases), and four in five had been emotionally abusive towards (79%, n=15; information missing for four cases) and/or controlling of the victim (84%, n=16; information missing for three cases).

A key characteristic of the FT/PD cases, which was notably similar to the FT trajectory, was that the violence and abuse in the relationship appeared to increase and change in nature when the offender perceived that his level of control over the victim was slipping. This was primarily attributable to the victim leaving the offender, or indicating her intent to do so. In 47 per cent of FT/PD cases the offender was separated from the victim at time of the lethal incident (n=9; information missing for two cases).

Where a separation occurred or was imminent, the offender became more controlling of the victim, and would stalk her in an attempt to reconcile with her. Again consistent with the FT trajectory, the level of risk posed by offenders appeared to have been anticipated and understood by this period (44%, n=4; information missing for three cases).

Like many FT offenders, in the context of the offender’s increasingly desperate attempts to reconcile with the victim and/or increasing levels of anger at the victim, FT/PD offenders entered the same space as the victim with a clear intent to seriously harm her. For example, 16 per cent of FT/PD offenders engaged in planning activities prior to entering the same space as the victim (n=3; information missing for six cases) while one in three offenders threatened to kill the victim in the lead-up to the lethal incident (32%, n=6; information missing for 10 cases). Further, after the incident, one in two FT/PD offenders engaged in concealment activities, such as hiding the body and murder weapons (53%, n=10).

Deterioration/acute stressors/fixated threat

DAS/FT cases were diverse in nature; different cases involved different combinations of elements from both the DAS and FT trajectories. Consistent with the DAS trajectory, none of the DAS/FT offenders had been described as being abusive towards a former partner (information missing for eight cases), although some had been violent towards non-intimates (15%, n=2; information missing for five cases) and/or been involved in non-violent offending (31%, n=4; information missing for one case). Another similarity between DAS/FT and DAS offenders is the high prevalence of mental illness, prior to the relationship with the victim starting (54%, n=7; information missing for three cases), and LTHCs (31%, n=4; information missing for eight cases).

However, consistent with the FT trajectory, there was evidence that many DAS/FT offenders had been abusive towards the victim, the most common form of abuse being coercive control (e.g. Case 168). However, in other cases there did not appear to be any violence and abuse within the relationship, which is more consistent with the DAS trajectory (e.g. Case 170).

In all DAS/FT cases, there was a point within the relationship where the offender experienced the exacerbation of pre-existing stressors and physical and mental health issues, as well as the onset of new ones. In most cases, this stressor was financial in nature (e.g. Cases 168 and 192). For example, in Case 82, the offender ran his own company, was in significant debt to his friends and had just been refused a personal loan by his bank in the lead-up to the lethal incident.

Consistent with the DAS trajectory, it appeared that in some cases the lethal incident occurred in the context of the offender’s deteriorating mental and physical health and wellbeing, as well as increasing levels of conflict within the relationship. However, in other cases it appeared that while these factors were indeed present, the main contributing factor to the lethal incident was the offender’s increasing levels of anger towards the victim. In these cases, the offender appeared to blame the victim for their loss of control within the relationship, or in other domains. For example, in the above cited case (82), the offender’s stress was attributable to his ongoing affair with another woman who was pressuring him to leave the victim. However, he did not believe he could for financial reasons. It is possible that part of the offender’s motivation for killing the victim was because her mere presence was a source of stress in his life because he wanted to leave her to be with his mistress.

Persistent and disorderly/deterioration/acute stressors

In the small number of cases that involved elements of both the PD and DAS trajectories, all four offenders had reported histories of AOD and mental health-related issues, as well as contact with the criminal justice system for prior abuse towards former partners and non-intimates, and non-violent offending. This is consistent with the PD trajectory. Further, like in the PD trajectory, all four of the offenders had been violent towards their partner during the course of the relationship, including acts of physical violence and emotionally and verbally abusive behaviours.

Consistent with the DAS trajectory, at some point within the relationship the offender experienced escalating symptoms associated with underlying chronic mental and physical health conditions, as well as the onset of new stressors. For example, in Case 191 the offender’s stepfather had died shortly prior to the lethal incident, and his alcohol and drug use increased significantly.

At time of the lethal incident, it was unlikely that PD/DAS offenders had separated from the victim, so their access to them did not require subterfuge or force. Rather, the victim and offender came together in the context of their ongoing relationship and regular routines. While in the same space together the offender was typically intoxicated and became emotionally heightened due to an argument with the victim, and they made an instantaneous decision to kill them.

Fixated threat/persistent and disorderly/deterioration/acute stressors

Only two cases were classified as involving elements of all three trajectories. Consistent with the PD trajectory, both offenders had histories of traumatic experiences in their childhood and adolescence, although only one had detected AOD issues and had been violent towards a former partner (Case 165). In comparison, the other offender (Case 26) had not been violent towards a former partner. However, this may have been attributable to the offender’s very young age – he was an adolescent at time of the lethal incident occurring and it appeared that the victim was his first intimate relationship.

After the relationship started, both of the offenders experienced significant events that had a negative impact on their mental health and wellbeing. In both cases, this event was the death of a family member they were very close to. In Case 26 the offender also experienced the loss of his job which was viewed as having a deleterious impact on his wellbeing as he lost his sense of purpose.

In the context of the offenders’ deteriorating mental health and wellbeing, their relationship with the victims started to break down and the offenders’ level of control over the victims was challenged. In Case 165 the offender’s distress over his loss of control over the victim was exacerbated by her pregnancy to the offender. More specifically, the offender in this case repeatedly called and texted the victim and sought to reconcile, under the guise of concern for his unborn child and his suspicions that she was continuing to use drugs.

The loss of control and experience of significant and negative events appeared to lead these two offenders to become highly motivated to reassert their control over the victim, and punish them for their non-compliance. In both cases the offender entered the same space as the victim using subterfuge or force and killed them. In Case 26, the offender also killed a bystander who had attempted to assist the victim.

Themes associated with intimate partner homicide

Although the analysis identified three trajectories that highlighted some of the key differences in the IPH pathways, there were also several important themes that appeared to have a role in explaining the occurrence of IPH, regardless of the offender classification. These were:

• the emotional, mental and physical health of offenders
• traumatic experiences among offenders
• pre- and post-migration experiences of victims and offenders
• separation
• hegemonic masculinities and traditional gender norms.

Key theme 1:
The emotional, mental and physical health of offenders

The PIPH sample excluded cases where the offender was found not guilty by reason of mental illness. However overall, three quarters of IPH offenders were described as having experienced at least one emotional, mental or physical health condition during their lifetime (73%, n=146; information missing for 45 cases). Common conditions included chronic and acute physical illness, cognitive impairment, intellectual disability, mental health disorders and alcohol and other drug dependence. These lived experiences often overlapped, intersected and impacted on an offender’s capacity and agency to varying degrees.

The following discussion of offenders’ emotional, physical and mental health problems is not an attempt to rationalise, pathologise or make excuses for the harm they perpetrated. The aim is to explore the interrelated risk factors for IPV and IPH as they presented among offenders in this sample. In part, this will be achieved by exploring the role of emotional, mental and physical health in pathways to, and incidents of, IPH. It is important to balance the need for offender accountability with the reality that many of these individuals suffered severe health conditions that negatively impacted their capacity and agency.

Long-term health conditions

The study identified that many of the male offenders had experienced a range of physical and physiological health conditions. Overall, one in five offenders reported experiences of LTHC, including physical disability, at some point during their lives (19%, n=38; information missing for 135 cases). Meanwhile, one in 10 offenders experienced the onset or escalation of an LTHC while they were in the relationship with the victim (10%, n=19; information missing for 155 cases).

There was a notable interrelationship between age and experiences of physical health conditions among offenders. Among offenders aged 55 and over (n=29), approximately a third reported some kind of physical health condition (31%, n=9; information missing for 16 cases). In the other age groups, physical health problems were experienced by:

• one in five offenders who were 45–54 years old (21%, n=9; information missing for 24 cases)
• one in six offenders who were 34–45 years old (16%, n=11; information missing for 48 cases)
• one in six offenders who were 25–34 years old (17%, n=7; information missing for 28 cases).

In comparison, no offenders who were 18–24 years old reported physical health problems.

In cases where offenders had physiological health conditions these included diabetes, hypertension, renal disease, liver disease, tachycardia, prostate problems, blood diseases, paralysis, hemiplegia and leprosy. For other offenders, ongoing physical health issues were attributed to accidents or injuries that occurred in the workplace, while playing sport or in the context of non-IPV related violence (e.g. fighting, assault). For some, the consequences of traumatic accidents were severe and associated with ongoing pain and disability. For example, in Case 59 the offender survived a house fire but suffered serious burns to nearly half his body and the loss of fingers from his hands.

The long-standing health problems experienced by many of the offenders in the sample limited their agency to navigate and manage day-to-day life. Often, health challenges prevented them from working, supporting their families and contributing to society more generally. For example, in Case 58 the offender suffered a significant back injury and resulting chronic pain which prevented him from being able to continue working as a labourer which he had done all his life. His inability to support his family contributed to the onset of depression, which in turn led to the deterioration in the relationship between the victim and offender. This offender was classified as FT as the homicide was motivated by his anger and inability to deal with the breakdown of the relationship. However, intersections between chronic pain, depression and unemployment were important factors influencing his trajectory. While the psychological impact of a series of events such as these can be significant, it remains that the vast majority of people who experience chronic health conditions and severe stress related to these experiences do not kill their partners.

The findings from this research on the physical health of IPH offenders contribute new evidence to a sparse literature in this space. Other research has shown that one in 10 (10.5%) IPH offenders were on a sick or disability pension at the time of the incident (Caman et al., 2017). However, few studies have explored physical health characteristics and trends in chronic physical illness among IPH offenders in particular. While other established risk factors for IPH such as substance use and personality disorders have received more scholarly attention (Aldridge & Brown, 2003; Kivisto, 2015), there is a lack of research examining physical health correlates of IPH perpetrators against which to compare our findings (Singh et al., 2014). Improving understanding of the relationship between chronic health conditions and IPH is a critical area of focus for future research and necessary for identifying suitable intervention and prevention initiatives.

Cognitive disability and acquired brain injury

One in six offenders (21%, n=41; information missing for 132 cases) had a history of cognitive impairment or disability. This ranged from traumatic brain injuries to diagnosed learning or intellectual disabilities, average or borderline IQ scores, and issues with literacy and numeracy. Many of these issues were identified retrospectively through pre-sentencing assessments which occasionally revealed that the offender had difficulties with tasks related to executive functioning such as slowed processing speed and reduced working memory. Only one offender was reported to have a neurodegenerative disease at time of the lethal incident (i.e. dementia).

Late-onset cognitive problems typically occurred in the context of alcohol-related brain damage and disability resulting from traumatic events. In one in 10 cases (11%, n=21), accidents or assaults resulted in a suspected or diagnosed ABI (information missing for 148 cases). An ABI is defined as brain damage that occurs after birth which is distinct from, but related to, intellectual disability and mental illness (Australian Institute of Health and Welfare [AIHW], 2007). An ABI can result in structural and functional changes to the brain but the causes and consequences can vary greatly from concussions to surgery, stroke, aneurism or a brain tumour. Disabilities associated with ABI may be cognitive, sensory-motor or affective (Ciuffreda & Kapoor, 2012). Depending on the part of the brain affected and the extent of the injury, ABI can limit higher level processing and result in changes in thinking, memory and behaviour. Impairment to the frontal region of the brain is associated with heightened impulsivity; aggression; and deficits in self-regulation, planning and risk assessment (Brower & Price, 2001).

Research examining neuropsychological factors in IPV has found that more severely aggressive IPV offenders (i.e. borderline–dysphoric and generally violent antisocial) were more likely to report histories of head injuries and lower scores on verbal intelligence (Walling et al., 2012). This is in line with our research showing that ABI was more commonly reported for the PD pathway than the FT and DAS pathways (see Table 11). Reviews of the literature suggest that 36–58 per cent of family violence perpetrators have a traumatic brain injury, however this is based on a small number of studies (n=2; Ayton et al., 2021). While ABI is more common among offending populations there is limited evidence about ABI among family violence and IPH perpetrators in particular (State of Victoria, 2014–2016).

Caregiver roles and relationship dynamics

Previous research has found that alongside social maladjustment and related health conditions, a caregiver role is associated with increased risk of offending among older homicide offenders and IPH offenders in particular (Benbow et al., 2019; Nguyen et al., 2021). In cases where the offender or victim had a physical health condition, either the victim or the offender was the primary carer for the other. This was primarily the victim caring for the offender but there were instances where the victim had significant health challenges and the offender was their primary support. For example, in Case 150 the victim had multiple health conditions for which she was frequently hospitalised in the year prior to the homicide. The offender administered the victim a lethal dose of heroin when she was in significant pain after leaving hospital against medical advice.^[1]

Serious illness and disability constitute significant stressful events for both caregivers and those directly impacted. In response to such stress, people differ in their perceived capacity to cope. The current analysis found that caregiving dynamics were significant in cases where the victim and/or offender were older. For example, in Case 86 the couple had been together for over 25 years, and he had a chronic health problem for which she provided him daily care. He was classified as a DAS offender because the marriage was reportedly happy without longstanding animosity but conflict related to the caregiving dynamic and the escalation of his health problems precipitated the homicide. This is in line with research showing the level of dependency the offender has on the victim (balance of power) and their ability to handle stress influences the risk of IPV and IPH among older offenders (Benbow et al., 2019; Poole & Rietschlin, 2012).

Emotional and mental health

Mental health problems and disorders create significant and complex burdens for people who experience them, their families and their carers. Mental health disorders are defined in the Diagnostic and Statistical Manual of Mental Disorders (5th edition; DSM-5) as psychological disorders that impact on emotions, cognitions and/or behaviour; cause significant distress and impairment in various areas of life; and are associated with increased risk of pain, morbidity and mortality (American Psychiatric Association, 2013).

Mental health problems were commonly experienced by offenders in this study. In cases where the offender’s mental health was discussed, sentencing judges would generally refer to mental health assessments and pre-sentence reports in their remarks. Almost half of offenders in this study were described as having diagnosed or suspected mental health problems (43%, n=86; information missing for 96 cases). Further, over one in four offenders reported the onset or escalation of mental health problems during their relationship with the victim (28%, n=56; information missing for 129 cases; see Table 11). This is consistent with other research which has identified elevated rates of mental health problems among prisoners compared to the community, IPV offenders generally and IPH offenders in particular (Butler et al., 2007, Sesar et al., 2018; Spencer & Stith, 2018).

Our findings align with Australian and international research showing one third of IPH offenders are estimated to have a mental health disorder at some point in their life (Cullen et al., 2019), with lower rates of psychotic disorders, and higher rates of mood disorders (Oram et al., 2013). The role of mental health issues in the occurrence of IPH is unclear, although a number of key mechanisms emerged from the analysis. These are explored below.

Depression and psychological distress

The analysis study found that symptoms of depression and significant psychological distress were commonly experienced by IPH offenders in the PIPH sample. Offenders had suspected or diagnosed depression (26%, n=52) and self-harming behaviours (12%, n=23) at significantly higher rates than in the general population (Australian Bureau of Statistics, 2018). These findings are consistent with the international literature which has consistently found evidence of an association between mood disorders such as depression and IPH; global population estimates range from 17 to 56 per cent among men who kill their partners (Kivisto, 2015).

Despite the high rates of depression in this study and among IPH samples more broadly, the relationship between depression and violence is not straightforward (Elbogen & Johnson, 2009; Fazel et al., 2015). Like other mental health conditions, depression is experienced on a continuum with mild symptomology alongside relatively normal functioning at one end to chronic and severe disablement at the other. The analysis found that in most cases where an IPH offender had a diagnosis of depression, it was described as being severe and characterised by periods of acute symptomology and co-occurrence with other mental and physical health issues and AOD. For example, in Case 160 the offender struggled with severe depression and had a history of self-harming behaviour. He had attempted suicide previously which was almost successful and required hospitalisation. However, he did not engage with treatment and discontinued the use of his prescribed medication, self-medicating with alcohol instead. The nature of comorbidity as experienced by IPH offenders will be discussed in more detail in a following section of the report.

In understanding the links between depression and IPV and IPH, it has been noted that depressed IPH perpetrators have more risk factors than non-depressed IPH perpetrators (Cheng & Jaffe, 2019). For example, depressed IPH offenders are also older, have threatened suicide in the past, been victimised as a child and experienced sexual jealousy (Cheng & Jaffe, 2019). As such, depression may co-occur with other risk factors associated with IPH which increases the likelihood of lethal violence occurring.

Alternatively, the link between depression and IPH could be explained by the symptoms associated with the disorder itself. Depression as it is experienced by IPH offenders is often accompanied by an intense fear of being abandoned, instability, self-harm and suicidality (Belfrage & Rying, 2004). This was evident in Case 41 where the offender had a strong desire for acceptance by others but this was often prevented by his intense fear of humiliation and criticism, his emotional detachment and his tendency to isolate himself socially. All of these factors, the judge noted, did not provide an ideal foundation for him to deal with his illicit drug problem and aggressive tendencies.

Interestingly, despite the research linking depression with IPV and IPH, sentencing judges often minimised the likely impact of the condition on offenders’ decision-making. In some cases, judges normalised common mental health problems like depression, noting the high prevalence of this experience in the wider community. This is demonstrated in the following passage from Case 160:

I accept that the offender’s depressed state led him to take a bleak view of events and to feel pessimistic about his life and circumstances. I do not, however, regard the evidence as to his depressive state as being capable of significantly mitigating the offender’s moral culpability. Depression is a relatively common illness. The offender, in common with many other Australians, suffered and continues to suffer from it. It cannot however, in the circumstances of this case, provide some answer for the offender’s crimes. It provides context, but can do little to mitigate these offences.

However, experiences of depression among IPH perpetrators need to be understood within the context of other clustering risk factors, such as comorbid mental health and substance use problems. This work supports previous evidence that perpetrator depression may signal lethality risk for IPV and IPH and constitutes an important intervention point and treatment need for high-risk offenders.

Personality disorders

One in 14 offenders included in the PIPH sample had a suspected personality disorder that was ongoing at time of the lethal incident (7%, n=13; information missing for 96 cases). This is higher than national community estimates of 6.5 per cent (Jackson & Burgess, 2004). Personality disorders are characterised by disturbances in how individuals think about themselves and others, and are commonly associated with difficulty in psychosocial functioning and navigating interpersonal relationships (Tyrer et al., 2015). The primary personality disorders identified were borderline personality disorder, antisocial personality disorder and personality disorders characterised by schizoid, paranoid, depressive and obsessive-compulsive traits.

Research shows that personality disorders are among the strongest predictors of interpersonal violence (Fazel et al., 2018). Increased risk of violence among individuals experiencing personality disorders has been attributed to symptoms such as impulsivity (emotional dysregulation), low IQ and attentional problems, antisocial attitudes, rigidity, psychopathy and delusional ideation that are experienced along a continuum of personality dysfunction (Glenn et al., 2013; Howard, 2015). This was evident in Case 50 where the offender was described as having a personality style characterised by a lack of empathy and remorse, impulsivity and psychopathic features. The offender’s disregard for the rights of others and the law was evidenced by multiple breaches of protection orders made against him. He had completed a court-ordered family violence program but did not benefit from it and was described as posing a high risk of violent reoffending.

Reviews of the research also suggest that borderline, over-controlled and dependent personalities are associated with IPV severity and IPH risk (Aldrige & Browne, 2003). According to this work, many personality disordered men who kill their partners do not have recorded criminal or otherwise antisocial behaviour histories (Aldrige & Browne, 2003). In Case 189, the offender was described as having a personality disorder characterised by obsessive compulsive traits, presenting as a strong need to control his external environment and difficulty managing stress. While he had minimal criminal history, he demonstrated a pattern of responding disproportionately, excessively and inappropriately (such as losing his temper and verbally abusing his co-workers) when a task was not done to his standard. This case highlights that men without criminal histories who are at risk of perpetrating IPH may demonstrate antisocial behaviour in other realms of their lives, not just in intimate relationships. This indicates an opportunity outside of the relationship to provide support such as help with anger management and emotional dysregulation.

Emotional dysregulation

While emotional regulation is the automatic capacity to intentionally or unconsciously manage emotional states, emotional dysregulation relates to difficulty managing emotional states in the moment (Gross & Thompson, 2007). This research identified that experiences of emotional dysregulation were commonly experienced by offenders in the PIPH sample. Difficulty regulating emotions was related to offenders’ emotional and mental health problems, physical health conditions and AOD. These findings are in line with research showing that IPV offenders with emotional dysregulation problems can present on different offending trajectories but are more likely to commit IPH compared to IPV offenders without emotional dysregulation (Vignola-Levesque & Lévéillee, 2021).

Research shows that deficits in higher order processing skills such as goal setting and impulse control are associated with the onset and maintenance of antisocial behaviour more broadly and IPV specifically (Bushman et al., 2001; Gillespie & Beech, 2016; Gratz & Roemer, 2004; Holtzworth-Munroe & Stuart, 1994; Marín-Morales et al., 2021). The interaction between biological and environmental vulnerabilities can make some people more sensitive to emotional stimuli in their environment, thus increasing the intensity of the emotional response and making it harder for them to return to their “normal” state after being aroused (Glenn & Klonsky, 2009). For example, in Case 41 the offender began using alcohol and other drugs from an early age. His continued heavy use detrimentally impacted on his ability to manage his emotions and impulses. Over the years he had developed a pattern of becoming angry and aggressive after using alcohol and other drugs, subsequently apologising and promising to cease use but then relapsing again soon after. Similar observations of early substance use detrimentally impacting offenders’ ability to manage their impulses and emotions was also noted in Cases 41 and 38. This evidence supports past work showing alcohol use problems mediate the relationship between emotional dysregulation and IPV and contribute to the cycle of violence (Grigorian et al., 2020).

“People who experience physiological and psychological health conditions can have more difficulty regulating their emotions and therefore may be more likely to use aggression as a way of communicating their feelings …”

The Catalyst Model of Aggression posits that some people have a lower threshold to violent behaviour due to individual and environmental risk factors (Ferguson & Dyck, 2012). People who experience physiological and psychological health conditions can have more difficulty regulating their emotions and therefore may be more likely to use aggression as a way of communicating their feelings (Gratz & Roemer, 2004). In the context of separation (a critical risk period for IPH), offenders who have mental health and emotional dysregulation problems are more vulnerable to difficulty processing the grief that can follow estrangement. Specifically, an inability to effectively regulate emotions such as distress, abandonment and rejection can increase the likelihood of self-destructive and violent behaviour (Vignola-Levesque & Léveillée, 2021). As demonstrated in Case 145, the offender and victim were only in a relationship for five months, however his attempts to control and isolate her began early in their relationship. He was verbally and emotionally abusive, he threatened to commit suicide if she left and when she did leave, he harassed her by calling her multiple times a day in the lead-up to the homicide. This is an example of how an offender who has a maladaptive coping style and difficulties with real and perceived rejection can pose a high risk of harm to significant others in the immediate period following estrangement.

Psychotic disorders and symptoms

Experiences of psychosis, delusions, morbid jealousy and paranoia had significant impacts on IPH offenders, their relationships and the homicide incidents. Some offenders had a history of diagnosed or suspected schizophrenia (4%, n=7) and a minority were diagnosed with a delusional disorder characterised by morbid jealousy or paranoia (2%, n=4). These rates are less than those identified in previous research showing that nearly one in 10 IPH offenders are experiencing delusions of some nature at the time of the homicide (Bricknell & Doherty, 2021). However, these lower rates may be due to cases where the offender was found not guilty by reason of mental illness being excluded from our sample. Drug- and alcohol-induced psychosis and schizophrenia featured in some pathways and incidents (e.g. Cases 191 and 102). In Case 154 the offender was determined to be experiencing methamphetamine-induced psychosis around the time of the homicide. He was described as having had delusional thinking for some time, a persecutory belief system and auditory hallucinations telling him to do things.

Delusional disorders experienced by IPH offenders in this sample typically presented as an extreme obsession with their partners’ perceived infidelity. In these cases, the victim was perceived by the offender to be unfaithful, planning to leave the relationship and take their children away. For example, in Case 98 the offender was diagnosed with a delusional disorder where he perceived that the victim was attacking him at the time of the murder. It was reported that he believed that she had drugged him so he was unable to sleep for three days preceding the homicide, that she was having an affair, and that she was scheming with relatives to take his money and his children and poison him. While it was determined that the delusional disorder would have impaired his capacity to control his actions, his behaviour during the incident indicated that he was not deprived of the capacity to know he should not have killed the victim and therefore he was fit for trial.

In cases where perceptions of infidelity, betrayal and abandonment preceded the lethal incident, these were described as a motivating force for the offender. The offender frequently fixated on the victim as the focus of these fictitious beliefs and for them this provided a rationale for their lethal violence.

Alcohol and other drug disorders

Just over half of offenders in the PIPH sample had histories of alcohol and other drug use in their lifetime (54%, n=107; information missing for 67 cases). One quarter of offenders (25%; n=49; information missing for 130 cases) experienced the onset of a new or exacerbation of a pre-existing alcohol and other drug problem during their relationship with the victim.

Alcohol use and intoxication

The offender had consumed alcohol or drugs in half of all IPH incidents analysed (48%, n=95; information missing for 33 cases). These findings are in line with reviews of the research showing that alcohol is implicated in 44 to 70 per cent of IPH incidents (Kivisto, 2015). In Australia, up to two thirds of all family violence incidents involve alcohol (Curtis et al., 2019; Laslett et al., 2015; Mayshak et al., 2020). Violence involving alcohol is more severe, is more likely to re-occur and is more likely to result in serious injury (Mayshak et al., 2020). Further, over half (60%) of all episodes occur within two hours of drinking alcohol and the risk of severe IPV is up to 11 times higher on days where alcohol is being used (Fals-Stewart, 2003).

Multiple meta-analyses examining experimental research have shown there is a causal relationship between alcohol, aggression and IPV, but that this relationship is influenced and mediated by individual-level and situation-specific factors (Exum, 2006; Tomlinson et al., 2016). Alcohol may increase risk of IPV through its disruption of normal cognitive processing and problem-solving capacity, so that people have more trouble resolving conflict and are more likely to misinterpret social cues as threats to masculinity and then respond aggressively (Giancola et al., 2005). Alcohol also may cause some people to become aggressive towards their partners because they have a preconceived idea that it is acceptable to behave in that way after they have been drinking (Wilson et al., 2014). If the social or cultural context is tolerant towards alcohol-related aggression, some people may intentionally be violent because they perceive they will be excused of such conduct because they were drunk (Graham et al., 1998).

Compared to people with other mental health disorders, alcohol-dependent people are less likely to be considered mentally ill and more likely to be held fully responsible for their behaviour (Schomerus et al., 2011). However, our research found that in many cases the offender’s alcohol dependence was a clinically significant and sometimes life-threatening condition that had a long-term negative impact on various areas of their lives. For example, in Case 172 the offender had a history of heavy alcohol use since his late teens. His alcoholism interfered with his work, he had difficulty retaining employment due to his drinking and he was known to make violent and aggressive statements when he was intoxicated. He had a history of alcohol-related violent offending and unsuccessful attempts at alcohol rehabilitation treatment. Other offenders reported self-medicating with alcohol. For example in Case 51, after the offender was seriously injured in an assault and his father and brother died, he used alcohol to cope with the pain.

Our findings also support past work showing alcohol is a significant risk factor for violence in Aboriginal and/or Torres Strait Islander peoples’ relationships and communities (Green, 2001). In cases where the offender was Aboriginal and/or Torres Strait Islander the majority (78%, n=40) had a history of AOD (compared to 45%, n=67 of non-Indigenous offenders; information missing for 84 cases). Alcohol and other drug involvement at the time of the homicide was also more common in cases where the offender was Aboriginal and/or Torres Strait Islander (90%, n=46 vs. non-Indigenous=44%, n=65).

In a number of cases involving Aboriginal and/or Torres Strait Islander victims and/or offenders, violence in the relationship was often associated with incidents of prolonged and heavy alcohol use. For example, in Case 48 the relationship between the victim and offender was defined by excessive alcohol use and violence; they reportedly had many arguments that were alcohol-fuelled but did not argue when sober. Conflict and fighting among Aboriginal and/or Torres Strait Islander peoples is influenced by cultural rules and obligations as well as circumstantial, community-level, intra-familial and individual-level factors (Nancarrow, 2019; Shore & Spicer, 2004). It has been suggested that compared to non-Indigenous people, IPV among Aboriginal and/or Torres Strait Islander communities may be more closely related to “fighting” than a pattern of domination and coercive control (Nancarrow, 2019). However, as noted in previous sections of this report, it is important to remain cautious when interpreting these findings as they may more reflect the dominant narrative relied on by judges in sentencing Indigenous offenders than the individualised and nuanced relationship dynamics present in each case.

Illicit drug use

This research found that 29 per cent (n=57) of offenders in the PIPH sample had histories of illicit drug problems (information missing for 56 cases). This was evidenced by the presence of drug charges or prosecutions in the offenders’ histories, participation in drug rehabilitation programs (including methadone programs), and references to histories of drug use that may not have resulted in contact with health systems or the criminal justice system.

Further, the offender had been using illicit drugs at time of the lethal incident in 18 per cent of cases (n=35; information missing for 33 cases). The primary drugs reported to have been used by offenders were cannabis and methamphetamines. While stimulants such as methamphetamines and cocaine have established links to violent behaviour, cannabis tends to be associated only in the context of polydrug use and withdrawal (Tomlinson et al., 2016). Other drugs such as buprenorphine, oxycodone, heroin and inhalants were less commonly reported by IPH offenders in this sample and many who used drugs had histories of polydrug use. For example, in Case 163 the offender began smoking cannabis at 14 and drinking alcohol when he left school in Year 10. As an adult he used methamphetamines regularly and he had been on methadone for 27 years. He had overdosed on heroin multiple times, and tried various drug rehabilitation when he was younger with no success.

As with alcohol, the relationship between illicit drug use and IPV and IPH is complex and may be explained by multiple mechanisms (Gadd et al., 2019). For example, illicit drug use may be an independent risk factor for violence among some people; others may use illicit drugs to cope with the consequences of their violent behaviour; or the relationship may go in both directions (Boles & Miotto, 2003). Alternatively, another factor (such as personality) may predict heightened risk of both IPV and illicit drug use.

Research has indicated there is a higher risk of injury in IPV incidents involving illicit drugs compared to those that don’t involve illicit drugs (Coomber, 2021). People who are dependent users may be more likely to respond aggressively during a state of withdrawal due to increased agitation and anxiety (Gilchrist et al., 2019). In Case 153 the offender had a long-standing illicit drug use dependence from his early teens that increased after his release from prison when he also developed alcohol dependence. The combination of his methamphetamine and steroid use was also related to his severe bouts of anger and violent behaviour.

In some cases analysed, the offender’s illicit drug use was contextualised within an overall deterioration in the health and wellbeing of the offender in the weeks and months leading up to the lethal violence. For example, in Case 154, the offender’s ex-partner outlined that their relationship had been “normal and happy” for the first 18 months until the offender started using ice and became paranoid, agitated and highly irritable. Similarly, in Case 63 the offender was described as performing well at school and leading a happy and satisfactory life before he started using drugs. At the time of the lethal violence he had been using heroin, cannabis and methamphetamines for the previous 10 years. He had a number of drug-related criminal offences, had unsuccessfully attempted to engage with treatment and had taken out a car loan of $10,000 that he used to further fund his drug use.

While some studies show illicit drug use is a better predictor of IPH risk than alcohol use, this is only significant before the effects of previous violence and abuse (including threats) are accounted for (Campbell et al., 2003). Taken together, these findings suggest that perpetrators’ use of illicit drugs and alcohol are important intervention targets for reducing the risk of IPH. However, while illicit drug use is an important situational risk factor, it ultimately holds limited explanatory power (Juodis et al., 2014). Drug use behaviours among IPH perpetrators must be understood and situated in the wider context of their comorbidities, life experiences and patterns of abusive behaviour.

Comorbidity: The co-occurrence of multiple health conditions

As has been discussed throughout this section of the report, mental health problems, substance use and physical health conditions were common experiences of IPH offenders in this study, and many of these men experienced more than one of these conditions. This is in line with previous research showing that people with co-occurring mental health and AOD have an increased risk of criminal offending (López-Castro et al., 2019). Forty-four per cent of IPH offenders in the PIPH sample reported a history of two or more physical and/or mental health conditions, cognitive disabilities or AOD (n=87; information missing for 99 cases; see Figure 13).

Unfortunately, the limitations of this dataset mean it is not possible to provide a true estimate of comorbidity at specific time points as that level of detail was not available. As a result, experiences of historical and current health challenges and diagnoses were grouped together for analysis.

Figure 13: Comorbidity among IPH offenders, by form of comorbidity (%)

Long-term health condition (n=38)

	n
MI	63
AOD	63
CI	45

Mental illness (n=86)

	n
LTHC	28
AOD	64
CI	27

Alcohol and drug use disorder (n=107)

	n
LTHC	22
MI	51
CI	29

Cognitive impairment (n=41)

	n
LTHC	41
AOD	76
MI	56

Note: Percentage totals do not equal 100 due to the categories not being mutually exclusive.
MI=mental illness; AOD=alcohol and other drug disorders; LTHC=long-term health condition; CI=cognitive impairment.

People with co-occurring AOD and other MI are at higher risk of committing homicide than people who do not have these issues (Fazel et al., 2009). Research suggests that the risk of violence and IPV is significantly heightened when particular forms of comorbidity are present, such as co-occurring personality disorders and substance use disorders (Yu et al., 2019). Indeed, the current analysis found that 69 per cent (n=9) of IPH offenders with a personality disorder diagnosis also reported a co-occurring AOD. For example, in Case 125 the offender had been diagnosed with borderline personality disorder, psychotic symptoms and anger management problems which were reportedly exacerbated by his heavy use of alcohol, cannabis and methamphetamines. Further research is needed to understand mechanisms underlying increased risk among people with particular comorbidity combinations beyond specific disorder dysfunction and symptomology (Yu et al., 2019).

As previously discussed and demonstrated in past research (Oliver & Jaffe, 2021), depression and alcohol dependence was a common form of comorbidity among the PIPH sample. Among offenders who had depression, nearly half (46%, n=24) were under the influence of alcohol at the time of the homicide incident. Intersections between depression, hostility and alcohol dependence among men have been observed in previous research (e.g. Cavanagh et al., 2017; Cheng & Jaffe, 2019). The “big build model” may help to explain these interrelationships: some men cope with emotional distress by first turning inwards (avoidance, numbing and escaping) and then turning outwards with externalising behaviours such as aggression and self-harm (Brownhill et al., 2005). The effect of these coping methods is that the source of the stress and distress is not resolved and so continues to grow over time, and the problematic behaviours become more extreme. While details about motivations for use of alcohol and other drugs were rarely provided in the data analysed, self-medication for unresolved trauma, pain management and psychological distress related to injuries and illness, unemployment and stress was relevant.

Among the 51 Aboriginal and/or Torres Strait Islander offenders in the sample, almost half reported more than one AOD and mental or physical health condition (45%, n=23; information missing for 28 cases). This is comparable to the rates reported for non-Indigenous offenders (43%; n=53; information missing for 60 cases). Research shows that Aboriginal and/or Torres Strait Islander offenders present with high rates of PTSD and AOD comorbidity (Heffernan et al., 2015). However, this study found that PTSD diagnoses were more commonly reported for non-Indigenous offenders (n=11) than Aboriginal and/or Torres Strait Islander offenders (n=1). In this sample, more than half (58%, n=7) of IPH offenders with suspected or diagnosed PTSD also reported AOD problems.

In the literature, treatment-seeking samples of people who have co-occurring PTSD and substance use disorders are more likely to perpetrate violence than those without PTSD (Barrett et al., 2013). Research by Barrett and colleagues (2011) showed that over half (55%) of people in treatment with co-occurring PTSD and AOD reported previous violent offending (Barrett et al., 2011). This has been hypothesised to occur because of the interaction between PTSD symptoms of hyperarousal (irritability, angry outbursts, anxiety) and the psychoactive effects of alcohol or other drugs to increase risk for aggressive behaviour (Barrett et al., 2015; Facer-Irwin et al., 2019). People with co-occurring PTSD and AOD often report that their substance use was a form of self-medication to manage symptoms associated with trauma experiences (Barrett et al., 2013; Elbogen et al., 2010; Leeies et al., 2010). Our findings reflect meta-analyses that argue treatment for IPV offenders (such as men’s behaviour change programs; MBCPs) must address men’s comorbid AOD and trauma in order to effectively prevent violence (Karakurt et al., 2019).

Table 11: Emotional, mental and physical health problems among IPH offenders, by pathway (%)

	FT (n=59)	PD (n=73)	DAS (n=19)	Overall (n=199)
Long-term health conditiona	12	16	32	19
Cognitive impairmentb	8	30	6	21
Acquired brain injuryc	7	23	26	16
Any mental illnessd	42	36	68	43
Alcohol and drug use disordere	29	74	53	54
Comorbidityf	25	49	68	44

Note: Overall percentages include the entire sample (n=199).
^a Denominators include cases where this information was missing (FT=37, PD=55, DAS=10, overall=135).
^b Denominators include cases where this information was missing (FT=41, PD=46, DAS=11, overall=132).
^c Denominators include cases where this information was missing (FT=42, PD=50, DAS=10, overall=138).
^d Denominators include cases where this information was missing (FT=26, PD=43, DAS=5, overall=96).
^e Denominators include cases where this information was missing (FT=27, PD=18, DAS=6, overall=107).
^f Denominators include cases where this information was missing (FT=35, PD=35, DAS=5, overall=99).
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

Key theme 2:
Offender experiences of trauma

Trauma is a person’s response to a significant event or ongoing experience which they are unable to process because it is so difficult, frightening, devastating or overwhelming. Depending on the nature of the traumatic event or experience, this can have negative consequences for families, communities and entire cultures. People who have experienced trauma have increased risk of developing chronic and disabling health conditions including MI and AOD, poor physical health and early mortality.

More than half (58%, n=116; information missing for 83 cases) of offenders in the PIPH sample had experienced traumatic life events, including war and conflict, homelessness, incarceration, abuse and neglect during childhood, violent crime and the death of significant family members. This is consistent with other research which has shown that individuals who have contact with the criminal justice system as victims and offenders report higher rates of traumatic experiences when compared to the general community (Angelakis et al., 2020; Driessen et al., 2006; Reavis et al., 2013).

War and conflict

The psychological impact of experiencing war can last long after combat has ceased, with many people suffering symptoms of trauma for years afterwards, or their entire lives. One in 10 IPH offenders in this research reported that they had been exposed to war zones during their lives (11%, n=21; information missing for 157 cases) either as civilians or members of the armed forces. In 13 of these cases (62%; information missing for one case), these experiences occurred during the offender’s childhood and adolescence.

Traumatic experiences associated with war and conflict varied significantly in nature, including abuse perpetrated by others towards the offender during military service, being tortured as a prisoner of war, and witnessing death and injury due to proximity to war zones. IPH offenders’ experiences of war and conflict were not unlike those of many veterans around the world who experience ongoing complex consequences from prolonged exposure to life-threatening situations (Taft et al., 2011). These findings also align with research showing the prevalence of IPV perpetration is three times higher among military populations and veterans compared to the general population (Kwan et al., 2020; Marshall et al., 2005).

Overall, it was more common for migrant IPH offenders to report experiences of war and conflict (25%, n=15; information missing for 39 cases) than it was for IPH offenders who were born in Australia (5%, n=6; information missing for 91 cases). Further, some offenders and their families had experienced animosity, prejudice and social discrimination while growing up in their country of origin. For example, in Case 156 the offender grew up in an Islamic country and his family was discriminated against due to their Baha’i faith. They were consequently restricted in their ability to engage in education and employment and interact easily within society. This was reportedly related to the offender developing significant mental health problems when he was a teenager (depression, self-harm) that continued into his adult life. The role of post-migration experiences in IPH trajectories will be discussed in more depth in later sections of this report.

Incarceration

Experiences of incarceration (time spent serving prison sentences in correctional facilities in Australia and overseas) were described as traumatic for some IPH offenders. One in four offenders (26%, n=52; information missing for 99 cases) had been incarcerated at some point in their lifetime. One in five offenders had been incarcerated for violent offending (21%, n=41; information missing for 96 cases), and one in 10 (10%, n=19; information missing for 92 cases) had been incarcerated for non-violent offending.

Research suggests that offenders are at high risk of exposure to trauma during imprisonment (Piper & Berle, 2019; Sindicich et al., 2014). Not only are inmates likely to be exposed to frequent physical and sexual violent behaviour, verbal abuse and suicide (DeVeaux, 2013), but the emotional and psychological punishment inflicted by design via deprivation of liberty can have enduring effects long after an individual is released. The compounded stress related to the ongoing threat of violence and always needing to be “on guard” may lead people to be in a constant state of hypervigilance, a key symptom of PTSD (American Psychological Association, 2013; DeVeaux, 2013).

Therefore, in some of the cases involving an offender who had previously been incarcerated, the experience was described as having “changed” them. In particular, it was suggested that imprisonment had contributed to the onset or exacerbation of underlying MI and AOD, and impacted their ability to find employment and reintegrate back into society once released. For example, in Case 153 the offender reported that after spending time in prison he was nervous, anxious and started having panic attacks. He was subsequently diagnosed with depression and PTSD that was described as chronic, complex and severe. This is in line with research showing that personality changes and mental health conditions such as PTSD, anxiety, depression and paranoia are common among people exiting prison (Jamieson & Grounds, 2004; Rhodes, 2005).

Further, it was also apparent that prior negative experiences while imprisoned meant that some of the offenders in the sample had a morbid fear associated with re-incarceration. In situations where the offender was convicted of another offence and was returned to the community awaiting the start of the imprisonment period or sentencing, the anticipation of returning to prison appeared to have a significant and negative impact on the offender’s behaviours and personality. In Case 102, the victim and offender had previously spent time imprisoned for drug offences and were facing another sentence of imprisonment for further drug convictions. They both found the initial prison experience extremely difficult and suffered related mental health problems, including PTSD, in the aftermath. At the time of the homicide the couple were reportedly under extreme stress, experiencing a revival of their trauma symptoms and discussing suicide together regularly. Further, in Case 198 the offender was on bail pending investigation of his involvement in the death of his roommate when he met the victim. After a six-month relationship, the offender killed the victim when she threatened to call the police after an incident where he damaged her property. The offender said he killed the victim because if she had called the police, he would have been returned to prison for breaching the conditions of his bail.

By acknowledging that prisons are traumatic environments, we are not suggesting that offenders should not be incarcerated for their crimes. However, the above described cases signal missed opportunities for intervention and prevention. Acknowledging that people facing prison time are likely under significant stress and mental distress means that they should be provided with support to facilitate the transition and the impending significant life change. In particular, when these offenders are returned to the community awaiting sentencing or the start of their incarceration period, the heightened risk that they may pose to their partners and families should be acknowledged and minimised through additional support, safety planning and monitoring.

Childhood experiences of abuse and neglect

In many cases, offenders were reported to have experienced relatively “ordinary” upbringings. In these cases, sentencing judges used words like “normal”, “solid”, “supportive”, “uneventful”, “happy” and “loving” to describe their family and background (e.g. Cases 168, 160, 151, 149, 63 and 50). However, in other cases, offenders were described as having childhoods characterised by abuse and neglect. Consistent with international estimates, overall 32 per cent of offenders had experienced some form of childhood abuse and neglect during their childhood and adolescence (n=63; information missing for 136 cases; Krivisto, 2015; Stout, 1993). This included experiences of sexual victimisation, family violence, witnessing IPV between carers, carer arrangements changing, and extreme socio-economic deprivation (including homelessness). In situations where the offender was removed from his biological parents as a child or adolescent and placed in alternative care arrangements, the assumption was made that this was due to experiences of abuse and neglect.

Family violence and sexual violence victimisation experiences

Overall, one in eight offenders in the PIPH sample reported being a target of family violence growing up (13%, n=25; information missing for 164 cases). In most of these cases, their biological father was usually identified as perpetrating the abuse, although in some situations their mother was also identified as being abusive towards them (e.g. Cases 26 and 193), as well as other family members such as uncles and aunts (e.g. Case 21). Some abuse reported by offenders was very severe in nature, such as being punched to the point of unconsciousness. Abuse was often situated by offenders in the context of “discipline”, such as in Case 35, where the offender reported receiving a “bit of a hiding” on occasions as a child.

The abuse experienced by offenders was not only physical in nature; many offenders had reportedly experienced emotional abuse growing up from their parents and caregivers. For example, in Case 45 the offender described being physically assaulted, emotionally abused and socially humiliated by his father on a regular basis. Further, in Case 29 the offender’s mother and siblings described his father as being unemotional, cold and dismissive of the offender during his childhood.

One in eight offenders witnessed intimate partner and family violence between family members during their childhood and adolescence (14%, n=27; information missing for 163 cases). This is higher than international estimates of three per cent (Valabdass et al., 2021). The abuse witnessed most often took the form of the offender’s father being violent towards their mother or their siblings (particularly female siblings), and often co-occurred with carer alcohol use and mental health problems.

Sexual abuse victimisation among offenders was not frequently identified, only being reported in a small number of cases overall. For example, the offender in Case 125 said he had been raped by his two brothers when he was a child. Another offender in Case 41 reported being sexually abused by an adult stranger when he was four. As community-based sexual victimisation estimates suggest that between one in six and one in 10 people experience sexual assault and sexual abuse in childhood, the low estimates from this study may point to underreporting (Barrett et al., 2015; Stout, 1993).

Removal from home and changes in carer arrangements

Young people who are removed from their homes and placed in out-of-home care are likely to have experienced complex and multiple traumas (Department of Families, Housing, Community Services and Indigenous Affairs, 2011). When young people need to be removed from their home, they may be placed in residential, foster or kinship care until they can return home again. In 15 per cent of cases, the offender’s carer arrangement changed during their childhood and adolescence (n=29; information missing for 139 cases). In almost half of these cases (41%, n=12), the carer change involved the offender being removed from the home and entering non-familial placements (e.g. refuges, foster homes, orphanages and group homes).

Crucially, 27 per cent of Aboriginal and/or Torres Strait Islander offenders experienced a change in carer arrangements during their childhood and adolescence (n=14; information missing for 36 cases) compared to 10 per cent of non-Indigenous perpetrators (n=12; information missing for 90 cases). Although detailed information about the context or reasons for carer arrangements was limited in many cases, there is consistent research suggesting that the high proportion of child removals and child protection involvement in First Nations communities may be attributable in part to systemic racism and bias, as well as the intergenerational impacts of the Stolen Generation and associated child removal policies (Cuneen & Libesman, 2000; Funston & Herring, 2016; Newton, 2020).

Children who experience trauma may recover and lead healthy, fulfilled lives if they experience positive and nurturing care environments (Esaki et al., 2013). Conversely, the experience of being removed from home can be a detrimental and traumatic experience in its own right, compounding the adversity that led them there in the first place. For example, in Case 125 the offender was removed from his home when he was three and returned at the age of seven. While he was in out-of-home care, his parents had divorced and his mother had remarried a member of a motorcycle gang. His home life was characterised by dysfunction and heavy alcohol use, and he eventually dropped out of school in Year 9. This example shows how disruptions to care arrangements and home environments can negatively impact a young person’s development and contribute to their pathway to IPH.

Links between childhood abuse and neglect and IPV and IPH

Evidence shows that the number of adverse childhood experiences experienced by an individual is independently associated with violent behaviour even after controlling for other risk factors such as personality and heavy alcohol use (Lawler et al., 2021). The links between child abuse and neglect experiences and IPV and IPH have been explained in the literature in a few key ways (Roberts et al., 2011). First, it has been suggested that adverse childhood experiences can have a negative impact on an individual’s attachment to their primary carer figures (i.e. their parents), which then influences their subsequent attachment to intimate partners and others. In particular, individuals with poor or dysfunctional attachment to their primary carers during childhood may be cold and withholding within their relationships with intimate partners, or overly attached to them and anxious about being abandoned. In both of these scenarios, offenders may use violence as a means of managing their emotional distress related to their relationships with intimate partners (Babcock et al., 2000; Doumas et al., 2008; Henderson et al., 2005).

Alternatively, social learning theory suggests that men who witness IPV between their carers begin to view these behaviours as appropriate and acceptable means of managing conflict within relationships. Within this framework, use of violence against their intimate partners is “learnt” by watching the interactions between their parents (Aldarondo & Sugarman, 1996; Daniels & Murphy, 1997; Giordano et al., 2015; Nowakowski-Sims, 2019).

More generally, negative and adverse childhood experiences can have negative consequences for individuals’ psycho-social development, and are linked to the onset and exacerbation of mental health issues like PTSD and AOD, as well as criminal offending more generally and low levels of impulse control (Baidawi & Sheehan, 2019; Edwards, 2018). As such, the relationship between childhood abuse and neglect and IPV and IPH may be direct, or mediated by other individual and social factors (see for example Gay et al., 2013).

Death of family members

The deaths of family members, either suddenly or as a result of long-term health problems, had a significant impact on the lives of many IPH offenders. Deaths of primary caregivers, such as biological parents or grandparents, and family members (e.g. siblings) were reported in 16 per cent of cases (n=31; information missing for 165 cases). In over half of these cases, this experience occurred during offenders’ childhood or adolescence (61%, n=19; information missing for six cases). A number of offenders lost loved ones in the context of accidents, suicide, natural disasters and drug overdoses. For example, in Case 98 the offender’s two-year-old brother was killed in a flood when the offender was six.

The death of family members was more common in cases involving an Aboriginal and/or Torres Strait Islander offender. One in four Aboriginal and/or Torres Strait Islander offenders experienced the loss of a family member (28%, n=14; information missing for 37 cases), compared to one in 15 non-Indigenous offenders (7%, n=10; information missing for 112 cases). Further, in many cases involving an Aboriginal and/or Torres Strait Islander offender it was reported that they had experienced the loss of multiple family members. For example, in Case 150, the offender’s father and sister both died from drug overdoses and his brother had passed away after a car accident.

The length of time between the death of a family member and the lethal incident varied considerably across the dataset. In some cases, the death occurred in the months leading up to the lethal incident, and the event appeared to be associated with the onset or exacerbation of acute emotional distress for the perpetrator. In other cases, the death of a family member occurred earlier in the perpetrator’s life, often years prior to the lethal incident.

Despite this lack of temporal proximity to the lethal incident, the death of family members appeared to have a long-term and lasting impact on the perpetrator, affecting their emotional development and relationships with others. In particular, the grief associated with the loss of close relationships with kinship carers such as grandparents who died at significant points in their development (i.e. early adolescence) was described as having a significant impact on offenders. The impact of his grandmother’s death on one offender’s pathway was described in Case 38. The offender went to live with his grandmother when he was nine months old and he was very close to her growing up. She died when he was 13 and after this he did not have a safe or stable place to live. He was then exposed to significant violence and alcohol use in the community; his behaviour deteriorated and he started offending and using inhalants, alcohol and cannabis.

The reported impact of losing a loved one during their childhood and adolescence on offenders is consistent with research which has shown that the unexpected death of a loved one is often rated by individuals as the worst of all traumas, regardless of other types experienced (Keyes et al., 2014). Experiencing the death of a loved one, particularly when this is sudden or unexpected, can result in significant psychological distress and PTSD-related symptoms. If left unmanaged, unresolved grief can lead to complex and ongoing problems for sufferers and their families. For example, in Case 51 it was reported that the offender’s parents provided him with a strong moral grounding which guided his behaviour and identity formation. When his father and brother both died around the same time, he used alcohol to cope with his grief which resulted in the development of an alcohol use disorder. It is important to recognise the role of grief and loss associated with trauma in order to provide support to people in their recovery and rehabilitation. The period after people experience such an event is a critical time for the development or onset of various mental health and substance use problems (Keyes et al., 2014).

Trauma experienced by Aboriginal and/or Torres Strait Islander offenders: Intergenerational trauma and systemic racism and discrimination

Intergenerational trauma is experienced when an original trauma goes unresolved for the person or peoples who experienced or witnessed it and the emotional and psychological wounds are then passed down through the next generation surrounding that person or community (Menzies, 2010). The cumulative impact of this is also known as historical response trauma and it is a collective experience of diverse cultures and Indigenous peoples around the world (Salzman & Halloran, 2004). Trauma that becomes shared in this way can be embedded within how people associate with each other socially and their practices as well as institutional and systemic processes (Kirmayer et al., 2000). Functional and relational challenges often follow trauma where people experience subsequent dysregulation that impacts on their capacity to effectively manage stress.

Racism and discrimination are established causal factors for poor health in the international literature (Paradies, 2006). In Australia, research shows that one in three Aboriginal and/or Torres Strait Islander peoples reported unfair treatment based on their identity (ABS, 2016) and one in six reported experiencing racism in the past year (Markwick et al., 2019). Judges partly alluded to the cumulative impact of systemic racism over time when they assessed the relationship between community-level, socio-economic factors and IPH offending. Many Aboriginal and/or Torres Strait Islander offenders reflected positively on their communities of origin in regard to learning about culture and custom, developing and fostering a strong sense of cultural identity and mentoring young people. While diverse developmental experiences were reported by Aboriginal and/or Torres Strait Islander offenders, the role of recent and historical trauma was discussed by judges in more detail in cases where the offender was Aboriginal and/or Torres Strait Islander. This manifested in bias related to the additional description of lifestyle factors and community-level disadvantage for these cases.

Unfortunately, based on the nature of the information analysed here we cannot comment on the relationship between trauma stemming from systemic racism and IPH offending. Specific experiences of systemic racism as related to trauma were not addressed on the individual level but were assessed for Aboriginal and/or Torres Strait Islander peoples by way of their Indigeneity. For example, in Case 35 it was reported that the offender’s substance use and its role in his offending was reflective of the socio-economic environment where he grew up, a remote Aboriginal community where alcohol use and violence were prevalent. The judge reported that his personal history, deprivation and difficulties associated with his Aboriginality served to mitigate the sentence but this was also weighed against the seriousness of the offence.

Experiences of deprivation, poverty, trauma and abuse influence a person’s moral culpability. Sentencing courts take these factors into account by considering the evidence or information provided by the defence as part of the pre-sentence process. The ongoing experiences of trauma among Aboriginal and/or Torres Strait Islander peoples are multifaceted and varied, related to disadvantage, colonisation, dispossession and systemic racism. Anger among Aboriginal and/or Torres Strait Islander prisoners is related to the context of historical trauma, emotional dysregulation and experiences of discrimination (Day et al., 2008). Aboriginal and/or Torres Strait Islander peoples in Australia experience disproportionately poor health and social inequality compared to the general population presenting through disproportionate rates of victimisation and criminal justice involvement, harmful substance use, economic hardship, increased morbidity and early mortality (AIHW, 2005). The wider literature rarely distinguishes between trauma from colonisation specifically and its secondary impacts such as poverty, criminal justice engagement, alcohol dependence and violence (Nelson & Wilson, 2017). As the above research indicates, this is a complex and multifaceted problem which requires future study.

Table 12: Traumatic experiences among IPH offenders, by pathway (%)

	FT (n=59)	PD (n=73)	DAS (n=19)	Overall (n=199)
Any incarcerationa	17	36	21	26
War and conflictb	7	7	16	11
Child abuse and neglectc	27	41	16	32
Family violence victimisation (includes sexual violence) d	12	16	0	13
Non-family violence-related violence victimisation (includes sexual violence) e	3	5	0	4
Carer arrangement changesf	8	22	5	15
Witnessing family violence between carersg	7	23	0	14
Death of a family memberh	4	17	0	16
Any traumai	49	70	42	58

Note: Overall percentages include the entire sample (n=199).
^a Denominators include cases where this information was missing (FT=27, PD=42, DAS=8, overall=99).
^b Denominators include cases where this information was missing (FT=48, PD=61, DAS=15, overall=157).
^c Denominators include cases where this information was missing (FT=43, PD=43, DAS=16, overall=136).
^d Denominators include cases where this information was missing (FT=48, PD=58, DAS=17, overall=164).
^e Denominators include cases where this information was missing (FT=52, PD=67, DAS=18, overall=182).
^f Denominators include cases where this information was missing (FT=44, PD=47, DAS=15, overall=139).
^g Denominators include cases where this information was missing (FT=51, PD=54, DAS=17, overall=163).
^h Denominators include cases where this information was missing (FT=53, PD=56, DAS=18, overall=165).
ⁱ Denominators include cases where this information was missing (FT=30, PD=22, DAS=11, overall=83).
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

Key theme 3: Separation

Of the 199 cases included in the sample, one third involved a victim and offender who were separated (32%, n=63) or negotiating reconciliation (1%, n=1) at the time of the lethal violence (information missing for 36 cases). For the remainder of this analysis, the case in which the victim and offender were negotiating reconciliation will be included in the separation total (n=64).

The proportion of cases that involved a separated victim and offender is consistent with previous research, which suggests that between 20 and 57 per cent of IPH incidents involve a victim and offender who were no longer in a relationship (Campbell et al., 2007; Kivisto, 2015; Wilson & Daly, 1993). This rate may be higher when prior separations and reconciliations are considered (Campbell et al., 2003; Campbell et al., 2007). Indeed, a further nine per cent (n=17) of cases in this study involved a victim and offender who had been separated previously and reconciled prior to the lethal incident. What this means is that overall, 41 per cent of cases included in the final sample involved a victim and offender who had separated on at least one occasion during their relationship.

In 38 per cent of cases involving a separated victim and offender, the lethal violence occurred within three months of separation (n=24; information missing for 19 cases), and one in two (56%, n=36) occurred within the first year of separation (see Figure 14). This is consistent with previous research (Aldridge & Browne, 2003; Campbell et al., 2007; Dobash et al., 2007; Spencer & Stith, 2018; Wilson & Daly, 1993).

Figure 14: Period of time between separation and lethal incident (limited to cases where the victim and offender were separated or negotiating reconciliation at time of the lethal incident; n=64; %)

Period of time	Cases (n)	Cases (%)
3 months or less	24	38
4 to 6 months	6	9
7 to 12 months	6	9
13 to 18 months	2	3
19 months or longer	7	11
Not stated/unknown	19	30

Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

Escalating patterns of violence and abuse post-separation

IPV perpetrated by the offender against the victim was common during the period following a separation. Overall, almost two thirds (63%, n=40) of offenders perpetrated violence during this period (information missing for 36 cases). Of these cases, 68 per cent (n=27) involved the escalation of previous patterns of IPV. The most common type of violence perpetrated by offenders who were separated from the victim was emotional abuse (44%, n=28; information missing for 30 cases), followed by threats to harm the victim (31%, n=20; information missing for 38 cases; see Table 13). Further, five per cent of victims (n=3) perpetrated violence during this period (information missing for 46 cases). Of these, one involved the onset of IPV, and the remaining two involved the escalation of previous violence.

Table 13: Intimate partner violence perpetration during the period following separation (limited to cases where the victim and offender were separated or negotiating reconciliation at time of the lethal incident; n=64)

	n	%
Emotional abusea	28	44
Threats to harm the victim (including threats with a weapon and threats to kill)b	20	31
Physical violencec	19	30
Stalkingd	16	25
Coercive controle	14	22
Threats to harm othersf	9	14
Violence and abuse towards children (including threats)g	9	14
Non-fatal strangulationh	6	9
Financial abusei	4	6
Sexual violencej	2	3
Any IPV	40	63

^a Denominator includes 30 cases where this information is missing.
^b Denominator includes 38 cases where this information is missing.
^c Denominator includes 28 cases where this information is missing.
^d Denominator includes 41 cases where this information is missing.
^e Denominator includes 42 cases where this information is missing.
^f Denominator includes 46 cases where this information is missing.
^g Denominator includes 45 cases where this information is missing.
^h Denominator includes 43 cases where this information is missing.
ⁱ Denominator includes 50 cases where this information is missing.
^j Denominator includes 52 cases where this information is missing.
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

One in four (25%, n=16) victims obtained a protection order during the post-separation period, typically in response to the onset or escalation of IPV during this time. Protection orders may also have been sought in response to violence perpetrated just prior to separation (often causing or contributing to the separation). As such, offenders may have been known to police during this time.

Conflict about care arrangements for shared children

As noted earlier, in approximately one in two cases (48%, n=31) where the victim and offender were separated at the time of the lethal violence, they had at least one child together. In 42 per cent of these cases (n=13; information missing for 18 cases), the victim and offender were described as being in conflict about issues related to the care of shared children. These included:

• disputes relating to the custody of shared children (n=7)
• the victim or offender limiting access to shared children (n=8)
• handover issues (n=7)
• disputes relating to the financial support of children (n=2).

The victim was the primary carer for shared children in 85 per cent (n=11) of these cases, and the offender was the primary carer in one case (information missing for one case). The care arrangements for shared children were typically agreed between the victim and offender by way of an informal arrangement, with only three cases involving formal parenting orders. In two of these three cases, the victim had recently applied to move within the state (Case 119) or overseas (Case 129), which would have potentially limited the offender’s contact with their children in future. In two further cases, the victim was in the process of seeking formal custody of the children (Cases 162 and 36), and in Case 200, the offender was anticipating that the victim would apply for full custody of the children.

In 23 percent of cases where the victim and offender had shared children, the victim had experienced issues during handover (n=7; information missing for 21 cases). For example, as described in Case study 6 below, in one case the offender refused to hand the children back to the victim, resulting in the victim seeking a recovery order.

In many cases where the victim and offender were involved in dispute regarding the care of their children, the lethal violence was in part motivated by the offender’s extreme anger at the victim for their perceived role in “withholding” or limiting their access to the children. This anger may have been exacerbated by the victim applying for a protection order against the offender due to his ongoing and persistent abuse of her and their children. These findings are consistent with international literature which has found that changes in child custody arrangements are a potentially triggering event that ultimately leads to the offender’s loss of control and perpetration of lethal violence (as evidenced by Case study 6, and reported by Ellis, 2017; Harden et al., 2019; Sheehan et al., 2015). Ultimately, separation and associated issues may act as both a risk factor and motivation for IPH (Harden et al., 2019).

Re-partnering and concerns about “love rivals”

During the post-separation period, the victim had re-partnered in 28 per cent of cases (n=18; information missing for 30 cases). In six of these cases, the offender suspected (often erroneously) that the victim had been unfaithful to him, even if her new relationship started after she had separated from the offender.

Regardless of whether they believed that the victim had been unfaithful or not, in one in two cases where the victim had re-partnered, the offender was described as being distressed, jealous and angry as a result (56%, n=10). For example, in Case 69, the offender received a photo of the victim and her new partner. This resulted in him texting the victim to say that he would kill both her and her new partner.

In some situations, the distress and anger experienced by the offender was heightened when he was confronted with evidence that the victim’s new relationship had become more “serious”, and it was therefore unlikely she would reconcile with him. For example, in Case 106, the offender killed the victim shortly after she told him that she had become engaged to her new partner.

Critically, the offender had begun a new relationship in seven cases. There was no evidence that the offender re-partnering had contributed to conflict between the victim and offender in any of these cases. However, in Case 62, the offender told the women he dated after separating from the victim that he was still in love with the victim, and wanted to get back together.

Financial disputes

Of the 64 cases in which the victim and offender were separated at the time of the lethal violence, 14 were known to have involved financial disputes (22%; information missing for 44 cases). These disputes predominately related to:

• the offender’s concern that the victim had received or would receive a greater share of shared assets through legal proceedings (n=4). For example, after killing the victim, the offender in Case 151 told police that it was a shame that the victim couldn’t accept “a fair deal” in relation to financial settlement and that if she had, the conflict could have been solved in a different manner
• what would be done with the family home (n=4). For example, in Case 136 the victim wanted to sell the house and divide the money equally, but the offender did not want to sell the house
• the offender withdrawing financial support from the victim, or withholding her access to shared assets (n=4). For example, in Case 187 the victim attempted to retrieve her belongings from the offender but he would not cooperate.

Threats of separation

There were also a number of cases where the victim and offender were in the process of separating at the time of the lethal violence. In these situations, the victim and offender were continuing to live together, but were described as sleeping separately (e.g. Cases 98, 86 and 32). Further, victims may have raised the possibility of separation prior to the lethal violence but been met with a hostile response from the offender. Such responses identified through the analysis included:

• the offender refusing to accept the victim’s decision to end their relationship (for example, in Case 145 the offender refused to move out of the family home)
• the offender assaulting the victim (for example, in Case 194 the offender attempted to choke the victim when she suggested that they separate)
• the offender threatening the victim (for example, in Cases 148 and 90 the offender told the victim that he would kill her before he divorced her).

Key theme 4:
Hegemonic masculinities and traditional gender norms

Consistent with other IPH research (see for example Polk, 1994), regardless of the pathway, there was evidence that in many cases the offender may have been motivated to kill (or at least physically harm) their partner because of the victim’s perceived violation of gendered norms associated with femininity, and in turn the challenge to the offender’s masculinity. In particular, in many cases, a key stage in the IPH trajectory involved the offender perceiving that the victim was challenging their authority and dominance within the relationship in some way. Certainly, as noted in the description of the FT pathway and the above section of this report focusing on separation, the violence and abuse experienced by some women escalated considerably in the period following the relationship ending and/or as a result of other actions taken by the victim to limit the level of control the offender had over them. This was demonstrated in Case 45: in the post-separation period, the victim asked a friend to collect some property from the offender because she did not want to have contact with him. The offender became very angry about this because he believed she was being disrespectful by withdrawing contact, and was “ungrateful” for all that he had done for her and their children. He then decided to “punish” her by killing her.

Other forms of “challenge” that emerged from the analysis of the entire sample included:

• the victim returning to work or succeeding in their career
• the victim refusing to acquiesce to the offender’s demands
• the victim “fighting” back against the offender during an incident of violence
• the victim having an affair
• the victim re-partnering (see section on separation).

Crucially, victims were entirely within their rights to engage in these behaviours, and by consciously or unconsciously challenging the offender’s control over them they in no way should be viewed as “provoking” or excusing the offender for their subsequent violence. However, that these acts appeared to contribute to the IPH trajectory of many offenders demonstrates the extent to which these men felt they were entitled to have control over their victims, and the level of anger they experienced when their masculinity and role within the relationship was challenged. This information can assist in understanding the underlying causes for the lethal violence occurring.

Research has shown that men who condone violence generally, and endorse traditional gender norms, are more likely to be abusive towards their intimate partners (Golden et al., 2013). Traditional gender norms place socially defined constraints around the expected behaviours of individuals, defined by their identified gender. For example, in many societies, traditional gender norms limit women’s influence to the private domain, particularly through domestic duties, submitting to their partner and raising children. Meanwhile, men are expected to exert their influence in public spaces, through employment, political engagement and in the broader community and, crucially, be the dominant partner and have control within their relationships (Lamont, 2014; McCarthy et al., 2018).

Although this information was missing for the majority of cases, there were some instances where the offenders in the current sample voiced opinions or behaved in ways which demonstrated that they endorsed traditional gender norms (17%, n=33; information missing for 165 cases). Often, these views were expressed in very subtle ways. For example, in Case 179 the offender was raised in a household where he had a privileged position relative to his female siblings. In particular, his mother was very protective of him, and had even lied to the police on his behalf when they were investigating his involvement in criminal offending. When the offender would stay with his parents he said that his mother waited on him “hand and foot”, and that it was like living in a hotel.

In other cases, the views of offenders regarding gender norms were more overt in nature. For example, in Case 102, during his police interview after killing the victim, the offender said that women were “wonderful creatures” but also said that sometimes, it was the responsibility of men to take charge and solve “problems”. The implication here was that the victim’s perceived disobedience was a problem that the offender had to “solve” with lethal violence. Further, in Case 179 the offender was raised in a household where his father made all the decisions regarding the family business, and his mother ran the household and raised the children. When he took over the business and his own wife wanted to be involved in its day-to-day operations, he responded with aggression and hostility. This was attributed to his belief that as a man, he did not need his wife’s assistance to run the business and that she should not try to be involved.

Adherence to and endorsement of traditional gender roles by some men in the PIPH sample sits within and is shaped by broader community attitudes in Australia. It is crucial to recognise that Australia is still a patriarchal society that prioritises the interests of men over women in various domains. For example, in the most recent National Community Attitudes towards Violence against Women Survey, a notable proportion of respondents endorsed beliefs consistent with traditional gender norms (Webster et al., 2018). For example, 25 per cent of respondents agreed that women prefer a man to be in charge of the relationship, 16 per cent agreed that men should take control in relationships and be the head of the household, and 14 per cent agreed that on the whole, men make better political leaders than women (Webster et al., 2018).

Further, although it has declined over the last 20 years, there is still a significant wage gap between men and women in Australia. The wage gap was calculated by the Australian Government Workplace Gender Equality Agency as 14.2 per cent in May 2021, meaning that Australian women earned on average A$261.50 per week less than men (Workplace Gender Equality Agency, 2021). Further, according to research conducted by the Australian Institute of Family Studies, in Australian households, women are much more likely to be the primary carer of children, and take on the majority of household-related activities (Hand et al., 2020).

However, the vast majority of men in Australia will never perpetrate violence against their partners, and only a tiny minority will kill them. As such, it is important to look at individual-specific risk factors that may illustrate why these men killed their partners. Such approaches are consistent with public health models which recognise that individualistic behaviours are likely influenced by societal-, relationship- and individual-level factors (Heise, 2013). One such risk factor is the internalisation of gender norms and exposure and adherence to hegemonic masculinities.

First, as described in previous sections of this report, approximately one in seven offenders were exposed to IPV perpetrated by their male carers towards female members of their family, particularly their mothers (14%, n=27; information missing for 163 cases). Although many men who witness IPV in their families of origin do not go on to become perpetrators of violence, research has again shown that these experiences are strong predictors of subsequent perpetration of IPV (Mahalik et al., 2005; McDermott & Lopez, 2013; Reidy et al., 2014).

In explaining this link, researchers have shown that men who are abused or witness IPV as children are more likely to have strong normative beliefs about the role of men within relationships, and endorse traditional values associated with masculinity (Eriksson & Mazerolle, 2015; Mahalik et al., 2005; Reidy et al., 2014). In these situations, female partners behaving in ways that they believe challenge their authority within the relationship could be viewed as an assault on their masculinity more generally. Gender role strain theory suggests that men who believe their masculinity is being challenged often experience emotional distress as a result, and use violence as a means of both reasserting their masculinity and mitigating this emotional distress (Mahalik et al., 2005; McDermott & Lopez, 2013; Reidy et al., 2014). Certainly, the analysis showed that a larger proportion of IPH offenders who had witnessed IPV or been a target of family violence voiced opinions or behaved in ways which demonstrated that they endorsed traditional gender norms (21% vs. 11%). However, this information was not available for the majority of cases (n=190).

Second, one in four offenders included in the sample were described as being exposed to communities and cultural groups characterised by hegemonic masculinities at different stages of their life (25%, n=50; information missing for 149 cases). Hegemonic masculine communities that offenders were affiliated with included:

• sport teams and groups (e.g. Cases 193 and 36)
• body-building cultures (e.g. Case 44)
• male-dominated work industries (e.g. in Cases 200 and 189 the offenders worked in the construction industry)
• religious and cultural groups that endorsed traditional gender norms (e.g. in Case 45 the offender was a member of an outlaw motorcycle gang)
• military service (e.g. Cases 156, 145 and 37).

It is important to acknowledge that being a part of these groups or organisations does not mean that offenders took on the attitudes of work colleagues and peers around them that endorsed traditional gender norms – nor even that these attitudes existed among members in the first place. However, there is evidence that endorsement of traditional gender norms is concentrated among specific social groups, particularly those associated with traditional notions of masculinity, and that rates of IPV may be higher among these groups (see for example Eriksson & Mazerolle, 2015; McCarthy et al., 2018; Nydegger et al., 2017; Santana et al., 2006). Further, research has demonstrated the important role of peers and workplaces for informing individuals’ attitudes and belief systems, including towards gender equity and gender norms (Copp et al., 2019). As such, it is possible that the attitudes of offenders within the current sample may have been influenced by their peers and work colleagues, and in turn contributed to their IPH trajectory.

Third, as noted in the next section of this report which focuses on the role of pre- and post-migration experiences on IPH pathways, a significant proportion of offenders included in the final PIPH sample were from countries and cultures where hegemonic masculinities and endorsement of gender norms were strong, reinforced in some situations by religious beliefs. In these cases, the process of migration and changing power dynamics within the relationship may have been viewed as culturally inappropriate by the offenders, and in part motivated their use of lethal violence against the victim.

This discussion highlights the potential importance of traditional gender norms and hegemonic masculinities in understanding the occurrence of IPH in many cases. However, the ability to undertake a more nuanced analysis was limited by the absence of information about the beliefs of offenders regarding gender equality and gender norms, and their affiliation with communities and groups characterised by adherence to hegemonic masculinities. Future research should continue to unpack these relationships, as these groups may provide highly effective sites for intervention

Key theme 5:
Pre- and post-migration experiences

As shown in Table 14, of the 199 cases included in the PIPH sample, 31 per cent (n=61) of offenders were immigrants born overseas in one of 34 unique countries. Of those 61 offenders, 47 were born in locations where English is not the primary language. The remainder of overseas-born offenders (n=13) were born in countries where English is considered the primary language (e.g. the United Kingdom; information missing for one case).

Relevant to this study, it is important to differentiate between migrants and refugees. While the United Nations (2021) acknowledges there is no formal legal definition, most experts agree that an international migrant is someone who changes his or her country of usual residence by choice. Refugees are persons who are forced to flee their country of origin for reasons of feared persecution, conflict, generalised violence, or other circumstances (United Nations, 2021). For the general discussion in this report, the terms “migrants” and “immigrants” will collectively incorporate those individuals who entered Australia as refugees or otherwise. However, where relevant, an individual’s or family’s refugee status will be described.

For the purposes of this section, we focus specifically on cases where the offender was from a culturally and linguistically diverse (CALD) community (n=47); that is, offenders who were born overseas in predominately non-English speaking countries. Offenders born in countries where English is the primary language (Canada, New Zealand, the United Kingdom) generally identify strongly with Anglo-Australian culture and are likely to experience migration very differently to those individuals and families from CALD backgrounds.

Consistent with the international research, the current analysis found that pre- and post-migration experiences of migrant offenders had an important role in their pathway to IPH (Balica & Stöckl, 2016; David & Jaffe, 2021; Guruge et al., 2020). The following section describes these experiences, including the broader processes of acculturation, in IPH pathways.

Table 14: CALD status of offenders included in the PIPH sample, by pathway (%)

Pathway	CALD offenders	English PL overseas
FT (n=26)a	77	19
PD (n=8)	63	38
DAS (n=7)	71	29
Overall (n=61)a	77	21

Note: Percentage totals may not equal 100 due to rounding.
^a Denominator includes one case where this information is not available.
Source: “Pathways to intimate partner homicide” project 2020–21 [Computer file]

Pre-migration trauma

The process of migration, regardless of reason, not only involves leaving social support networks but can also lead to experiencing a sense of loss, dislocation, alienation and isolation, which will in turn influence processes of acculturation in destination countries (Bhugra, 2004). As discussed in earlier sections of this report, some offenders included in the PIPH sample were forcibly displaced from their birth country by war, political conflict, or risks of ethnic, racial or religious persecution. This was the case for 30 per cent (n=14; information missing for 29 cases) of CALD offenders in this study.

Many refugees who flee due to conflict or to escape persecution have been subjected to and/or witnessed violence (David & Jaffe, 2021). They have often suffered physical and sexual abuse and face separation from and loss of family members. These individuals not only leave their country but also their established lives, often losing employment, property and social status. As a result, many experience hardships in transit countries as well as destination countries. This was demonstrated in Case 116 which involved an offender who had lived through civil war, and had witnessed extreme violence at a young age. General education was limited in his transit country, stunting his educational growth and offering little opportunity to learn English. On arrival in Australia, these limitations hindered the offender’s prospects to support his wife, and he began to abuse her shortly after the birth of their son. The added emotional, psychological and financial stresses of parenthood were beyond the offender’s coping ability.

Aside from war, conflict and persecution, other traumatic experiences reported for migrant offenders included death of a family member (17%, n=8; information missing for 37 cases), extreme levels of financial stress (36%, n=17; information missing for 26 cases; see Case study 8) and incarceration (21%, n=10; information missing for 18 cases). The roles of these types of traumatic experiences are covered in previous sections of this report.

Pre-migration trauma for offenders can have long-reaching impacts, and also influence the extent to which they are able to settle in their new country (Guruge et al., 2020; see Case study 7). Bhugra (2004) notes that an individual’s ability to cope with stressors associated with the immigration process is influenced by their reasons for migration, preparation prior to the journey and social support before, during and after.

Post-migration stressors

On arrival in a new country, individuals and families from CALD backgrounds can face a number of complex and additional stressors. These include chronic unemployment, social isolation, changes in the power dynamics within relationships, and uncertainty and instability regarding their residential status (David & Jaffe, 2021; Guruge et al., 2020). However, it is also important to recognise more broadly that the process of acculturation can be a source of emotional distress and stress for migrants, even in the absence of specific and acute barriers and problems. Being accepted in their new nation is also a significant contributor to the genesis of stress and how the individual deals with such stress.

Acculturation is the ongoing and fluid process in which individuals from CALD groups must balance their conflicting needs for cultural preservation and cultural adaptation (Sawrikar & Katz, 2008). Research suggests that the process of acculturation has the potential to positively or negatively affect behaviour, largely due to the changes and stressors that underlie this process (Gupta et al., 2010). The challenge of acculturation can be intergenerational within CALD families. This is demonstrated in Case 135, where the offender was a first-generation, Australian-born member of a CALD community. He was subjected to harsh discipline by his very religious father whom the Court deemed a “dominant influence”. The offender met and married the victim in his parents’ country of birth, temporarily living in the region, but was eventually kicked out of the victim’s family home due to personality conflict. The offender believed that he was caught between two cultures. Upon returning to Australia, the offender’s father offered unwelcomed intervention into his relationship, causing anguish and stress, which the offender took out on the victim. The abuse escalated to the point where the victim wanted to leave the relationship. This led to the offender killing the victim.

There are strong links between culture and personality. The acculturation process may be experienced by an individual as pressure to change some aspects of their identity and concept of self. These cultural adaptation processes can be emotionally distressing, increasing risk of long-term mental health problems (Bhugra, 2004). Immigrants may also experience emotional distress and symptoms associated with mental health issues due to post-migration racism, discrimination and social bias, which can be re-traumatising and/or increase risk of vulnerability for migrants (Baobaid et al., 2018). Certainly, nearly half of all CALD offenders in this study suffered from mental health issues (49%, n=23; information missing for 16 cases), and just over one in four had an AOD (26%, n=12; information missing for 22 cases). Three offenders had confirmed diagnoses of PTSD while this diagnosis was suspected for several others. This is not to suggest that these individuals’ mental illness was a direct cause of the lethal incident – as noted in previous sections of this report the vast majority of persons with a mental illness are not violent. Rather, it is demonstrative of the fact that some offenders in the PIPH sample who had migrated to Australia during their lifetimes may have experienced pre- and post-migration stressors that contributed in some way to the lethal incident.

Collectivism versus individualism: Barriers to help-seeking and shame

The relative collectivism or individualism of a particular cultural or ethnic group can play a significant role in personal and family experiences of IPV, and attitudes more broadly about accessing community services that address family violence. Australia generally takes an individualistic view in its family service provision, where individuals are only expected to look after themselves and their immediate family. CALD communities are generally collectivistic. Collectivism refers to a society that embraces a “we” consciousness: collective identity, emotional interdependence, group solidarity, sharing, duties and obligations, the need for stable and predetermined friendships, group decisions and particularism (Bhugra, 2004). Social status, age and gender determine who has authority and leads social interactions. Complexities of interdependence and individuation, cultural context and collective influences are key to understanding a person’s actions, more so than individual psychological processes (Ashbourne & Baobaid 2019).

While family responsibilities are important in both individualist and collectivist societies, collectivist cultures generally emphasise family obligations over individual autonomy, prioritising social harmony and support. Hierarchies based on age and gender are more socially acceptable and families do not disclose or discuss their family-related concerns to outsiders, “saving face” or protecting the family name (Sawrikar & Katz, 2008). For example, Sawrikar and Katz (2008) remind us how for some communities, mental illness is viewed as shameful and reflecting badly on the family. This said, of the 23 CALD offenders recorded as having mental health issues, two in five had sought treatment for those issues (43%, n=10; information missing for nine cases).

Family violence in particular is viewed as a private issue which limits options for the victim. Mistrust of authorities is also a tremendous contributor to the reluctance of victims to seek assistance as many members of CALD communities may have migrated from countries where police corruption is common. There was evidence that 70 per cent (n=33) of CALD offenders had been abusive towards the victim prior to the lethal violence (information missing for 13 cases). However, of these cases, only 24 per cent (n=8) had protection orders issued against the offender (information missing for eight cases) and only two victims (4%) had sought support from domestic violence services.

Importantly, First Nations researchers have also described Aboriginal and Torres Strait Islander peoples as commonly having “holistic and collectivist worldviews that prioritise the wellbeing of the group above one’s own individual needs” (Garvey et al., 2021, p. 1; see also Gee et al. 2014). However, the extent to which these collectivist viewpoints influence the attitudes of First Nations peoples towards IPV and help-seeking for IPV has not been the subject of extensive research. This said, several studies have highlighted that key barriers to reporting of IPV among Aboriginal and Torres Strait Islander peoples include fear of censure by community members. As noted by Bryant and Willis, “In a collectivist society, more attention may be given to what other family members think and withdrawal of their support can have devastating impacts on social connectivity and personal wellbeing” (2008, p. 63). Further research into the relationship between collectivism and attitudes towards IPV in Aboriginal and Torres Strait Islander families and communities may improve understanding of, responses to and prevention of IPV.

Changes in power dynamics within the relationship

Migration itself can change the dynamics of the family unit, exacerbated by acculturation gaps between individual family members; levels of adjustment to a new country and way of life will be different for each individual and group (Bhugra, 2004). The issue of power relations within couples has been seen as an important aspect of IPV. Jin and Keat (2010) describe power as a multidimensional construct including bases (resources/power), processes (communication control), and outcomes (decision-maker). It is possible that some immigrant men from more patriarchal societies will perceive that they are losing decision-making power as a result of immigration (Jin & Keat, 2010). This is especially true of CALD offenders who had confirmed exposure to hegemonic masculine communities (30%, n=14; information missing for 33 cases). This diverse group included offenders originating from northern Europe, south-eastern Europe, South America, southern East Africa, Central Asia, north-east Asia, South Asia, South-East Asia, Melanesia and the Middle East.

When considering power dynamics within CALD communities, honour, shame and “saving face” cannot be ignored. Honour is not a unidimensional concept and varies by definition and expression across groups to include both valued and destructive behaviours (honour-related violence; Ashbourne & Baobaid, 2019). Blaming a woman for being a victim of violence extends blame also to her mother, sisters and children. Women may feel forced to remain quiet about violence and may first turn to family or members of their community for assistance. The support offered may further disempower them when it is tied to conditions meant to preserve reputation and avoid confrontation (Ashbourne & Baobaid, 2019). Families may resist seeking help as it conflicts with their social norms, compounded further by traditional gender roles of women as carers rather than as those who are cared for (Sawrikar & Katz, 2008). Traditional gender roles may also prevent men from engaging with services or discussing family difficulties.

Changes in the power dynamics within relationships may be partially attributable to differences in gender role expectations between the country of origin and the new country, as well as the availability of new opportunities for women that may not have been accessible previously (Baobaid et al., 2018). Marriage is heavily weighted in collective communities, with registered marriages making up 70 per cent of CALD cases in this study (n=33). For some groups, breaking of a marriage contract (including promises to marry) is unacceptable. Divorce is not an option and is believed to bring shame to the family. Gender role changes within relationships may contribute to higher levels of marital conflict following post-migration settlement. This may be particularly likely to occur in situations where male partners experience issues finding suitable employment in the new country or in instances where women may have challenged the offender’s perceived control over them in the post-migration context.

The analysis identified that in many cases involving a CALD offender, changes in the power dynamics within the relationship post-migration was a key component of the pathway to IPH. For example, in Case 53 the victim refused to attend morning prayers with the family which was tradition; further, in Case 168 conflict within the relationship increased after the victim chose to attend a separate place of worship than the offender. Finally, in Case 158 the offender was considered very traditional and religious with a strong patriarchal view of gender roles. He would discipline his daughters but not his sons, which the victim openly did not agree with. The victim refused to wear a hijab and demanded the offender contribute to household duties or move out. This led to the offender killing the victim.

In other cases, the change in power within the relationship may have been influenced by the relative ease of assimilation experienced by the victim when compared to the offender. In particular, if the offender believed that the victim was integrating too well or too easily, they began to feel threatened and used violence as a means of re-establishing their control over her. This was demonstrated in Case 131. The offender was a refugee who returned to his country of origin to marry a “traditional” wife. When the victim eventually joined the offender in Australia, she became pregnant quickly, prompting the offender to question her fidelity. The victim also seemed to adjust easily to life in Australia which did not sit well with the offender. His paranoia and jealousy increased and he became aggrieved that she was not behaving like the traditional wife he had expected, which led to him killing her (see also Case study 9).

Residency instability

IPV is one of the most common victimisations experienced by migrants regardless of immigration status (Erez et al., 2009). However, Adams and Campbell (2012) discuss how fear of deportation and limited access to services increase fear of reporting abuse and seeking help among female migrants. For refugees, fear of being returned to the country from which they fled to escape conflict or persecution is a powerful motivator that prevents them from seeking help from family services.

Numerous Australian studies have shown that in situations where one partner’s residential status is dependent on the other, IPV may be more likely to occur (Lyneham & Richards, 2014; Segrave, 2017; Segrave & Pfitzner, 2020). Concerns about deportation may be exploited by abusers who threaten to withdraw their support or make false claims about the victim as a means of making them more dependent on the offender and limiting their help-seeking options.

However, as demonstrated in the current study, fear of deportation for non-citizens, especially those on spousal sponsorship visas, is not limited to immigrant women. In some cases included in the PIPH sample, the offender’s residency in Australia was dependent on the victim who would make threats to have them deported. For example, Case 170 (crossover FT/DAS) involved an offender who was a refugee and previous prisoner of war who fled to the United States to escape conflict. He immigrated to Australia to pursue a relationship with the victim, which started to deteriorate after his arrival. The victim threatened the offender with revoking his partner visa which appeared to contribute to his decision to kill the victim.

Discussion

The three offender cohorts and pathways identified in the analysis of the PIPH dataset were differentiable from one another in several key ways. These differences were apparent when looking across the offender’s life course, including their life before starting the relationship with the victim, the relationship between the victim and offender, and the major events that appeared to be significant “turning points” in the offender’s life and relationship that increased their risk of perpetrating lethal violence.

Before the relationship started

When looking at the life histories of offenders prior to starting their relationship with the victim, the PD and DAS offenders were distinguishable from the FT cohort primarily because of their experiences of co-occurring physical and mental health conditions. Although FT offenders reported high levels of mental health issues, particularly depression and anxiety, unlike the PD and DAS cohorts they were unlikely to be experiencing co-occurring and complex health issues (e.g. AOD and mental health issues).

However, there were also some key differences between the DAS and PD offenders regarding their health status. While the PD offenders were primarily described as having significant and chronic AODs, it was more commonly reported that DAS offenders had significant LTHCs and mental health issues. Further, while it was frequently identified that AODs had started during PD offenders’ childhood and adolescence, DAS offenders’ mental and physical health conditions appeared to start during adulthood.

Another difference between the three offender cohorts was their reported involvement in violent offending against former intimate partners, as well as family members, friends and strangers. A significant proportion of PD offenders had been violent towards former partners and others. In comparison, a smaller proportion of FT offenders had previously been violent towards former partners but violence towards non-intimate partners was rarely reported, and there were very low levels of reported violent tendencies among the DAS offender cohort generally.

Finally, while FT offenders were described as being functional in some public-facing domains of their life, this was not always the case for the DAS and PD offenders. In particular, most of the FT offenders were employed in well-respected industries or owned their own businesses. In comparison, PD offenders had experienced extended periods of unemployment and had very low levels of education. Further, while some DAS offenders had been employed at different stages of their life, it was noted that some were on disability pensions or had to retire early due to their chronic health conditions.

After the start of the relationship

A key point of difference between the three offender cohorts was the nature of their relationship with the victim. Overall, FT offender relationships were characterised as involving power imbalances between the victim and offender, exacerbated by the often significant age difference between them. However, FT offender relationships were typically long-term and committed, with marriage and the presence of shared children being common. PD offenders’ relationships with victims were often short-lived and characterised as “toxic” and marked by persistent arguments and conflict. In comparison to both of these cohorts, the DAS offenders’ relationships were very long-standing and were described overall as positive and “happy” by family members and others.

Changes in patterns of violence and abuse within the relationship

While FT offenders were controlling, jealous and manipulative, PD offenders were physically aggressive and violent, and had inflicted significant injuries on the victim. However, DAS offenders were unlikely to have been described as abusive towards the victim, and when abuse was reported, it was described as episodic and minor in nature. Although PD offenders had often been reported to the police for the violence they had inflicted against the victim, it was unlikely that FT and DAS offenders were reported to the police prior to the lethal incident for violence towards the victim.

Interestingly, the escalation of violence and abuse within the relationship was primarily characteristic of FT offenders; in the context of losing control over the victim – including separation – they would engage in stalking and emotionally abusive behaviours, as well as threaten to kill the victim. The increased risk to victims of being seriously harmed by the offender was understood by some women, with some starting to reach out to domestic violence services and the police for the first time. In comparison, PD offenders’ violence and abuse did not appear to change significantly in the lead-up to the lethal incident. Rather, the lethal incident occurred in the context of ongoing and persistent patterns of abuse, with the lethal incident itself looking very similar to previous assaults perpetrated by the offender. Finally, for many of the DAS offenders, the first and only incident of IPV that they had perpetrated against the victim within the relationship was the lethal incident itself.

Key events that increased the likelihood of lethal violence occurring

The key events that led to the offender’s decisions, instantaneous or otherwise, to kill the victim differed across the three groups. For FT offenders, their primary motivation for killing the victim was to regain control over them. In these cases, the victim was perceived by the offender as challenging their authority and dominance within the relationship. This was primarily attributable to the victim leaving the offender and refusing to reconcile with them, but other actions perceived as “challenging” by the offender included seeking employment opportunities, succeeding in their career and refusing to acquiesce to their (the offender’s) demands.

For DAS offenders, the main events that contributed to their pathway was the deterioration in their physical and mental health, which may have been attributable to the exacerbation of underlying conditions, or the onset of new issues. The deterioration in the offender’s health appeared to have a negative impact on their relationship with the victim, and led to increased conflict.

In comparison to both the FT and DAS groups, PD offenders were noteworthy for the lack of significant events in the lead-up to the lethal incident that may have helped to explain why they killed their partner. Rather, as noted previously, the violence appeared to be consistent with previous assaults that may have occurred within the relationship. However, the risk of lethal outcomes may have been attributable to the presence of situation-specific vulnerabilities such as their intoxication levels and the presence of a weapon.

Implications for policy and practice

Taken together, the three pathways highlight that to prevent IPH, different responses need to be implemented at different stages of each trajectory.

Fixated threat: Making invisible men visible

In responding to and disrupting FT offender trajectories, two issues of immediate concern need to be addressed. First, FT offenders – and the risk that they pose to victims – are not visible to law enforcement. FT abusers do not share many similarities with male abusers in general who come into contact with the police (see for example Hulme et al., 2019). In particular, FT abusers were typically employed or ran their own businesses, were middle class and had low levels of AOD. Importantly, while some FT offenders had been in contact with the criminal justice system for violent and other offending behaviours, these contacts were described by individuals within the dataset as relatively minor, and not indicative of their potential for subsequent extreme levels of violence.

Further, a major barrier to these abusers coming to the attention of the police was that primarily, their abuse involved non-physical forms of coercive controlling behaviours, including stalking which is covert in nature, and emotional and verbal abuse. That criminal justice systems are not well placed to respond to the threat associated with FT abusers is partially the rationale for ongoing discussions within Australia about the criminalisation of coercive controlling behaviours (McMahon & McGorrery, 2016, 2020; Walklate & Fitz-Gibbon, 2019), as well as educating frontline staff to identify when coercive control is present, and to treat it seriously when it is detected.

The second issue that arises in relation to the disruption of FT offender trajectories is that the period of acute escalation observed for many FT offenders, that coincided with their loss of control over the victim that immediately preceded the lethal incident, was very short. As described in earlier sections of the report, among cases involving an FT offender where the victim and offender had separated, the median length of time between the separation and the lethal incident was only two months.

Taken together, these identified issues highlight the need for:

• new techniques for facilitating the detection of acute risk posed by FT offenders, using information that may not be collected by traditional law enforcement agencies
• targeted and timely responses that can be implemented quickly to protect the safety of high-risk victims and their families.

One such strategy which may address the first issue regarding detection is intelligence-led approaches to IPV risk assessment. Intelligence is a broad term that can be applied to real-time information collected by law enforcement analysts about individuals, in this case IPV perpetrators. Intelligence that would be of direct relevance to identifying FT offenders that may not be readily available in existing criminal justice datasets includes:

• GPS data – to determine when they may be stalking the victim by attending their premises or following them
• online activity data to determine whether they have been stalking the victim online
• mental health data to determine if offenders are seeking support for suicidal ideation, disturbed sleep and changes in mood
• family law processes information to determine whether there are family law processes that are occurring in the context of separation.

These intelligence-led approaches to identify individuals who pose a high risk to the safety of the community have been used in other contexts, most notably fixated threat assessment centres (FTACs). FTACs were first developed in the United Kingdom in 2006 and in recent years have been introduced in almost every jurisdiction in Australia. The FTAC is designed to risk-assess and case-manage individuals that are fixated on public figures, specifically politicians (James et al., 2013). Since their introduction, FTACs across Australia have managed and de-escalated the threat posed by fixated individuals in the community.

In the first instance it may be possible to identify individuals who should be the subject of more intensive risk assessment. In particular, several red flags emerged from the analysis of the FT offender trajectory which could assist in identifying this cohort of offender prospectively. These include the victim seeking a first-time protection order against the offender, in the context of her separation from the offender. This contact with the legal system could act as a potentially useful trigger for law enforcement and domestic violence services having contact with the victim and conducting a risk assessment. If there are also potential stalking behaviours present, and/or threats to kill, this may in turn result in the implementation of additional safety mechanisms to protect vulnerable victims during high-risk periods for homicide.

Once FT offenders have been made visible, the subsequent response needs to be implemented quickly and intensively to manage the threat posed by these individuals to their partners and families. This would necessarily involve close surveillance and monitoring of abusers, as well as safety planning with victims, including facilitating access to emergency housing and the replacement of technological devices that may have been compromised through the installation of monitoring software by the abuser.

Key components of responses to disrupt the FT trajectory
Timely	Surveillance and monitoring	Intelligence collection	Safety planning with victims

Persistent and disorderly: Provide targeted support early

The PD offenders were the most “visible” within the PIPH sample. PD offenders were typically well known to police, with extensive criminal offending histories for violent and non-violent crime. They also tended to have histories of domestic violence offending, often reported for multiple partners. PD offenders characteristically demonstrated high levels of mental health disorders, alcohol and other drug problems and chronic physical health conditions. Traumatic experiences during childhood and adolescence were also highest among the PD offender cohort, which included witnessing IPV between carers and being a target of family violence .

PD offenders presented a profile of complex, chronic and co-occurring health issues early in life and during adulthood. However, as a cohort PD offenders had limited engagement with mental health and other services and low adherence to symptom management regimens. In other words, many PD offenders needed support for their (multiple) health conditions but the majority did not receive it. For example, although 36 per cent of PD offenders had a diagnosed or suspected MI (n=26; information missing for 43 cases), only half had received any form of mental health treatment (46%, n=12; information missing for 12 cases). Similarly, although three in four PD offenders had an AOD (74%, n=54; information missing for 18 cases), only 17 per cent had received support for alcohol and other drug problems (n=9; information missing for 36 cases). This said, among the small number of offenders who did receive treatment, it appeared to have limited if any benefit.

Despite PD offenders presenting with challenging and complex needs, the evidence base regarding “what works” with this type of offender (i.e. high-risk violent) is much stronger when compared to the other two pathways. The idea that violent offenders cannot be rehabilitated is a persistent myth despite evidence that appropriate treatment can reduce recidivism (Skeem et al., 2009). Incidents of IPH committed by PD offenders are preventable through early intervention during childhood and adolescence and by providing targeted, integrated and timely support in adulthood (Flood & Webster 2007; Karakurt et al., 2019).

The type of environment in which an individual grows up, and the nature of their experiences, has a significant impact on a person’s social and emotional development. In particular, the family unit is the setting where people’s ideas about how to interact with others are shaped and where conflict resolution skills are first learned (Bandura, 1977). As noted in previous sections of this report, attachment theory argues that children need to form close relationships with caregivers early in life in order for them to be able to connect emotionally with others and self-regulate their own emotions (Liem & Koenraadt, 2018). Developmental instability, trauma and not feeling safe as a child significantly impact on psycho-social development and make it difficult for people to have healthy adult relationships. People who have this history may deeply desire intimate connections with others but they also have an intense fear of being rejected or ridiculed. When they feel threatened, they may become easily angry, distrustful and violent in response (see for example Doumas et al., 2008; Mahalik et al., 2005).

Our findings show that risk factors associated with IPV and IPH often “clustered” together for PD offenders, particularly experiences of maltreatment and behavioural problems early in life. Prevention responses most relevant to PD offenders need to be multipronged and targeted across family, community and school settings (see e.g. Pathways to Prevention; Batchelor et al., 2006). For example, home-visiting programs can promote healthy social and emotional development and functioning by teaching parents to model healthy relationships; effectively manage their children’s behaviour without resorting to harsh punishment; and help their children to develop appropriate anger management, emotional regulation and social skills (Holzer et al., 2006). Similarly, school-based approaches that target communication and problem-solving, gender and power dynamics, safety and sexuality have demonstrated positive effects in reducing IPV perpetration among adolescents (De Koker et al., 2014). Programs delivered in the classroom are appealing as you can tailor interventions to be appropriate for students’ developmental level, and target multiple risk factors at once (i.e. alcohol use, mental health, decision-making; Cox et al., 2016). While both community and school-based approaches have been demonstrated to reduce aggression and violence, more evaluation is needed to demonstrate long-term benefits beyond adolescence (Averdijk et al., 2020; Kovalenko et al., 2020).

While preventing early adversity is key, it is also important to provide support to address contextual and situational risk factors for violence such as comorbidity and alcohol use (Lawler et al., 2021). As comorbidity is characteristic of many of the PD offenders, policy and practice approaches that aim to prevent IPH need to implement responses tailored to this need. Violent offenders with comorbidity are a particularly challenging group because they are difficult to engage with treatment and more likely to reoffend (Barrett et al., 2014). Further, despite strong associations between IPV and AOD, MBCPs and AOD treatment programs have been historically siloed and existing interventions to reduce IPV in the context of AOD lack effectiveness (Gilchrist et al., 2019; Gilchrist & Hegarty, 2017). This is problematic as male IPV offenders who drink hazardously are more likely to reoffend after completing an MBCP than IPV offenders without alcohol use problems (Stuart, 2005). Importantly, reviews of the research show that combined alcohol and violence treatment programs can have benefits for reducing IPV among perpetrators with alcohol problems (Wilson et al., 2014). However, the absence of compelling evidence of the efficacy of IPV perpetrator programs that are integrated with AOD interventions suggests that additional work is needed to understand the interrelationships between AOD and IPV (Stephens-Lewis et al., 2019).

Another policy challenge arising from our findings is that PD offenders on face value may appear similar to non-IPV violent offenders and other prisoners in the criminal justice system. Like most prisoners, they had high levels of disadvantage, maltreatment, alcohol and other drug problems, mental health issues, poor health generally and comorbidity. In light of this, the goal is now how to identify those men who demonstrate this typical antisocial behaviour personality profile who carry a high risk of IPH, compared to those who do not pose a lethal violence risk. One indicator may be the presence of a protection order and whether the offender has a history of breaching a protection order against a current or former partner. Our findings suggest relatively high rates of protection orders against offenders in the PD pathway (40%, n=29; information missing for 29 cases).

Although protection orders are a common response implemented to improve victim safety and reduce reoffending, that so many of the PD offenders were the subject of protection orders or other court orders at time of killing the victim indicates that they cannot be effective in all situations. This is supported by other research looking at the efficacy of protection orders in reducing IPV recidivism (see for example Dowling et al., 2018). In particular, in some cases involving a PD offender where the relationship was ongoing, the implementation of court orders and protection orders may actually increase the risk of lethal violence; as victims and offenders attempt to avoid detection, they may spend more time together by themselves in the absence of bystanders who may be able to intervene.

Taken together, the findings highlight the need for streamlined integrated service delivery for mental health and substance use problems among IPV perpetrators. Criminal justice engagement is an opportunity to provide rehabilitation to offenders with complex needs such as the PD offenders in this study. For some PD offenders, their time in prison was the only time they could stop using alcohol and other drugs. This signals an opportunity to intervene with such offenders to engage them with treatment and make plans to change. Unfortunately, there is a significant lack of effective programs for addressing substance use problems among prisoners (Doyle et al., 2019). The most common approach to treatment in custodial settings has been psychoeducation which has been shown to be ineffective in changing behavior as a standalone intervention (Belenko et al., 2013). Meta-analyses show that multimodal treatments that incorporate cognitive behavioural therapy (CBT) principals can be effective in reducing violent and non-violent/general recidivism when delivered in correctional and forensic mental health settings (Papalia et al., 2019). Unfortunately, the evidence base is limited by few well-designed evaluation studies and difficulty comparing evidence because of inconsistences in reporting findings (Papalia et al., 2019). However, research with Australian prisoners has shown promise for integrated treatment for comorbid substance use and PTSD using a CBT-based model, but larger scale trials are needed (Barrett et al., 2015).

Our research also supports previous work showing that IPH perpetrators with comorbidity have low engagement with mental health services (Oliver & Jaffe, 2018). This is a problem as research shows that prisoners with comorbidity are more likely to reoffend than those with either mental health or substance use disorders alone (Smith & Trimboli, 2010). Even among men who engage with treatment for domestic violence, only half of them feel they receive the support they need to reduce their violent behaviour (Campbell et al., 2010). Strategies for engaging abusive men in help-seeking include challenging ideas around masculinity and help-seeking, developing trust in non-judgmental therapeutic relationships, and engaging bystanders (friends, family, community leaders) through equipping them to intervene when necessary and provide support (Campbell et al., 2010). Treatment developed through co-design that targets feelings of anger, guilt and shame, alongside alcohol and other drug use, are most relevant for IPV offenders with substance use problems (Curtis et al., 2021).

Even with this evidence, our findings show that a key area for future research is to understand the barriers to help-seeking for this group and identify ways to motivate engagement with treatment. In order to prevent IPH among PD offenders, we must determine which treatment components are effective, in what combination and for whom.

Key components of responses to disrupt the PD trajectory
Early intervention	Integrate AOD and MI services in IPV intervention programs	Build safety planning around protection orders	Evidence-based IPV programs delivered in criminal justice settings

Responding to DAS offender trajectories

DAS offenders were the least “visible” of all offenders included in the PIPH sample. That is, previously identified risk factors were not (for the most part) present. For example, DAS offenders were unlikely to have had experiences of trauma, a history of offending or reported IPV perpetration. Further, DAS offenders tended to be in long-term and “happy” relationships with their victims. However, this changed when DAS offenders experienced a shift in their trajectory related to the onset or exacerbation of a significant life stressor(s). This triggered a rapid deterioration in their ability to cope with their circumstances and coincided with the onset and escalation of their (lethal) violent behaviour.

While FT offenders were similarly not visible to law enforcement, they did perpetrate IPV (particularly non-physical forms of coercive controlling behaviours) which escalated in the context of separation. As such, there was often a constellation of risk factors present within the relationship that could be identified through intelligence-led approaches to IPV risk assessment, as discussed above. However, the challenge with DAS offenders is that there do not appear to be clear risk factors that differentiate DAS offenders from “ordinary” people in the wider community. In particular, it is not clear how mental deterioration associated with increasing risk of IPH differs from mental deterioration that carries no risk of IPH. Therefore, there is a lack of obvious early intervention points for preventing IPH among DAS offenders. However, this study has provided some preliminary insights into how risk of lethal harm among DAS offenders could be identified, highlighting critical avenues for further investigation.

A defining characteristic of the DAS offender pathway was the presence of acute mental, emotional and physical health problems, particularly in the lead-up to the event. DAS offenders were likely to have been diagnosed with a mental illness throughout their lifetime. Notably, two thirds (68%) of DAS offenders had a diagnosed or suspected mental illness. This was a higher rate of mental illness than has been reported previously (11% to 46% of IPH offenders; Bridger et al., 2017; Campbell et al., 2007; Sebire, 2017). Further, over half of DAS offenders who had a diagnosed mental illness had received treatment (54%, n=7; information missing for five cases), a higher proportion than was found among FT (36%, n=9; information missing for 10 cases) and PD offenders (46%, n=12; information missing for 12 cases) who had been diagnosed with a mental illness. This is important evidence, as it shows that a large proportion of offenders on this pathway may be engaging with services, thus indicating a (missed) opportunity for intervention. Mental health professionals may be well placed to conduct risk assessment and identify those whose risk of IPH perpetration may increase in the context of deteriorating mental health. Spencer and Stith (2018) argue that it is imperative for practitioners to incorporate their own professional judgement when conducting risk assessments for IPH, as risk factors such as mental illness may not appear alongside more “expected” risk factors such as IPV perpetration.

Beyond mental health issues, DAS offenders were also likely to have had an LTHC, such as arthritis or chronic pain, which often co-occurred with mental illnesses and/or AOD. Indeed, one in three DAS offenders had an LTHC. Unfortunately, very little is known about the relationship between poor physical health and IPH (Singh et al., 2014). The lack of evidence hinders our ability to make recommendations regarding translation of this new evidence into clinical practice. Further research is needed to better understand the relationship between physical health conditions and risk of IPH, as well as how best to identify and intervene when this risk becomes heightened. However, similar to identifying increasing risk associated with mental illness, it may be that allied health professionals involved in an individual’s care have a unique opportunity to identify risk of IPH in the context of deteriorating physical health problems.

When the violence seems to come out of nowhere, there appears to be a lack of clear avenues that can be tailored for prevention. In many DAS cases, the best indicator of IPH risk was that family or friends of the offender noticed a change in the offender’s thinking or behaviour in the lead-up to the incident that was associated with stressors unrelated to the victim. Therefore, family and friends of individuals may also play an important role in identifying risk by noticing an individual’s deterioration in mental or physical health and associated changes in behaviours. This may be particularly important where offenders do not have contact with allied health professionals, as nearly half of DAS cases did not receive treatment prior to the homicide.

Here, bystander intervention programs may be beneficial. Bystander intervention programs for violence prevention aim to improve participant understanding of violence issues and increase people’s willingness and self-efficacy to intervene before, during or after the problem starts (Australian Human Rights Commission, n.d.). While our findings provide important evidence about how risk for IPH may be “flagged”, to prevent IPH bystanders would need to recognise such risk and act relatively quickly given the rapid decline in functioning most DAS offenders demonstrated. Further research would be needed to inform development of appropriate bystander intervention programs on the relationship between poor health, IPV and risk of IPH, particularly in regard to how potential risk factors for IPH are different to more established risk factors for this cohort (e.g. physical health deterioration in the context of IPV). However, it is positive that bystander intervention programs for IPV more broadly have been shown to be effective in helping participants both to understand IPV and to feel more confident in their ability to intervene safely in situations where violence is occurring or is likely to occur (McMahon & Dick, 2011). Of course, bystander intervention is an IPH prevention strategy that is relevant to all pathways identified in this research.

In summary, this research has identified that DAS offenders are a subset of IPH offenders who are not as clearly “visible” as PD offenders or as potentially “visible” as FT offenders. While they did not present with many established risk factors for IPH, it may be possible for health services and family and friends to identify escalating risk of IPH perpetration associated with mental and/or physical health deterioration. However, DAS offenders themselves are primarily responsible for their own behaviour, to monitor their own emotional state and to seek support for themselves if they are having difficulty coping. Therefore, information and awareness campaigns should also be targeted at men who are experiencing difficulty coping with stress, and also aim to address barriers to help-seeking among men. This includes challenging masculine tropes and stereotypes which may limit help-seeking among men, including that asking for help is akin to “weakness”.

Key components of responses to disrupt the DAS trajectory
Training of health professionals to identify risk of IPH	Research to understand the links between physical and health conditions and IPH	Information and awareness-raising targeted at men experiencing stress	Bystander training to support family members and friends to identify risk

Conclusion

The findings from this research provide detailed evidence and information about pathways to IPH in the Australian context. Three key IPH offending pathways were identified: FT, PD and DAS. While they share common characteristics, these distinct pathways require specific and tailored policy and practice responses to identify and disrupt potential IPH offenders before they perpetrate lethal harm. More generally, the findings from this research contribute to an already large body of evidence that highlights the importance of intervening early with children and young people to prevent the occurrence of IPV and IPH later in life. This includes programs and strategies that aim to prevent the occurrence of adverse childhood experiences and mitigate their impact when they do occur, as well as educating young men and women about gender norms, consent and respectful relationships. However, in light of the evidence that many of the key events, milestones or changes that occurred in the lives of offenders occurred once they reached adulthood, there is a need to focus our attention on how we respond during these stages of the life course. Particular stages or points in the life course that may signal an opportunity for intervention include:

• the offender’s contact with medical services, including methadone programs and alcohol and drug use rehabilitation programs, for escalating alcohol and drug use problems
• the offender’s contact with medical services and statutory workers’ compensation schemes due to the escalation or onset of long-term health conditions, including physical injuries and chronic pain
• the offender’s and victim’s contact with the criminal justice and family law systems for IPV offending and applications for protection orders
• the offender’s and victim’s contact with the Australian migration system and government and non-government settlement programs during post-migration periods
• the offender’s affiliation or association with hegemonic masculine communities such as sport teams and clubs, body-building communities, the armed forces and motorcycle clubs.

Finally, in the event that there has been no opportunity for disrupting an offender’s trajectory towards IPH, or the offender has been unmotivated to take up these opportunities, the focus needs to shift towards safety planning and protecting women and children from lethal violence during high-risk periods. This includes the post-separation period, as well as during disputes related to the custody of shared children and financial settlement.

Overall, the research found there is not one single, universal pathway to IPH. Indeed, the journey to IPH is instead a series of pathways that branch, weave, and intertwine, depending on circumstance. Despite the potential complexity and diversity of the pathways to IPH, it is possible to identify intervention points in many cases. Taken together, this body of work provides evidence and guidance to better respond to prevention of men’s lethal violence against women.

However, it is important to recognise again that the PIPH dataset included significant amounts of missing data. In particular, a key limitation of the current analysis was the absence of victim information and “voice”. This meant that our understanding of who victims were and the major events and milestones in their lives that may have led them to their relationship with the offender was very limited. Further, considering that IPV, particularly non-physical forms such as coercive control, may not be detected by family members and friends or be reported to the police, in the absence of the victim’s testimonials and voice, the information we have about the history of violence and abuse within the relationship was also limited.

The absence of victim information and its implications for current understanding of IPH has been noted in multiple studies (Erikkson et al., 2021; Johnson et al., 2017). Some researchers have attempted to address this knowledge gap by engaging with the friends and family members of victims who are asked to describe the victim and their life histories, and the nature of their relationship with the offender. While these studies have provided valuable information about victims, they are also limited by small sample sizes (Erikkson et al., 2021; Monckton Smith, 2019). This is in part attributable to the difficulty of both recruiting victims’ family members and friends to participate in research and conducting these interviews.

This raises the question of how we include the voices of women of homicide in research. This information is not only crucial for developing accurate trajectories and pathways, but is also necessary to ensure that we do not lose sight of the fact that underneath these statistics, there are real women whose lives were taken by their partners.

Footnotes

1. This case was unique and therefore categorised as an outlier for the sentencing judge determined it to be a mercy killing ↑

2. Please note: URLs are correct as of September 2021. ↑

References

Adams, M. E., & Campbell, J. (2012). Being undocumented & intimate partner violence (IPV): Multiple vulnerabilities through the lens of feminist intersectionality. Women’s Health and Urban Life, 11. https://tspace.library.utoronto.ca/handle/1807/32411^[2]

Agnew, R. (1992). Foundation for a general strain theory of crime and delinquency. Criminology, 30(1), 47–88. https://doi.org/10.1111/j.1745-9125.1992.tb01093.x

Aldarondo, E., & Sugarman, D. B. (1996). Risk marker analysis of the cessation and persistence of wife assault. Journal of Consulting and Clinical Psychology, 64(5), 1010–1019. https://doi.org/10.1037/0022-006X.64.5.1010

Aldridge, M. L., & Browne, K. D. (2003). Perpetrators of spousal homicide: A review. Trauma, Violence, & Abuse, 4(3), 265–276. https://doi.org/10.1177/1524838003004003005

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). APA.

Angelakis, I., Austin, J. L., & Gooding, P. (2020). Childhood maltreatment and suicide attempts in prisoners: A systematic meta-analytic review. Psychological Medicine, 50(1), 1–10. https://doi.org/10.1017/S0033291719002848

Ashbourne, L. M., & Baobaid, M. (2019). A collectivist perspective for addressing family violence in minority newcomer communities in North America: Culturally integrative family safety response. Journal of Family Theory & Review, 11. http://doi.org/10.1111/jftr.12332

Australian Bureau of Statistics. (2008). National Survey of Mental Health and Wellbeing: Summary of results. https://www.abs.gov.au/statistics/health/mental-health/national-survey-mental-health-and-wellbeing-summary-results/2007

Australian Bureau of Statistics. (2016). National Aboriginal and Torres Strait Islander Social Survey 2014–2015 (No. 4714.0). https://www.abs.gov.au/ausstats/abs@.nsf/mf/4714.0

Australian Bureau of Statistics. (2018). National Health Survey: First results, 2017–18. https://www.abs.gov.au/statistics/health/health-conditions-and-risks/national-health-survey-first-results/latest-release

Australian Bureau of Statistics. (2021). National, state and territory population. https://www.abs.gov.au/statistics/people/population/national-state-and-territory-population/latest-release

Australian Domestic and Family Violence Death Review Network. (2018). Australian Domestic and Family Violence Death Review Network: Data report 2018. https://coroners.nsw.gov.au/documents/reports/ADFVDRN_Data_Report_2018%20(2).pdf

Australian Human Rights Commission. (n.d.). Part 4: Bystander interventions in violence prevention. https://humanrights.gov.au/our-work/part-4-bystander-interventions-violence-prevention

Australian Institute of Health and Welfare. (2007). Disability in Australia: Acquired brain injury. https://www.aihw.gov.au/reports/disability-services/disability-australia-acquired-brain-injury/summary

Australian Institute of Health and Welfare, & Australian Bureau of Statistics. (2005). The health and welfare of Australia’s Aboriginal and Torres Strait Islander peoples 2005. https://www.aihw.gov.au/reports/indigenous-health-welfare/health-welfare-australias-indigenous-peoples-2005/contents/table-of-contents

Australian Law Reform Commission. (2018). Pathways to justice—An inquiry into the incarceration rate of Aboriginal and Torres Strait Islander peoples. https://www.alrc.gov.au/wp-content/uploads/2019/08/final_report_133_amended1.pdf

Averdijk, M., Eisner, M. P., Luciano, E. C., Valdebenito, S., & Obsuth, I. (2020). Effective violence prevention: An overview of the international evidence. Violence Research Centre, Institute of Criminology, University of Cambridge. https://www.vrc.crim.cam.ac.uk/files/vp_report_-_digital_version_-_not_for_printing.pdf

Ayton, D., Pritchard, E., & Tsindos, T. (2021). Acquired brain injury in the context of family violence: A systematic scoping review of incidence, prevalence, and contributing factors. Trauma, Violence, & Abuse, 22(1), 3–17. https://doi.org/10.1177/1524838018821951

Babcock, J. C., Jacobson, N. S., Gottman, J. M., & Yerington, T. P. (2000). Attachment, emotional regulation, and the function of marital violence: Differences between secure, preoccupied, and dismissing violent and nonviolent husbands. Journal of Family Violence, 15(4), 391–409. https://doi.org/10.1023/A:1007558330501

Baidawi, S., & Sheehan, R. (2019). “Crossover kids”: Offending by child protection-involved youth . Trends and Issues in Crime and Criminal Justice, no. 582. Australian Institute of Criminology. https://www.aic.gov.au/publications/tandi/tandi582

Balica, E., & Stöckl, H. (2016). Homicide-suicides in Romania and the role of migration. European Journal of Criminology, 13(4), 517–534. https://doi.org/10.1177/1477370816633258

Bandura, A., & Walters, R. H. (1977). Social learning theory (Vol. ١). Englewood Cliffs.

Banyard, V. L., Rizzo, A. J., & Edwards, K. M. (2020). Community actionists: Understanding adult bystanders to sexual and domestic violence prevention in communities. Psychology of Violence, 10(5), 531–541. https://doi.org/10.1037/vio0000281

Barrett, E. L., Teesson, M., & Mills, K. L. (2014). Associations between substance use, post-traumatic stress disorder and the perpetration of violence: A longitudinal investigation. Addictive Behaviors, 39(6), 1075–1080. https://doi.org/10.1016/j.addbeh.2014.03.003

Baobaid, M., Ashbourne, L., Tam, D., Badahdah, A., & Jamal, A. A. (2018). Pre- and post-migration stressors and marital relations among Arab refugee families in Canada. Doha International Family Institute. https://www.difi.org.qa/publications/pre-and-post-migration-stressors-and-marital-relations-among-arab-refugee-families-in-canada/

Barrett, E. L., Mills, K., & Teesson, M. (2011). Hurt people who hurt people: Violence among individuals with comorbid substance use disorder and post traumatic stress disorder. Addictive Behaviors, 36(7), 721–728. https://doi.org/10.1016/j.addbeh.2011.02.005

Barrett, E. L., Mills, K. & Teesson, M. (2013). Mental health correlates of anger and violence among individuals entering substance use treatment. Mental Health & Substance Use, 6(4), 287–302. https://doi.org/10.1080/17523281.2012.725425

Barrett, E. L., Teesson, M., Chapman, C., Slade, T., Carragher, N., & Mills, K. (2015). Substance use and mental health consequences of childhood trauma: An epidemiological investigation. Drug and Alcohol Dependence, 100(146), e217–e218. https://doi.org/10.1016/j.drugalcdep.2014.09.059

Batchelor, S., Carr, A., Elias, G., Freiberg, K., Hay, I., Homel, R., Lamb, C., Leech, M., & Teague, R. (2006). The Pathways to Prevention project: Doing developmental prevention in a disadvantaged community. Trends & Issues in Crime and Criminal Justice, no. 323. Australian Institute of Criminology. https://www.aic.gov.au/publications/tandi/tandi323

Belenko, S., Hiller, M., & Hamilton, L. (2013). Treating substance use disorders in the criminal justice system. Current Psychiatry Reports, 15(11), 1–11. https://doi.org/10.1007/s11920-013-0414-z

Belfrage, H., & Rying, M. (2004). Characteristics of spousal homicide perpetrators: A study of all cases of spousal homicide in Sweden 1990–1999. Criminal Behaviour and Mental Health, 14(2), 121–133. https://doi.org/10.1002/cbm.577

Benbow, S. M., Bhattacharyya, S., & Kingston, P. (2019). Older adults and violence: An analysis of domestic homicide reviews in England involving adults over 60 years of age. Ageing & Society, 39(6), 1097–1121. https://doi.org/10.1017/S0144686X17001386

Bhugra D. (2004). Migration, distress and cultural identity. British Medical Bulletin, 69(1), 129–141. https://doi.org/10.1093/bmb/ldh007

Block, C. R., & Christakos, A. (1995). Intimate partner homicide in Chicago over 29 years. Crime and Delinquency, 41(4), 496–526. https://doi.org/10.1177/0011128795041004008

Boles, S. M., & Miotto, K. (2003). Substance abuse and violence: A review of the literature. Aggression and Violent Behavior, 8(2), 155–174. https://doi.org/10.1016/S1359-1789(01)00057-X

Boxall, H., Boyd, C., Dowling, C., & Morgan, A. (2018). Understanding domestic violence incidents using crime script analysis. Trends and Issues in Crime and Criminal Justice, no. 558. Australian Institute of Criminology. https://www.aic.gov.au/publications/tandi/tandi558

Boxall, H., & Lawler, S. (2021). How does domestic violence escalate over time? Trends and Issues in Crime and Criminal Justice, no. 626. Australian Institute of Criminology. https://www.aic.gov.au/publications/tandi/tandi626

Boxall, H., & Morgan, A. (2021a). Experiences of coercive control among Australian women. Statistical Bulletin, no. 30. Australian Institute of Criminology. https://www.aic.gov.au/publications/sb/sb30

Boxall, H., & Morgan, A. (2021b). Who is most at risk of physical and sexual partner violence and coercive control during the COVID-19 pandemic? Trends and Issues in Crime and Criminal Justice, no. 618. Australian Institute of Criminology. https://www.aic.gov.au/publications/tandi/tandi618

Bowen, E., Heron, J., Waylen, A., Wolke, D., & Team, A. S. (2005). Domestic violence risk during and after pregnancy: Findings from a British longitudinal study. BJOG: An International Journal of Obstetrics & Gynaecology, 112(8), 1083–1089. https://doi.org/10.1111/j.1471-0528.2005.00653.x

Brady, P. Q., & Hayes, B. E. (2018). The intersection of stalking and the severity of intimate partner abuse. Violence and Victims, 33(2), 218–238. https://doi.org/10.1891/0886-6708.VV-D-16-00193

Bricknell, S. (2020a). Homicide in Australia 2017–18. Statistical Report, no. 23. Australian Institute of Criminology. https://doi.org/10.52922/sr04459

Bricknell, S. (2020b). Homicide in Australia 2016–17. Statistical Report, no. 22. Australian Institute of Criminology. https://doi.org/10.52922/sr04435

Bricknell, S., & Doherty, L. (2021). Homicide in Australia 2018–19. Statistical Report, no. 34. Australian Institute of Criminology. https://doi.org/10.52922/sr78153

Bridger, E., Strang, H., Parkinson, J., & Sherman, L.W. (2017). Intimate partner homicide in England and Wales 2011–2013: Pathways to prediction from multi-agency domestic homicide reviews. Cambridge Journal of Evidence-based Policing, 1, 93–104. https://doi.org/10.1007/s41887-017-0013-z

Brignell, G., & Donnelly, H. (2015). Sentencing in NSW: A cross-jurisdictional comparison of full-time imprisonment (Research monograph 39). Judicial Commission of NSW. https://www.judcom.nsw.gov.au/wp-content/uploads/2016/07/research-monograph-39.pdf

Brower, M. C., & Price, B. H. (2001). Neuropsychiatry of frontal lobe dysfunction in violent and criminal behaviour: A critical review. Journal of Neurology, Neurosurgery & Psychiatry, 71(6), 720–726. https://doi.org/10.1136/jnnp.71.6.720

Brownhill, S., Wilhelm, K., Barclay, L., & Schmied, V. (2005). “Big build”: Hidden depression in men. Australian & New Zealand Journal of Psychiatry, 39(10), 921–931. https://doi.org/10.1080/j.1440-1614.2005.01665.x

Bryant, C., & Willis, M. (2008). Risk factors in Indigenous violent victimisation (No. 30; Technical and background paper). Australian Institute of Criminology. https://www.aic.gov.au/sites/default/files/2020-05/tbp030.pdf

Bushman, B. J., Baumeister, R. F., & Phillips, C. M. (2001). Do people aggress to improve their mood? Catharsis beliefs, affect regulation opportunity, and aggressive responding. Journal of Personality and Social Psychology, 81, 17–32. https://doi.org/10.1037/0022-3514.81.1.17

Butler, T., Allnutt, S., Kariminia, A., & Cain, D. (2007). Mental health status of Aboriginal and non-Aboriginal Australian prisoners. Australian & New Zealand Journal of Psychiatry, 41(5), 429–435. https://doi.org/10.1080/00048670701261210

Caman, S., Howner, K., Kristiansson, M., & Sturup, J. (2016a). Differentiating male and female intimate partner homicide perpetrators: A study of social, criminological and clinical factors. International Journal of Forensic Mental Health, 15(1), 26–34. https://doi.org/10.1080/14999013.2015.1134723

Caman, S., Howner, K., Kristiansson, M., & Sturup, J. (2016b). Differentiating intimate partner homicide from other homicide: A Swedish population-based study of perpetrator, victim, and incident characteristics. Psychology of Violence, 7(2), 306–315. https://doi.org/10.1037/vio0000059

Caman S., Kristiansson, M., Granath, S., & Sturup, J. (2017). Trends in rates and characteristics of intimate partner homicides between 1990 and 2013. Journal of Criminal Justice, 49, 14–21. https://doi.org/10.1016/j.jcrimjus.2017.01.002

Campbell, J. C., Glass, N., Sharps, P. W., Laughon, K., & Bloom, T. (2007). Intimate partner homicide: Review and implications of research and policy. Trauma, Violence, & Abuse, 8(3), 246–269. https://doi.org/10.1177/1524838007303505

Campbell, J. C., Webster, D., Koziol-McLain, J., Block, C., Campbell, D., Curry, M. A., Gary, F., Glass, N., McFarlane, J., Sachs, C., Sharps, P., Ulrich, Y., Wilt, S. A., Manganello, J., Xu, X., Schollenberger, J., Frye, V., & Laughon, K. (2003). Risk factors for femicide in abusive relationships: Results from a multisite case control study. American Journal of Public Health, 93(7), 1089–1097. https://doi.org/10.2105/AJPH.93.7.1089

Campbell, M., Neil, J. A., Jaffe, P. G., & Kelly, T. (2010). Engaging abusive men in seeking community intervention: A critical research & practice priority. Journal of Family Violence, 25(4), 413–422. https://doi.org/10.1007/s10896-010-9302-z

Capaldi, D. M., Shortt, J. W., & Crosby, L. (2003). Physical and psychological aggression in at-risk young couples: Stability and change in young adulthood. Merrill-Palmer Quarterly, 49(1), 1–27. https://doi.org/10.1353/mpq.2003.0001

Carmichael, H., Steward, L., & Velopulos, C. G. (2019). It doesn’t just happen to “other” people – An exploration of occupation and education level of women who die from intimate partner violence. American Journal of Surgery, 218(4), 744–748. https://doi.org/10.1016/j.amjsurg.2019.07.021

Cavanagh, A., Wilson, C. J., Kavanagh, D. J., & Caputi, P. (2017). Differences in the expression of symptoms in men versus women with depression: A systematic review and meta-analysis. Harvard Review of Psychiatry, 25(1), 29–38. https://doi.org/10.1097/HRP.0000000000000128

Cheng, P., & Jaffe, P. (2019). Examining depression among perpetrators of intimate partner homicide. Journal of Interpersonal Violence, 36(19–20), 9277–9298. https://doi.org/10.1177/0886260519867151

Chiu, Y.-N., Leclerc, B., & Townsley, M. (2011). Crime script analysis of drug manufacturing in clandestine laboratories: Implications for prevention. British Journal of Criminology, 51(2), 355–374. https://doi.org/10.1093/bjc/azr005

Ciuffreda, K. J., & Kapoor, N. (2012). Acquired brain injury. In M. Taub, M. Bartuccio & D. Maino (Eds.), Visual diagnosis and care of the patient with special needs (pp. 95–100).Wolters Kluwer Health/Lippincott Williams & Wilkins.

Cliffe, C., Miele, M., & Reid, S. (2019). Homicide in pregnant and postpartum women worldwide: A review of the literature. Journal of Public Health Policy, 40, 180–216. https://doi.org/10.1057/s41271-018-0150-z

Coomber, K., Mayshak, R., Liknaitzky, P., Curtis, A., Walker, A., Hyder, S., & Miller, P. (2021). The role of illicit drug use in family and domestic violence in Australia. Journal of Interpersonal violence, 36(15–16), NP8247–NP8267. https://doi.org/10.1177/0886260519843288

Copp, J. E., Giordano, P. C., Longmore, M. A., & Manning, W. D. (2019). The development of attitudes toward intimate partner violence: An examination of key correlates among a sample of young adults. Journal of Interpersonal Violence, 34(7), 1357–1387. https://doi.org/10.1177/0886260516651311

Cornish, D. B. (1994). The procedural analysis of offending and its relevance for situational prevention. In R. V. Clarke (Ed.), Crime Prevention Studies (Vol. 3; pp. 151–196). https://popcenter.asu.edu/content/crime-prevention-studies-volume-3-volume-3

Corradi, C., & Stöckl, H. (2014). Intimate partner homicide in 10 European countries: Statistical data and policy development in a cross-national perspective. European Journal of Criminology, 11(5), 601–618. https://doi.org/10.1177/1477370814539438

Cox, E., Leung, R., Baksheev, G., Day, A., Toumbourou, J. W., Miller, P., Kremer, P., & Walker, A. (2016). Violence prevention and intervention programmes for adolescents in Australia: A systematic review. Australian Psychologist, 51(3), 206–222. https://doi.org/10.1111/ap.12168

Cullen, P., Vaughan, G., Li, Z., Price, J., Yu, D., & Sullivan, E. (2019). Counting dead women in Australia: An in-depth case review of femicide. Journal of Family Violence, 34(1), 1–8. https://doi.org/10.1007/s10896-018-9963-6

Cunneen, C., & Libesman, T. (2000). Postcolonial trauma: The contemporary removal of Indigenous children and young people from their families in Australia. Australian Journal of Social Issues, 35(2), 99–115. https://doi.org/10.1002/j.1839-4655.2000.tb01088.x

Curtis, A., Booth, B., Gruenert, S., Long, C. M., Karantzas, G., Harries, T., Mullins, E., & Miller, P. G. (2021). Identified support needs for intimate partner violence engagement in an alcohol and other drug treatment sample. Journal of Substance Use, 1–8. https://doi.org/10.1080/14659891.2020.1867660

Curtis, A., Vandenberg, B., Mayshak, R., Coomber, K., Hyder, S., Walker, A., Liknaitzky, P., & Miller, P. G. (2019). Alcohol use in family, domestic and other violence: Findings from a cross‐sectional survey of the Australian population. Drug and Alcohol Review, 38(4), 349–358. https://doi.org/10.1111/dar.12925

Cussen, T., & Bryant, W. (2015). Indigenous and non-Indigenous homicide. Research in Practice, no. 37. Australian Institute of Criminology.

Daly, M., & Wilson, M. (1988). Homicide. Aldine.

Daniels, J. W., & Murphy, C. M. (1997). Stages and processes of change in batterers’ treatment. Cognitive and Behavioral Practice, 4(1), 123–145. https://doi.org/10.1016/S1077-7229(97)80015-6

David, R., & Jaffe, P. (2021). Pre-migration trauma and post-migration stress associated with immigrant perpetrators of domestic homicide. Journal of Family Violence, 36, 551–561. https://doi.org/10.1007/s10896-021-00259-4

Dawson, M., Bunge, V. P., & Balde, T. (2009). National trends in intimate partner homicides: Explaining declines in Canada, 1976 to 2001. Violence Against Women, 15(3), 276–306. https://doi.org/10.1177/1077801208330433

Day, A., Davey, L., Wanganeen, R., Casey, S., Howells, K., & Nakata, M. (2008). Symptoms of trauma, perceptions of discrimination, and anger: A comparison between Australian Indigenous and non-Indigenous prisoners. Journal of Interpersonal Violence, 23(2), 245–258. https://doi.org/10.1177/0886260507309343

De Koker, P., Mathews, C., Zuch, M., Bastien, S., & Mason-Jones, A. J. (2014). A systematic review of interventions for preventing adolescent intimate partner violence. Journal of Adolescent Health, 54(1), 3–13. https://doi.org/10.1016/j.jadohealth.2013.08.008

Department of Families, Housing, Community Services and Indigenous Affairs. (2011). An outline of national standards for out-of-home care: A priority project under the National Framework for Protecting Australia’s Children 2009–2020. Department of Families, Housing, Community Services and Indigenous Affairs.

DeVeaux, M. I. (2013). The trauma of the incarceration experience. Harvard Civil Rights–Civil Liberties Law Review, 48(1), 257. https://harvardcrcl.org/volumes-40-through-present-2/

Dixon, L., & Graham-Kevan, N. (2011). Understanding the nature and etiology of intimate partner violence and implications for practice and policy. Clinical Psychology Review, 31(7), 1145–1155. https://doi.org/10.1016/j.cpr.2011.07.001

Dixon, L., Hamilton-Giachritsis, C., & Browne, K. (2008). Classifying partner femicide. Journal of Interpersonal Violence, 23(1), 74–93. https://doi.org/10.1177/0886260507307652

Dobash, R. E., & Dobash, R. P. (2011). What were they thinking? Men who murder an intimate partner. Violence Against Women, 17(1), 111–134. https://doi.org/10.1177/1077801210391219

Dobash, R. E., & Dobash, R. P. (2015). When men murder women. Oxford University Press.

Dobash, R. E., Dobash, R. P., Cavanagh, K., & Lewis, R. (2004). Not an ordinary killer – Just an ordinary guy: When men murder an intimate woman partner. Violence Against Women, 10(6), 577–605. https://doi.org/10.1177/1077801204265015

Dobash, R. E., Dobash, R. P., Cavanagh, K., & Medina-Ariza, J. (2007). Lethal and nonlethal violence against an intimate female partner: Comparing male murderers to nonlethal abusers. Violence Against Women, 13(4), 329–353. https://doi.org/10.1177/1077801207299204

Doumas, D. M., Pearson, C. L., Elgin, J. E., & McKinley, L. L. (2008). Adult attachment as a risk factor for intimate partner violence: The “mispairing” of partners’ attachment styles. Journal of Interpersonal Violence, 23(5), 616–634. https://doi.org/10.1177/0886260507313526

Dowling, C., Morgan, A., Hulme, S., Manning, M., & Wong, G. (2018). Protection orders for domestic violence: A systematic review. Trends and Issues in Crime and Criminal Justice, no. 551. Australian Institute of Criminology. https://www.aic.gov.au/publications/tandi/tandi551

Doyle, M. F., Shakeshaft, A., Guthrie, J., Snijder, M., & Butler, T. (2019). A systematic review of evaluations of prison-based alcohol and other drug use behavioural treatment for men. Australian and New Zealand Journal of Public Health, 43(2), 120–130. https://doi.org/10.1111/1753-6405.12884

Dragiewicz, M., Burgess, J., Matamoros-Fernández, A., Salter, M., Suzor, N. P., Woodlock, D., & Harris, B. (2018). Technology-facilitated coercive control: Domestic violence and the competing roles of digital media platforms. Feminist Media Studies, 18(4), 609–625. https://doi.org/10.1080/14680777.2018.1447341

Driessen, M., Schroeder, T., Widmann, B., von Schönfeld, C. E., & Schneider, F. (2006). Childhood trauma, psychiatric disorders, and criminal behavior in prisoners in Germany: A comparative study in incarcerated women and men. Journal of Clinical Psychiatry, 67(10). https://doi.org/10.4088/jcp.v67n1001

Dugan, L., Nagin, D. S., & Rosenfeld, R. (1999). Explaining the decline in intimate partner homicide: The effects of changing domesticity, women’s status, and domestic violence resources. Homicide Studies, 3(3), 187–214. https://doi.org/10.1177/1088767999003003001

Dugan, L., Nagin, D. S., & Rosenfeld, R. (2003). Exposure reduction or retaliation? The effects of domestic violence resources on intimate-partner homicide. Law & Society Review, 37(1), 169–198. https://doi.org/10.1111/1540-5893.3701005

Edwards, D. (2018). Childhood sexual abuse and brain development: A discussion of associated structural changes and negative psychological outcomes. Child Abuse Review, 27(3), 198–208. https://doi.org/10.1002/car.2514

Eggers Del Campo, I., & Steinert, J. I. (2020). The effect of female economic empowerment interventions on the risk of intimate partner violence: A systematic review and meta-analysis. Trauma, Violence, & Abuse. https://doi.org/10.1177/1524838020976088

Elbogen, E., & Johnson, S. (2009). The intricate link between violence and mental disorder: Results from the national epidemiologic survey on alcohol and related conditions. Archives of General Psychiatry, 66(2), 152–161. https://doi.org/10.1001/archgenpsychiatry.2008.537

Elbogen, E., Wagner, R., Fuller, S., Calhoun, P., Kinneer, P., & Beckham, J. (2010). Correlates of anger and hostility in Iraq and Afghanistan war veterans. American Journal of Psychiatry, 167(9), 1051–1058. https://doi.org/10.1176/appi.ajp.2010.09050739

Elisha, E., Idisis, Y., Timor, U., & Addad, M. (2010). Typology of intimate partner homicide: Personal, interpersonal, and environmental characteristics of men who murdered their female intimate partner. International Journal of Offender Therapy and Comparative Criminology, 54(4), 494–516. https://doi.org/10.1177/0306624X09338379

Ellis, D. (2017). Marital separation and lethal male partner violence. Violence Against Women, 23(4). 503–519. https://doi.org/10.1177/1077801216644985

Erez, E., Adelman, M., & Gregory, C. (2009). Intersection of immigration and domestic violence: Voices of battered immigrant women. Feminist Criminology, 4(1). https://doi.org/10.1177/1557085108325413

Eriksson L., & Mazerolle, P. (2013). A general strain theory of intimate partner homicide. Aggression and Violent Behaviour, 18(5), 462–470. https://doi.org/10.1016/j.avb.2013.07.002

Eriksson, L., & Mazerolle, P. (2015). A cycle of violence? Examining family-of-origin violence, attitudes, and intimate partner violence perpetration. Journal of Interpersonal Violence, 30(6), 945–964. https://doi.org/10.1177/0886260514539759

Eriksson, L., Mazerolle, P., Wortley, R., Johnson, H., & McPhedran, S. (2018). The offending histories of homicide offenders: Are men who kill intimate partners distinct from men who kill other men? Psychology of Violence, 9(4), 471–480. https://doi.org/10.1037/vio0000214

Esaki, N., Benamati, J., Yanosy, S., Middleton, J. S., Hopson, L. M., Hummer, V. L., & Bloom, S. L. (2013). The sanctuary model: Theoretical framework. Families in Society: The Journal of Contemporary Social Services, 94(2), 87–95. https://doi.org/10.1606/1044-3894.4287

Exum, L. (2006). Alcohol and aggression: An integration of findings from experimental studies. Journal of Criminal Justice, 34(2), 131–145. https://doi.org/10.1016/j.jcrimjus.2006.01.008

Facer-Irwin, E., Blackwood, N. J., Bird, A., Dickson, H., McGlade, D., Alves-Costa, F., & MacManus, D. (2019). PTSD in prison settings: A systematic review and meta-analysis of comorbid mental disorders and problematic behaviours. PLoS One, 14(9), e0222407. https://doi.org/10.1371/journal.pone.0222407

Fals-Stewart, W. (2003). The occurrence of partner physical aggression on days of alcohol consumption: A longitudinal diary study. Journal of Consulting and Clinical Psychology, 71(1), 41. https://doi.org/10.1037/0022-006X.71.1.41

Fazel, S., & Danesh, J. (2002). Serious mental disorder in 23,000 prisoners: A systematic review of 62 surveys. Lancet, 359(9306), 545–550. https://doi.org/10.1016/S0140-6736(02)07740-1

Fazel, S., Smith, E. N., Chang, Z., & Geddes, J. R. (2018). Risk factors for interpersonal violence: An umbrella review of meta-analyses. British Journal of Psychiatry, 213(4), 609–614. https://doi.org/10.1192/bjp.2018.145

Fazel, S., Wolf, A., Chang, Z., Larsson, H., Goodwin, G. M., & Lichtenstein, P. (2015). Depression and violence: A Swedish population study. Lancet Psychiatry, 2(3), 224–232. https://doi.org/10.1016/S2215-0366(14)00128-X

Federal Bureau of Investigation. (n.d.). Expanded homicide data table 10: Murder circumstances, by relationship, 2018. https://ucr.fbi.gov/crime-in-the-u.s/2018/crime-in-the-u.s.-2018/tables/expanded-homicide-data-table-10.xls

Felson, R. B., & Lane, K. J. (2010). Does violence involving women and intimate partners have a special etiology? Criminology, 48(1), 321–338. https://doi.org/10.1111/j.1745-9125.2010.00186.x

Felson, R. B., Messner, S. F., Hoskin, A. W., & Deane, G. (2002). Reasons for reporting and not reporting domestic violence to the police. Criminology, 40(3), 617–647. https://doi.org/10.1111/j.1745-9125.2002.tb00968.x

Ferguson, C. J., & Dyck, D. (2012). Paradigm change in aggression research: The time has come to retire the General Aggression Model. Aggression and Violent Behavior, 17(3), 220–228. https://doi.org/10.1016/j.avb.2012.02.007

Flood, M., & Webster, K. (2007). Preventing violence before it occurs: A framework and background paper to guide the primary prevention of violence against women in Victoria. VicHealth.

Fox, J. A., & Zawitz, M. W. (2007). Homicide trends in the United States. Bureau of Justice Studies. https://www.bjs.gov/content/pub/pdf/htius.pdf

Funston, L., & Herring, S. (2016). When will the stolen generations end?: A qualitative critical exploration of contemporary “child protection” practices in Aboriginal and Torres Strait Islander communities. Sexual Abuse in Australia and New Zealand, ٧(1), 51–58.

Gadd, D., Henderson, J., Radcliffe, P., Stephens-Lewis, D., Johnson, A., & Gilchrist, G. (2019). The dynamics of domestic abuse and drug and alcohol dependency. British Journal of Criminology, 59(5), 1035–1053. https://doi.org/10.1093/bjc/azz011

Gannoni, A., & Cussen, T. (2014). Same-sex intimate partner homicide in Australia. Trends and Issues in Crime and Criminal Justice, no. 469. Australian Institute of Criminology. https://www.aic.gov.au/publications/tandi/tandi469

Garcia, L., Soria, C., & Hurwitz, E. L. (2007). Homicides and intimate partner violence: A literature review. Trauma, Violence, & Abuse, 8(4), 370–383. https://doi.org/10.1177/1524838007307294

Garvey, G., Anderson, K., Gall, A., Butler, T. L., Whop, L. J., Arley, B., Cunningham, J., Dickson, M., Cass, A., & Ratcliffe, J. (2021). The fabric of Aboriginal and Torres Strait Islander wellbeing: A conceptual model. International Journal of Environmental Research and Public Health, 18(15), 7745. https://doi.org/10.3390/ijerph18157745

Gates, M. L., Hunter, E. G., Dicks, V., Jessa, P. N., Walker, V., & Yoo, W. (2018). Multimorbidity patterns and associations with functional limitations among an aging population in prison. Archives of Gerontology and Geriatrics, 77, 115–123. https://doi.org/10.1016/j.archger.2018.03.012

Gay, L. E., Harding, H. G., Jackson, J. L., Burns, E. E., & Baker, B. D. (2013). Attachment style and early maladaptive schemas as mediators of the relationship between childhood emotional abuse and intimate partner violence. Journal of Aggression, Maltreatment & Trauma, 22(4), 408–424. https://doi.org/10.1080/10926771.2013.775982

Gee, G., Dudgeon, P., Schultz, C., Hart, A., & Kelly, K. (2014). Aboriginal and Torres Strait Islander social and emotional wellbeing. In P. Dudgeon, H. Milroy, & R. Walker (Eds.), Working together: Aboriginal and Torres Strait Islander mental health and wellbeing principles and practice (Vol. 2; pp. 55–68). Australian Government Department of Prime Minister and Cabinet.

Gelles, R. J. (1991). Physical violence, child abuse, and child homicide. Human Nature, 2, 59–72. https://doi.org/10.1007/BF02692181

Giancola, P. R., Godlaski, A., & Parrott, D. J. (2005). “So I can’t blame the booze?”: Dispositional aggressivity negates the moderating effects of expectancies on alcohol related aggression. Journal of Studies on Alcohol, 66(6), 815–824. https://doi.org/10.15288/jsa.2005.66.815

Gilchrist, G., Dennis, F., Radcliffe, P., Henderson, J., Howard, L. M., & Gadd, D. (2019). The interplay between substance use and intimate partner violence perpetration: A meta-ethnography. International Journal of Drug Policy, 65, 8–23. https://doi.org/10.1016/j.drugpo.2018.12.009

Gilchrist, G., & Hegarty, K. (2017). Tailored integrated interventions for intimate partner violence and substance use are urgently needed. Drug and Alcohol Review, 36(1), 3–6. https://doi.org/10.1111/dar.12526

Gillespie, L. K., & Reckdenwald, A. (2017). Gender equality, place, and female-victim intimate partner homicide: A county-level analysis in North Carolina. Feminist Criminology, 12(2), 171–191. https://doi.org/10.1177/1557085115620479

Gillespie, S. M., & Beech, A. R. (2016). Theories of emotion regulation.  In D. P. Boer, A. R. Beech, T. Ward, L. A. Craig, M. Rettenberger, L. E. Marshall, & W. L. Marshall (Eds.), The Wiley handbook on the theories, assessment, and treatment of sexual offending (pp. ٢٤٥–٢٦٣). Wiley Blackwell.

Giordano, P. C., Johnson, W. L., Manning, W. D., Longmore, M. A., & Minter, M. D. (2015). Intimate partner violence in young adulthood: narratives of persistence and desistance. Criminology, 53(3), 330–365. https://doi.org/10.1111/1745-9125.12073

Glass, N., Laughon, K., Campbell, J., Block, C. R., Hanson, G., Sharps, P. W., & Taliaferro, E. (2008). Non-fatal strangulation is an important risk factor for homicide of women. Journal of Emergency Medicine, 35(3), 329–335. https://doi.org/10.1016/j.jemermed.2007.02.065

Glenn, A. L., Johnson, A. K., & Raine, A. (2013). Antisocial personality disorder: A current review. Current Psychiatry Reports, 15(12), 427. https://doi.org/10.1007/s11920-013-0427-7

Glenn, C. R., & Klonsky, E. D. (2009). Emotion dysregulation as a core feature of borderline personality disorder. Journal of Personality Disorders, 23(1), 20–28. https://doi.org/10.1521/pedi.2009.23.1.20

Golden, S. D., Perreira, K. M., & Durrance, C. P. (2013). Troubled times, troubled relationships: How economic resources, gender beliefs, and neighborhood disadvantage influence intimate partner violence. Journal of Interpersonal Violence, 28(10), 2134–2155. https://doi.org/10.1177/0886260512471083

Graham, K., Leonard, K. E., Room, R., Wild, T. C., Pihl, R. O., Bois, C., & Single, E. (1998). Current directions in research on understanding and preventing intoxicated aggression. Addiction, 93(5), 659–676. https://doi.org/10.1046/j.1360-0443.1998.9356593.x

Gratz, K. L., & Roemer, L. (2004). Multidimensional assessment of emotion regulation and dysregulation: Development, factors structure, and initial validation of the Difficulties in Emotion Regulation Scale. Journal of Psychopathology and Behavioral Assessment, 26, 41–54. https://doi.org/10.1023/B:JOBA.0000007455.08539.94

Green, S. (2001). Aboriginal and Torres Strait Islander Women’s Taskforce on Violence Report, 2000. Aboriginal and Islander Health Worker Journal, 25(1).

Grigorian, H. L., Brem, M. J., Garner, A., Florimbio, A. R., Wolford-Clevenger, C., & Stuart, G. L. (2020). Alcohol use and problems as a potential mediator of the relationship between emotion dysregulation and IPV perpetration. Psychology of Violence, 10(1), 91–99. https://doi.org/10.1037/vio0000237

Gross, J. J., & Thompson, R. A. (2007). Emotion regulation: Conceptual foundations. In J. J. Gross (Ed.), Handbook of emotion regulation (Vol. 3; p. 24). Guildford Press.

Gupta, J., Acevedo-Garcia, D., Hemenway, D., Decker, M. R., Raj, A., & Silverman, J. G. (2010). Intimate partner violence perpetration, immigration status, and disparities in a community health center-based sample of men. Public Health Reports, 125(1). https://doi.org/10.1177%2F003335491012500111

Guruge, S., Al Jamal, A., Yercich, S., Dhillon, M., Rossiter, K., David, R., & Kulasinghe, M. (2020). Domestic homicide in immigrant communities: Lessons learned. In P. Jaffe, K. Scott, & A.-L. Straatman (Eds.), Preventing domestic homicides (pp. 111–135). Academic Press.

Hall, M. A., & Wright, R. F. (2008). Systematic content analysis of judicial opinions. California Law Review, ٩٦(1), 63–122.

Hand, K., Baxter, J., Carroll, M., & Budinski, M. (2020). Families in Australia Survey: Life during COVID-19. Australian Institute of Family Studies. https://aifs.gov.au/media-releases/new-report-reveals-how-aussie-families-are-adjusting-during-covid-19

Harden, J., Du, J., Spencer, C. M., & Stith, S. M. (2019). Examining attempted and completed intimate partner homicide: A qualitative synthesis. Violence and Victims, 34(6), 869–888. https://doi.org/10.1891/0886-6708

Hasin, D., Van Rossem, R., McCloud, S., & Endicott, J. (1997). Alcohol dependence and abuse diagnoses: Validity in community sample heavy drinkers. Alcoholism: Clinical and Experimental Research, 21(2), 213–219. https://doi.org/10.1111/j.1530-0277.1997.tb03752.x

Heffernan, E., Andersen, K., Davidson, F., & Kinner, S. A. (2015). PTSD among Aboriginal and Torres Strait Islander people in custody in Australia: Prevalence and correlates. Journal of Traumatic Stress, 28(6), 523–530. https://doi.org/10.1002/jts.22051

Heise, L. L. (1998). Violence against women: An integrated, ecological framework. Violence against Women, 4(3), 262–290.

Henderson, A. J., Bartholomew, K., Trinke, S. J., & Kwong, M. J. (2005). When loving means hurting: An exploration of attachment and intimate abuse in a community sample. Journal of Family Violence, 20(4), 219. https://doi.org/10.1007/s10896-005-5985-y

Holtzworth-Munroe, A., & Stuart, G. L. (1994). Typologies of male batterers: Three subtypes and the differences among them. Psychological Bulletin, 116(3), 476–497. https://doi.org/10.1037/0033-2909.116.3.476

Holzer, P., Higgins, J. R., Bromfield, L., & Higgins, D. (2006). The effectiveness of parent education and home visiting child maltreatment prevention programs (Child Abuse Prevention Issues no. 24). Australian Institute of Family Studies. http://www.aifs.gov.au/nch/pubs/issues/issues24/issues24.html

Howard, R. (2015). Personality disorders and violence: What is the link? Borderline Personality Disorder and Emotional Dysregulation, 2(12). https://doi.org/10.1186/s40479-015-0033-x

Hughes, C., Bolis, M., Fries, R., & Finigan, S. (2015). Women’s economic inequality and domestic violence: Exploring the links and empowering women. Gender & Development, 23(2), 279–297. https://doi.org/10.1080/13552074.2015.1053216

Hulme, S., Morgan, A., & Boxall, H. (2019). Domestic violence offenders, prior offending and reoffending in Australia. Trends & Issues in Crime and Criminal Justice, no. 580. Australian Institute of Criminology. https://aic.gov.au/publications/tandi/tandi580

Jackson, H. J., & Burgess, P. M. (2000). Personality disorders in the community: A report from the Australian National Survey of Mental Health and Wellbeing. Social Psychiatry and Psychiatric Epidemiology, ٣٥(12), 531–538. https://doi.org/10.1007/s001270050276

Jacobson, N. S. & Gottman, J. M. (1994). When men batter women: New insights into ending abusive relationships. Simon & Schuster.

James, D. V., Farnham, F. R., & Wilson, S. P. (2013). The fixated threat assessment centres: The joint policing and psychiatric approach to risk assessment and management in cases of public figure threat and lone actor grievance-fueled violence. In J. Reid Meloy & J. Hoffman (Eds.), International handbook of threat assessment (Vol. 299; p. 471). Oxford University Press. https://doi.org/10.1093/med-psych/9780190940164.003.0027

Jamieson, R., & Grounds, A. (2005). Release and adjustment: Perspectives from studies of wrongly convicted and politically motivated prisoners. The Effects of Imprisonment, 33–65.

Jin, X., & Keat, J. E. (2010). The effects of change in spousal power on intimate partner violence among Chinese immigrants. Journal of Interpersonal Violence, 25(4), 610–625. https://doi.org/10.1177/0886260509334283

Johnson, H., Eriksson, L., Mazerolle, P., & Wortley, R. (2017). Intimate femicide: The role of coercive control. Feminist Criminology, 14(1), 1–21. Https://doi.org/10.1177/1557085117701574

Johnson, M. P. (1995). Patriarchal terrorism and common couple violence: Two forms of violence against women. Journal of Marriage and the Family, 57, 283–294.

Johnson, M.P. (2006). Conflict and control: Gender symmetry and asymmetry in domestic violence. Violence Against Women, 12(11), 1003–1018. https://doi.org/10.1177/1077801206293328

Juodis, M., Starzomski, A., Porter, S., & Woodworth, M. (2014). A comparison of domestic and non-domestic homicides: Further evidence for distinct dynamics and heterogeneity of domestic homicide perpetrators. Journal of Family Violence, 29(3), 299–313.

Karakurt, G., Koç, E., Çetinsaya, E. E., Ayluçtarhan, Z., & Bolen, S. (2019). Meta-analysis and systematic review for the treatment of perpetrators of intimate partner violence. Neuroscience & Biobehavioral Reviews, 105, 220–230. https://doi.org/10.1016/j.neubiorev.2019.08.006

Keatley, D. (2018). Pathways in crime: An introduction to behaviour sequence analysis. Springer.

Keyes, K. M., Pratt, C., Galea, S., McLaughlin, K. A., Koenen, K. C., & Shear, M. K. (2014). The burden of loss: Unexpected death of a loved one and psychiatric disorders across the life course in a national study. American Journal of Psychiatry, 171(8), 864–871. https://doi.org/10.1176/appi.ajp.2014.13081132

Kirmayer, L. J., Brass, G. M., & Tait, C. L. (2000). The mental health of Aboriginal peoples: Transformations of identity and community. Canadian Journal of Psychiatry, 45(7), 607–616.

Kivisto, A. J. (2015). Male perpetrators of intimate partner homicide: A review and proposed typology. Journal of the American Academy of Psychiatry and the Law online, 43(3), 300–312.

Kovalenko, A. G., Abraham, C., Graham-Rowe, E., Levine, M., & O’Dwyer, S. (2020). What works in violence prevention among young people?: A systematic review of reviews. Trauma, Violence, & Abuse. https://doi.org/10.1177/1524838020939130

Kwan, J., Sparrow, K., Facer-Irwin, E., Thandi, G., Fear, N. T., & MacManus, D. (2020). Prevalence of intimate partner violence perpetration among military populations: A systematic review and meta-analysis. Aggression and Violent Behavior, 53. https://doi.org/10.1016/j.avb.2020.101419

Lamont, E. (2014). Negotiating courtship: Reconciling egalitarian ideals with traditional gender norms. Gender & Society, 28(2), 189–211. https://doi.org/10.1177/0891243213503899

Laslett, A. M., Mugavin, J., Jiang, H., Manton, E., Callinan, S., MacLean, S., & Room, R. (2015). The hidden harm: Alcohol’s impact on children and families. Foundation for Alcohol Research and Education.

Lawler, S., Stapinski, L., Prior, K., Basto-Periera, M., Newton, N., Teesson, M., & Barrett, E. (2021). Unpacking violence in young adulthood: The relative importance of hazardous alcohol use. Journal of Interpersonal Violence. https://doi.org/10.1177/08862605211044103

Leeies, M., Pagura, J., Sareen, J., & Bolton, J. M. (2010). The use of alcohol and drugs to self‐medicate symptoms of posttraumatic stress disorder. Depression and Anxiety, 27(8), 731–736.

Liem, M., & Koenraadt, F. (2008). Familicide: A comparison with spousal and child homicide by mentally disordered perpetrators. Criminal Behaviour and Mental Health, 18, 306–318. https://doi.org/10.1002/cbm.710

Liem, M., & Koenraadt, F. (2018). Domestic homicide: Patterns and dynamics. Routledge.

López-Castro, T., Smith, K. Z., Nicholson, R. A., Armas, A., & Hien, D. A. (2019). Does a history of violent offending impact treatment response for comorbid PTSD and substance use disorders? A secondary analysis of a randomized controlled trial. Journal of Substance Abuse Treatment, 97, 47–58.

Lyneham, S., & Richards, K. (2014). Human trafficking involving marriage and partner migration to Australia. Research and Public Policy Series, no. 124. Australian Institute of Criminology. https://www.aic.gov.au/publications/rpp/rpp124

Lysell, H., Dahlin, M., Långström, N., Lichtenstein, P., & Runeson, B. (2016). Killing the mother of one’s child: Psychiatric risk factors among male perpetrators and offspring health consequences. Journal of Clinical Psychiatry, 77(3), 342–347. https://doi.org/10.4088/JCP.14m09564

Mahalik, J. R., Aldarondo, E., Gilbert-Gokhale, S., & Shore, E. (2005). The role of insecure attachment and gender role stress in predicting controlling behaviors in men who batter. Journal of Interpersonal Violence, 20(5), 617–631. https://doi.org/10.1177/0886260504269688

Marín-Morales, A., Pérez-García, M., Catena-Martínez, A., & Verdejo-Román, J. (2021). Lower brain volume and poorer emotional regulation in partner coercive men and other offenders. Psychology of Violence. https://doi.org/10.1037/vio0000393

Markwick, A., Ansari, Z., Clinch, D., & McNeil, J. (2019). Experiences of racism among Aboriginal and Torres Strait Islander adults living in the Australian state of Victoria: A cross-sectional population-based study. BMC Public Health, 19, 309. https://doi.org/10.1186/s12889-019-6614-7

Marshall, A. D., Panuzio, J., & Taft, C. T. (2005). Intimate partner violence among military veterans and active duty servicemen. Clinical Psychology Review, 25(7), 862–876.

Matias, A., Gonçalves, M., Soeiro, C., & Matos, M. (2019). Intimate partner homicide: A meta-analysis of risk factors. Aggression and Violent Behaviours, 50, 1–12. https://doi.org/10.1016/j.avb.2019.101358

Mayshak, R., Cox, E., Costa, B., Walker, A., Hyder, S., Day, A., Coomber, K., Taylor, N., & Miller, P. (2018). Alcohol/drug-involved family violence in Australia (ADIVA). NDLERF Research Bulletin, no. 7. Australian Institute of Criminology. https://www.aic.gov.au/publications/ndlerfbulletin/ndlerfbulletin7

Mayshak, R., Curtis, A., Coomber, K., Tonner, L., Walker, A., Hyder, S., Liknaitzky, P., & Miller, P. (2020). Alcohol-involved family and domestic violence reported to police in Australia. Journal of Interpersonal Violence. https://doi.org/10.1177/0886260520928633

McCarthy, K. J., Mehta, R., & Haberland, N. A. (2018). Gender, power, and violence: A systematic review of measures and their association with male perpetration of IPV. PLoS One, 13(11), e0207091. https://doi.org/10.1371/journal.pone.0207091

McDermott, R. C., & Lopez, F. G. (2013). College men’s intimate partner violence attitudes: Contributions of adult attachment and gender role stress. Journal of Counseling Psychology, 60(1), 127. https://doi.org/10.1037/a0030353

McFarlane, J. M., Campbell, J. C., Wilt, S., Sachs, C. J., Ulrich, Y., & Xu, X. (1999). Stalking and intimate partner femicide. Homicide Studies, 3(4), 300–316. https://doi.org/10.1177/1088767999003004003

McMahon, S., & Dick, A. (2011). “Being in a room with like-minded men”: An exploratory study of men’s participation in a bystander intervention program to prevent intimate partner violence. Journal of Men’s Studies, 19(1), 3–18. https://doi.org/10.3149/jms.1901.3

McMahon, M., & McGorrery, P. (2016). Criminalising emotional abuse, intimidation and economic abuse in the context of family violence: The Tasmanian experience. University of Tasmania Law Review, 35, 1.

McMahon, M., & McGorrery, P. (2020) Criminalising coercive control: An introduction. In M. McMahon & P. McGorrery (Eds.), Criminalising coercive control. Springer. https://doi.org/10.1007/978-981-15-0653-6_1

McPhedran, S., Eriksson, L., Mazerolle, P., & Johnson, H. (2018). Victim-focussed studies of intimate partner femicide: A critique of methodological challenges and limitations in current research. Aggression and Violent Behavior, 39, 61–66. https://doi.org/10.1016/j.avb.2018.02.005

Menzies, P. (2010). Intergenerational trauma from a mental health perspective. Native Social Work Journal, 7, 63–85.

Monckton Smith, J. (2019). Intimate partner femicide: Using Foucauldian analysis to track an eight stage progression to homicide. Violence against Women, 26(11), 1267–1285.

Myhill, A. (2015). Measuring coercive control: What can we learn from national population surveys? Violence Against Women, 21(3), 355–375. https://doi.org/10.1177/1077801214568032

Nancarrow, H. (2019). Unintended consequences of domestic violence law: Gendered aspirations and racialised realities. Springer Nature.

Nelson, S. E., & Wilson, K. (2017). The mental health of Indigenous peoples in Canada: A critical review of research. Social Science & Medicine, 176, 93–112.

New Zealand Police. (2020). Historic murder offences in New Zealand, 1926–2017 [Data table]. https://www.police.govt.nz/sites/default/files/publications/historic-new-zealand-murder-rates.pdf

Newton, B. (2020). Aboriginal parents’ experiences of having their children removed by statutory child protection services. Child & Family Social Work, ٢٥(4), 814–822. https://doi.org/10.1111/cfs.12759

Nowakowski-Sims, E. (2019). An exploratory study of childhood adversity and delinquency among youth in the context of child-to-parent and sibling-to-sibling violence. Journal of Family Social Work, 22(2), 126–145.

Nguyen, H., Haeney, O., & Galletly, C. (2021). The characteristics of older homicide offenders: A systematic review. Psychiatry, Psychology and Law, 1–18.

Nydegger, L. A., DiFranceisco, W., Quinn, K., & Dickson-Gomez, J. (2017). Gender norms and age-disparate sexual relationships as predictors of intimate partner violence, sexual violence, and risky sex among adolescent gang members. Journal of Urban Health, 94(2), 266–275. https://doi.org/10.1007/s11524-016-0068-3

Office for National Statistics. (2021a). Appendix tables: Homicide in England and Wales [Dataset]. https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/datasets/appendixtableshomicideinenglandandwales

Office for National Statistics. (2021b). Estimates of the population for the UK, England and Wales, Scotland and Northern Ireland [Dataset]. https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationestimates/datasets/populationestimatesforukenglandandwalesscotlandandnorthernireland

Oliver, C. L., & Jaffe, P. G. (2021). Comorbid depression and substance abuse in domestic homicide: Missed opportunities in the assessment and management of mental illness in perpetrators. Journal of Interpersonal Violence, 36(11–12), NP6441–NP6464.

Oram, S., Flynn, S. M., Shaw, J., Appleby, L., & Howard, L. M. (2013). Mental illness and domestic homicide: A population-based descriptive study. Psychiatric Services, 64(10), 10061011.

Papalia, N., Spivak, B., Daffern, M., & Ogloff, J. R. (2019). A meta-analytic review of the efficacy of psychological treatments for violent offenders in correctional and forensic mental health settings. Clinical Psychology: Science and Practice, 26(2), e12282.

Paradies, Y. (2006). A systematic review of empirical research on self-reported racism and health. International Journal of Epidemiology, 35(4), 888–901.

Piper, A., & Berle, D. (2019). The association between trauma experienced during incarceration and PTSD outcomes: A systematic review and meta-analysis. Journal of Forensic Psychiatry & Psychology, 30(5), 854–875. https://doi.org/10.1080/14789949.2019.1639788

Polk, K. (1994). When men kill: Scenarios of masculine violence. Cambridge University Press.

Poole, C., & Rietschlin, J. (2012). Intimate partner victimization among adults aged 60 and older: An analysis of the 1999 and 2004 General Social Survey. Journal of Elder Abuse & Neglect, 24(2), 120–137.

Potas, I. (1991, October 29–31). The principles of sentencing violent offenders: Towards a more structured approach [Conference presentation]. Serious Violent Offenders: Sentencing, Psychiatry and Law Reform, Melbourne, Victoria, Australia.

Reavis, J.A., Looman, J., Franco, K.A., & Rojas, B. (2013). Adverse childhood experiences and adult criminality: How long must we live before we possess our own lives? Permanente Journal, 17(2), 44.

Reckdenwald, A., & Parker, K.F. (2012). Understanding the change in male and female intimate partner homicide over time: A policy- and theory-relevant investigation. Feminist Criminology, 7(3), 167–195. https://doi.org/10.1177/1557085111428445

Reidy, D. E., Berke, D. S., Gentile, B., & Zeichner, A. (2014). Man enough? Masculine discrepancy stress and intimate partner violence. Personality and Individual Differences, 68, 160–164. https://doi.org/10.1016/j.paid.2014.04.021

Rhodes, L. A. (2005) Pathological effects of the supermaximum prison. American Journal of Public Health, 95(10), 1692. https://doi.org/10.2105/AJPH.2005.070045

Roberts, A. L., McLaughlin, K. A., Conron, K. J., & Koenen, K. C. (2011). Adulthood stressors, history of childhood adversity, and risk of perpetration of intimate partner violence. American Journal of Preventive Medicine, 40(2), 128–138.

Robinson, S. R., Ravi, K., & Voth Schrag, R. J. (2020). A systematic review of barriers to formal help seeking for adult survivors of IPV in the United States, 2005–2019. Trauma, Violence, & Abuse. https://doi.org/10.1177/1524838020916254

Salzman, M. B., & Halloran, M. J. (2004). Cultural trauma and recovery: Cultural meaning, self-esteem, and the reconstruction of the cultural anxiety buffer. In J. Greenberg, S. L. Koole, & T. Pyszczynski (Eds.), Handbook of experimental existential psychology (pp. ٢٣١–٢٤٦). Guilford Press.

Santana, M. C., Raj, A., Decker, M. R., La Marche, A., & Silverman, J. G. (2006). Masculine gender roles associated with increased sexual risk and intimate partner violence perpetration among young adult men. Journal of Urban Health, 83(4), 575–585. https://doi.org/10.1007/s11524-006-9061-6

Sawrikar, P., & Katz, I. (2008). Enhancing family and relationship service accessibility and delivery to culturally and linguistically diverse families in Australia (AFRC Issues No. 3). Australian Institute of Family Studies. https://aifs.gov.au/cfca/publications/enhancing-family-and-relationship-service-accessibility-and-delivery-cultural

Schomerus, G., Lucht, M., Holzinger, A., Matschinger, H., Carta, M. G., & Angermeyer, M. C. (2011). The stigma of alcohol dependence compared with other mental disorders: A review of population studies. Alcohol and Alcoholism, 46(2), 105–112.

Scottish Government. (2019). Homicide in Scotland 2018–2019 [Bulletin]. https://www.gov.scot/publications/homicide-scotland-2018-19/documents/

Sebire, J. (2017). The value of incorporating measures of relationship concordance when constructing profiles of intimate partner homicides: A descriptive study of IPH committed within London, 1998–2009. Journal of Interpersonal Violence, 32(10), 1476–1500. https://doi.org/10.1177/0886260515589565

Segrave, M. (2017). Temporary migration and family violence: An analysis of victimisation, vulnerability and support. School of Social Sciences, Monash University. https://www.monash.edu/arts/gender-and-family-violence/research-and-projects/completed-projects/temporary-migration-and-family-violence

Segrave, M., & Pfitzner, N. (2020). Family violence and temporary visa holders during COVID-19. Monash Gender and Family Violence Prevention Centre, Monash University.

Sesar, K., Dodaj, A., & Šimić, N. (2018). Mental health of perpetrators of intimate partner violence. Mental Health Review Journal, 23(4), 221–239 https://doi.org/10.1108/MHRJ-08-2017-0028

Sheehan, B. E., Murphy, S. B., Moynihan, M. M., Dudley-Fennessey, E., & Stapleton, J. G. (2015). Intimate partner homicide: New insights for understanding lethality and risks. Violence Against Women, 21(2), 269–288. https://doi.org/10.1177/1077801214564687

Shore, J. H., & Spicer, P. (2004). A model for alcohol-mediated violence in an Australian Aboriginal community. Social Science & Medicine, 58(12), 2509–2521. https://doi.org/10.1016/j.socscimed.2003.09.022

Sindicich, N., Mills, K. L., Barrett, E. L., Indig, D., Sunjic, S., Sannibale, C., Rosenfeld, J., & Najavits, L. M. (2014). Offenders as victims: Post-traumatic stress disorder and substance use disorder among male prisoners. Journal of Forensic Psychiatry & Psychology, 25(1), 44–60. https://doi.org/10.1080/14789949.2013.877516

Singh, V., Petersen, K., & Singh, S. R. (2014). Intimate partner violence victimization: Identification and response in primary care. Primary Care: Clinics in Office Practice, 41(2), 261–281.

Skeem, J., Eno Louden, J., Manchak, S., Vidal, S., & Haddad, E. (2009). Social networks and social control of probationers with co-occurring mental and substance abuse problems. Law and Human Behavior, 33(2), 122–135. https://doi.org/10.1007/s10979-008-9140-1

Skott, S. (2019). Disaggregating homicide: Changing trends in subtypes over time. Criminal Justice and Behavior, 46(1), 1650–1668. https://doi.org/10.1177/0093854819858648

Smith, N., & Trimboli, L. (2010). Comorbid substance and non-substance mental health disorders and re-offending among NSW prisoners. Crime and Justice Bulletin, no. 140. NSW Bureau of Crime Statistics and Research. https://search.informit.org/doi/abs/10.3316/agispt.20110202

Spencer, C. M., & Stith, S. M. (2020). Risk factors for male perpetration and female victimization of intimate partner homicide: A meta-analysis. Trauma, Violence, & Abuse, 21(3), 527–540. https://doi.org/10.1177/1524838018781101

Stark, E. (2007). Coercive control: How men entrap women in personal life. Oxford University Press.

State of Victoria. (2014–2016). Royal Commission into Family Violence: Summary and recommendations (Parl Paper No. 132). Royal Commission into Family Violence.

Statistics Canada. (2019). Table 9: Homicides by closest accused to victim relationship and sex, Canada, 2018. https://www150.statcan.gc.ca/n1/pub/85-002-x/2019001/article/00016/tbl/tbl09-eng.htm

Statistics Canada. (2021). Table 17-10-0005-01: Population estimates on July 1st, by age and sex. https://www150.statcan.gc.ca/t1/tbl1/en/cv.action?pid=1710000501

Stats NZ. (n.d.). Table: Estimated resident population by age and sex (1991+) (Annual–Dec) [Dataset]. http://infoshare.stats.govt.nz/SelectVariables.aspx?pxID=578e1fa5-c86b-42a7-adeb-f34f18517323

Stephens-Lewis, D., Johnson, A., Huntley, A., Gilchrist, E., McMurran, M., Henderson, J., Feder, G., Howard, L. M., & Gilchrist, G. (2019). Interventions to reduce intimate partner violence perpetration by men who use substances: A systematic review and meta-analysis of efficacy. Trauma, Violence, & Abuse. https://doi.org/10.1177/1524838019882357

Stith, S. M., Smith, D. B., Penn, C. E., Ward, D. B., & Tritt, D. (2004). Intimate partner physical abuse perpetration and victimization risk factors: A meta-analytic review. Aggression and Violent Behavior, 10(1), 65–98. https://doi.org/10.1016/j.avb.2003.09.001

Stöckl, H., Devries, K., Rotstein, A., Abrahams, N., Campbell, J., Watts, J., & Moreno, C.G. (2013). The global prevalence of intimate partner homicide: A systematic review. Lancet, 382, 859–65.

Stout, K. D. (1993). Intimate femicide: A study of men who have killed their mates. Journal of Offender Rehabilitation, 19(3-4), 81–94.

Stuart, G. L. (2005). Improving violence intervention outcomes by integrating alcohol treatment. Journal of Interpersonal Violence, 20(4), 388–393. https://doi.org/10.1177/0886260504267881

Taft, C. T., Walling, S. M., Howard, J. M., & Monson, C. (2011). Trauma, PTSD, and partner violence in military families. In S. M. Wadsworth & D. Riggs (Eds.), Risk and resilience in US military families (pp. ١٩٥–٢١٢). Springer.

Tomlinson, M., Brown, M. & Hoaken, P. (2016). Recreational drug use and human aggressive behavior: A comprehensive review since 2003. Aggression and Violent Behavior, 27, 9–29.

Tyrer, P., Reed, G. M., & Crawford, M. J. (2015). Classification, assessment, prevalence, and effect of personality disorder. Lancet, 385(9969), 717–726.

United Nations. (2021). Definitions: Refugees and migrants [UN website]. https://refugeesmigrants.un.org/definitions

United States Census Bureau (2021). National population by characteristics: 2010–2019 [Dataset]. https://www.census.gov/data/tables/time-series/demo/popest/2010s-national-detail.html

Valabdass, S. N., Subramaney, U., & Edge, A. (2021). Characteristics of persons accused of intimate partner homicide amongst forensic psychiatric observations.  South African Journal of Psychiatry, ٢٧(9756), 1675. https://doi.org/10.4102/sajpsychiatry.v27i0.1675

Vatnar, S. K. B., & Bjørkly, S. (2008). An interactional perspective of intimate partner violence: An in-depth semi-structured interview of a representative sample of help-seeking women. Journal of Family Violence, 23, 265–279. https://doi.org/10.1007/s10896-007-9150-7

Vatnar, S. K. B., Friestad, C., & Bjørkly, S. (2017). Intimate partner homicide in Norway 1990–2012: Identifying risk factors through structured risk assessment, court documents, and interviews with bereaved. Psychology of Violence, 7(3), 395–405. https://doi.org/10.1037/vio0000100

Vickerman, K. A., & Margolin, G. (2008). Trajectories of physical and emotional marital aggression in midlife couples. Violence and Victims, 23(1), 18–34.

Vignola-Lévesque, C., & Léveillée, S. (2021). Intimate partner violence and intimate partner homicide: Development of a typology based on psychosocial characteristics. Journal of Interpersonal Violence. https://doi.org/10.1177/08862605211021989

Voce, I., & Boxall, H. (2018). Who reports domestic violence to police? A review of the evidence. Trends & Issues in Crime and Criminal Justice, no. 559. Australian Institute of Criminology. https://www.aic.gov.au/publications/tandi/tandi559

Walklate, S., & Fitz-Gibbon, K. (2019). The criminalisation of coercive control: The power of law? International Journal for Crime, Justice and Social Democracy, 8(4), 94.

Walling, S. M., Meehan, J. C., Marshall, A. D., Holtzworth‐Munroe, A., & Taft, C. T. (2012). The relationship of intimate partner aggression to head injury, executive functioning, and intelligence. Journal of marital and family therapy, 38(3), 471–485.

Webster, K., Diemer, K., Honey, N., Mannix, S., Mickle, J., Morgan, J., Parkes, A., Politoff, V., Powell, A., Stubbs, J., & Ward, A. (2018). Australians’ attitudes towards violence against women and gender equality: Findings from the 2017 National Community Attitudes towards Violence against Women Survey (NCAS) (Research Report, 03/2018). ANROWS. https://ncas.anrows.org.au/findings/#target_dl

Whaley, R.B., & Messner, S.F. (2002). Gender equality and gendered homicides. Homicide Studies, 6(3), 188–210.

White, H. R., Conway, F. N., & Ward, J. H. (2019). Comorbidity of substance use and violence. In M. Krohn, N. Hendrix, G. Penly Hall & A. Lizotte (Eds.), Handbook on crime and deviance (pp. 513–532). Springer.

Wilson, I. M., Graham, K., & Taft, A. (2014). Alcohol interventions, alcohol policy and intimate partner violence: A systematic review. BMC Public Health, 14(1), 1–11.

Wilson, M., Johnson, H., & Daly, M. (1995). Lethal and nonlethal violence against wives. Canadian Journal of Criminology, ٣٧(3), 331–361. https://doi.org/10.3138/cjcrim.37.3.331

Wilson, M., & Daly, M. (1993). An evolutionary psychological perspective on male sexual proprietariness and violence against wives. Violence and Victims, 8(3), 271–94.

Wilson, M.I., & Daly, M. (1992). Who kills whom in spouse killings? On the exceptional sex ratio of spousal homicides in the United States. Criminology, 30(2), 189–216.

Workplace Gender Equality Agency. (2021). Australia’s gender pay gap statistics: August 2021. Australian Government. https://www.wgea.gov.au/sites/default/files/documents/Gender_pay_gap_factsheet_august2021.pdf

Yu, R., Nevado-Holgado, A. J., Molero, Y., D’Onofrio, B. M., Larsson, H., Howard, L. M., & Fazel, S. (2019). Mental disorders and intimate partner violence perpetrated by men towards women: A Swedish population-based longitudinal study. PLoS medicine, 16(12), e1002995.

APPENDIX A:
Literature review

The following review summarises the contemporary literature on intimate partner homicide, specifically the male-perpetrated homicide of a female intimate partner. The purpose of the review was to inform the development of the coding framework that in turns informed the data extraction processes undertaken as part of the project.

The literature search was conducted by the Australian Institute of Criminology’s (AIC) J.V. Barry Library in consultation with project team members. Searches were conducted in the following databases: Proquest Criminal Justice Database; EBSCO’s Criminal Justice Abstracts with full text, Psychology and Behavioral Sciences Collection, Violence & Abuse Abstracts, and SocINDEX; PubMed; and the AIC Library catalogue, including CINCH. Broad search terms included femicide or homicide (intimate or partner or domestic). The subsequent terms – review, risk, assessment, characteristics, behaviour and sequence – were added to refine the output to literature on the prevalence, characteristics and risk factors of female intimate partner homicide. The original search was limited to English-language papers published between 2010 and 2020 and based on studies conducted in Australia, the United Kingdom, Canada, the United States, New Zealand and European countries. This search identified 162 records, which were screened for inclusion. This screening process resulted in 89 full-text articles. An additional 49 full-text articles published before 2010 were sought where identified as relevant to the purpose of the review, bringing the total number of articles reviewed to 138.

Trends in intimate partner homicide

Intimate partner homicide (IPH) is the most prevalent type of homicide in Australia and the most prevalent form of homicide victimisation among women. A systematic review of 227 studies of IPH across 66 countries estimated 38.6 per cent of female homicide incidents (i.e. female victimisation from homicide) and 13.5 per cent of all homicide incidents were IPH (Stöckl et al., 2013). Among the 18 high-income country studies included in the review, the female IPH rate was 41.2 per cent of all female homicide incidents and 14.9 per cent of all homicide incidents (Stöckl et al., 2013). IPH data from Australia are comparable. In 2018–19 IPH comprised 21 per cent of all homicide incidents in Australia (Bricknell & Doherty, 2021). Sixteen per cent of all incidents involved the murder (or manslaughter) of a female by a male intimate partner (Bricknell & Doherty, 2021). Female homicide victimisation predominantly occurs within the context of a domestic relationship and is specifically intimate partner-based: over half of female homicide victims in Australia are killed by a male intimate partner (Bricknell, 2020b; Bricknell & Doherty, 2021; Cussen & Bryant, 2015).

There are a small number of studies describing trends in IPH (Block & Christakos, 1995; Caman et al., 2017; Corradi & Stöckl, 2014; Dawson et al., 2009; Fox & Zawitz, 2007; Reckdenwald & Parker, 2012) but these are confounded by the use of different types of data, varied presentation of numbers or rates based on incidents of perpetration or victimisation (assuming male perpetration roughly equates to female victimisation, and vice versa), and different definitions of IPH. For example, Reckdenwald and Parker (2012) included all victims and offenders who were in an intimate relationship, whereas Dawson et al. (2009) included killings between current and former spouses or common-law partners (i.e. de facto) only.

Studies examining trends in IPH in North America describe notable decreases in male IPH victimisation between the 1970s and early to mid-2000s which are not mirrored in female victimisation rates. Rather, rates of female IPH victimisation exhibited successive periods of slow decline or stasis (Block & Christakos, 1995; Dawson et al., 2009; Fox & Zawitz, 2007; Reckdenwald & Parker, 2012). The picture is similarly “mixed” in the United Kingdom and Europe. Corradi and Stöckl (2014) compared IPH numbers in 10 countries but trend data were presented (or available) for just three countries. There was no change in England and Wales, a decrease in IPH in France, and a decrease in IPH in Finland in line with an overall decrease in female homicide victimisation. Conversely, the trend in Sweden for the period 1990–2013 has been a small but significant decrease in rates of male-perpetrated IPH (i.e. largely female victimisation) but no change in rates of female-perpetrated IPH (or male victimisation; Caman et al., 2017).

More recent recorded crime data demonstrate an overall decrease in female IPH victimisation in England and Wales, Canada and the United States and an increase in Scotland and New Zealand over the last decade. The decreases ranged from 10 per cent in the United States to 21 per cent in England and Wales. Larger decreases have occurred in Australia. In 2018–19, the female IPH rate decreased to 0.33 per 100,000, a 31 per cent decrease over the previous decade and a 62 per cent decrease since 1989–90 (Bricknell & Doherty, 2021).

Table A1: Rate of female IPH (per 100,000 women aged 18 years and over)

Country	Time period	Rate
Australiaa	2018–19	0.35
New Zealandb	2017	0.42
England and Walesc	2018–19	0.34
Scotlandd	2018–19	0.39
Canadae	2018	0.44
United Statesf	2018	0.81

^a Bricknell (2020a).
^b New Zealand Police (2020); Stats NZ (n.d.).
^c Office for National Statistics (2021a, 2021b).
^d Scottish Government (2019); Office for National Statistics (2021b).
^e Statistics Canada (2019, 2021)
^f Federal Bureau of Investigation (n.d.); United States Census Bureau (2021).

Irrespective of the trend in female IPH, the proportion of female homicides that are perpetrated by an intimate partner has remained the same – that is, if a woman is murdered, it is likely that the perpetrator will be a male current or former intimate partner. Further, changes in female IPH rates may also be better understood by comparing them with trends in female or other categories of homicide victimisation. Skott (2019), for example, used latent class analysis to compare absolute and relative changes over four time periods for four subtypes of homicide in Scotland: stabbing, bludgeoning, rivalry and domestic. Sub-types were derived based on victim, offender and incident variables with predominantly female intimate partner victims in the domestic subtype. Domestic homicide, like the other sub-categories, followed an absolute decrease (29 per cent) in 2012–15 compared with the index period 2000–03 – that is, there was a decrease in the number of homicides that occurred. However, the relative change varied, with significant relative increases in domestic homicide in 2004–07 and 2012–15 (Skott, 2019) – i.e. there was an increase in the proportion of domestic homicides across the later time periods. These findings compared with minimal change in the proportion of stabbing and bludgeoning homicides but a large decrease in rivalry homicides.

Theoretical frameworks for understanding intimate partner homicide

Several theoretical frameworks have been developed or “borrowed” from different disciplines to explain the occurrence of IPH. Two in particular have dominated the research that has been undertaken to date, and centre male homicidal behaviour towards a female intimate partner in the socio-biological and socio-cultural predominance of males within society, communities and intimate relationships. Within these frameworks, homicide may be viewed as a consequence experienced by the female intimate partner if that predominance is jeopardised. The first of these theories – drawn from evolutionary social psychology – establishes female IPH as a corollary of male sexual proprietariness (Daly & Wilson, 1988; Wilson & Daly, 1992, 1993; Wilson et al., 1995). Male fitness is contingent on exclusive sexual and reproductive rights to the intimate partner. Any perceived or actual threat to those rights incurs sexual jealousy that may manifest itself in violence and homicidal intent. Such proprietariness, expressed as possessiveness and jealousy, was evident in a large number of male-perpetrated IPHs reviewed by Daly and Wilson (1988) and preceded by the female partner’s apparent infidelity or decision to separate from their partner.

Alternatively, feminist theories argue that it is the socio-cultural construct of gender and the diminishment of the female gender within the patriarchal system that engenders risk of IPH. The elevation of one gender over another induces and regulates notions of control and entitlement that define the position and obligations of men and women in intimate relationships. Heightened male entitlement motivates behaviour that coerces and restrains the female partner, and where these methods fail, lethal violence is used as the ultimate demonstration of control. The intransigence of this behaviour is encapsulated in the concept of coercive control, framed by Stark (2007). The predominance of coercive controlling behaviours that can include physical violence, sexual coercion, intimidation, isolation and constraint in abusive intimate partner relationships is widely recognised as integral to the environment that increases the risk of IPH (Campbell et al., 2003; Dobash & Dobash, 2015; Gillespie & Reckdenwald, 2017; Myhill, 2015).

Explanations of trends in IPH have largely focused on gender equality and social, economic and policy changes that have contributed to parity. The amelioration–backlash theories used to describe patterns in gendered homicide posit that increases in gender equality can produce contradictory outcomes (Whaley & Messner, 2002). Improvements in gender equality provide women with options which ameliorates risk and reduces rates of homicide (Whaley & Messner, 2002). Conversely, those improved opportunities and resources for women may create a backlash effect in which perceived weakening of male privilege increases the risk of violent retaliation against women (Eggers del Campo & Steinart, 2020; Hughes et al., 2015; Whaley & Messner, 2002 ).

Specific to IPH is the exposure reduction theory proposed by Dugan and colleagues (1999). If IPH is characteristically preceded by sustained violence, enabling economic self-support, a means to leave the relationship and access to domestic violence services lessens a woman’s exposure to both lethal and non-lethal violence. Alternatively, there may be a retaliatory effect in which access to protective factors is countered by increases in violence. Studies testing the exposure reduction–retaliation theories produced mixed results. Dugan et al.’s (2003) research compared the absence or presence/form of 11 measures of domestic violence resources (such as types of arrest policies, consequences for breaches of protection orders, police domestic violence units, district attorney domestic violence specialisation, legal services and hotlines) in 48 US cities against rates of IPH between 1976 and 1996. There were reduced rates of IPH in jurisdictions where multiple resources were available, although this effect was not consistent across racial and marital status sub-groups, with some resources producing a retaliation rather than exposure reduction effect in these groups. Alternatively, Reckdenwald and Parker’s (2012) observation of 178 urban cities in the United States found a decrease in male but not female IPH victimisation. In effect, some of the proposed exposure reduction factors relieved the risk of male IPH victimisation but increased the risk of female victimisation. Improved educational, employment and economic factors had no effect on female victimisation. Both studies, however, were limited by the cross-sectional nature of the analysis and the sensitivity of the measures applied.

More recently, Australian researchers Li Eriksson and Paul Mazerolle (2013) suggested that IPH could be understood through the lens of general strain theory. Strain exists when a positively valued goal is not attained, a positive stimuli is lost or a negative stimuli is presented (Agnew, 1992). Negative emotional states mediate the effect of strain on an individual and, combined with moderating factors, influence the coping mechanism that is adopted by individuals in response. General strain theory elevates the significance of different negative emotions felt by men and women and the different sources of strain which these emotional conditions mediate (Eriksson & Mazerolle, 2013). Among men, strain comes in the form of widely documented precursors to IPH (for example, infidelity and estrangement) which are exacerbated if these strains are associated with negative emotions such as jealousy, anger and resentment (Eriksson & Mazerolle, 2013).

Risk factors for intimate partner homicide

There is a significant body of research examining the characteristics of male IPH perpetrators and the relationship and situational contexts in which IPH occurs. This research has further sought to distinguish these risk factors from those of non-fatal intimate partner violence (IPV) and other types of homicide. Most of this research utilises data from the United States and the United Kingdom and, to a lesser extent, Europe (particularly Scandinavia) and Australia.

The perpetrator

The research has identified several individual-level risk factors associated with male perpetrators of IPH, including childhood experience of family conflict and abuse, education and employment status, and history of mental illness. Risk factors supported by the larger complement of evidence are presented here.

Childhood experiences

Several studies have identified that male perpetrators of IPH are likely to have witnessed violence between their parents, or to have been directly victimised by family members (Aldridge & Browne, 2003; Kivisto, 2015). Indeed, in reviewing literature pertaining to male perpetrators of IPH, Kivisto (2015) found that approximately one quarter of male IPH perpetrators had been abused as children. However, some research suggests that, relative to IPV offenders not involved in fatal violence and to other homicide perpetrators, male perpetrators of IPH are less likely to have witnessed or experienced violence during childhood, and to have more “conventional” home lives. For example, Dobash and colleagues’ (2007) comparative study of 106 men convicted of IPH and 122 men convicted of non-fatal IPV found that IPH perpetrators were significantly less likely to come from families in which their father had used physical violence against their mother (OR=0.14), had a problem with alcohol use (OR=0.23) or had abused the perpetrator (OR=0.34). Further, IPH perpetrators were more likely than non-fatal IPV perpetrators to come from families in which their father had a skilled or white-collar job (OR=1.62) and their mother was a full-time homemaker (OR=5.41; Dobash et al., 2007). Similarly, Dobash and colleagues (2004) compared 106 men convicted of IPH with 424 men convicted of killing other men and found that IPH perpetrators were significantly less likely to come from families in which their parents’ relationship had broken down (OR=0.58) and in which the father abused alcohol (OR=0.35), had a criminal record (OR=0.24) or was violent towards their mother (OR=0.42).

Education and employment status

Limited educational attainment and unemployment have both been identified as risk factors for IPH perpetration among men, while higher levels of education and employment are protective factors. For example, compared with non-fatal IPV perpetration, the non-completion of secondary school education increases risk of IPH perpetration (OR=1.70; Spencer & Stith, 2020) whereas a tertiary education protects against IPH perpetration (OR=0.31; Campbell et al., 2003). Similarly, unemployment is associated with increased risk. Between 13 and 58 per cent of male IPH perpetrators were unemployed at the time of the homicide (Kivisto, 2015), which increased the likelihood of IPH perpetration by approximately five times compared with non-fatal IPV perpetration (OR=5.09; Campbell et al., 2003). Spencer and Stith’s (2020) meta-analysis of 17 studies of IPH found that employment reduced the likelihood of IPH perpetration by 50 per cent (OR=0.50) compared with non-fatal IPV perpetration.

Mental health

Previous studies have found that 11 to 46 per cent of male IPH perpetrators were reported to have diagnosed or suspected mental health conditions at the time of the incident (Bridger et al., 2017; Campbell et al., 2007; Sebire, 2017). This variation in estimates may be attributable to different study designs and the different ways in which mental health conditions are measured, such as by diagnoses, prior contact with a mental health professional or self-report. A history of mental illness may increase the likelihood of IPH perpetration by 30 per cent compared with non-fatal IPV perpetration (OR=1.30; Spencer & Stith, 2018) and almost six times compared with other homicide perpetration (OR=5.9; Lysell et al., 2016). Bridger and colleagues (2017) found that depression was the most common mental illness experienced by IPH perpetrators, with 27 per cent of male IPH perpetrators in England and Wales in 2011–2013 having been diagnosed with depression. This study further found that pre-offence suicidal ideation and/or suicide attempts were present in 40 per cent of male perpetrators, a significant overrepresentation compared with the general population (Bridger et al., 2017). Suicidal ideation and/or suicide attempts have further been demonstrated to almost double the risk of IPH perpetration, relative to non-fatal IPV perpetration (OR=1.62; Matias et al., 2019).

Alcohol and drug use

The risk assessment literature has produced mixed findings about the association between alcohol and drug use and IPH. Ten per cent of IPH perpetrators have been found to have lifetime diagnoses of substance dependence (including alcohol and illicit drugs; Kivisto, 2015) and 40 per cent of male IPH perpetrators have been known to misuse substances (Bridger et al., 2017). In comparison, five per cent of the general population had a past 12-month substance use disorder, and 25 per cent had a lifetime substance use disorder (Australian Bureau of Statistics, 2008). Further, compared with non-fatal IPV, substance misuse may increase the risk of IPH by 80 per cent (OR=1.80; Spencer & Stith, 2018). Campbell and colleagues (2003), however, found that illicit drug (but not alcohol) use was associated with the risk of IPH perpetration (OR=4.76). Male IPH perpetrators may be less likely to be dependent on alcohol or drugs than other homicide perpetrators (Dobash et al., 2004) and non-fatal IPV perpetrators (Dobash et al., 2007). Harden and colleagues (2019) found that perpetrator alcohol and drug use resulted in an escalation of IPV, which could result in an attempted or completed IPH.

Previous criminal history and reported intimate partner violence

Approximately one quarter to one half of all IPH perpetrators have a history of arrest for a violent crime (Kivisto, 2015). However, Spencer and Stith (2018) found that criminal history is not a significant risk factor for IPH and Dobash et al. (2007) established that previous convictions (for violent and all crimes) were associated with IPV but not IPH. Male IPH perpetrators also have a significantly less extensive offending history than other homicide perpetrators, in regard to the onset, prevalence, versatility and frequency of offending (Caman et al., 2016a; Dobash et al., 2004; Eriksson et al., 2018). However, IPH perpetrators are significantly more likely than other homicide perpetrators to have previously perpetrated violence against women (OR=11.2; Dobash et al., 2004). Violence against a previous partner is significant, increasing the odds that a man will murder their current partner by 3.4 times relative to non-fatal IPV perpetrators (OR=3.41; Dobash et al., 2007) and 2.4 times relative to other homicide perpetrators (OR=2.46; Dobash et al., 2004).

The victim

Relative to perpetrator-focused research, less is known about risk factors associated with IPH victimisation among women, which produces an incomplete analysis of the circumstances for IPH (McPhedran et al., 2018). Research focusing on individual-level risk factors for female IPH victimisation has been largely confined to socio-demographic characteristics, reproductive status and outcome.

One victim-related risk factor for IPH that has been extensively examined is the age of the victim. While findings about the age at which women are most at risk of IPH victimisation are mixed (Carmichael et al., 2019; Garcia et al., 2007), some studies suggest that victims who are younger than the male perpetrator are at a greater risk for IPH (Garcia et al., 2007). An analysis of 173 male-perpetrated and 34 female-perpetrated IPHs found that while female perpetrators were likely to be an average of four years younger than their male victim, male perpetrators were likely to be an average of three years older than their female victim (Sebire, 2017). Further, Aldridge and Browne (2003) found that women were most at risk of IPH victimisation when their male partner was more than 10 years older, and Garcia and colleagues (2007) identified that an age disparity of over 15 years posed the greatest risk. However, in a recent meta-analysis, Spencer and Stith (2018) found that the victim being younger than the perpetrator did not increase the risk of IPH victimisation.

A woman’s education has also been shown to impact their risk of IPH victimisation. In their meta-analysis of 17 studies, Spencer and Stith (2018) found that having less than a high-school education increased the likelihood of IPH victimisation by just over two times compared with non-fatal IPV victimisation (OR=2.45). In their meta-analysis of 28 studies, Matias and colleagues (2020) also compared the education level of victims of IPH to victims of non-fatal IPV and similarly found that having less than a high-school education increased the risk of IPH victimisation by just over two times (OR=2.35). Spencer and Stith (2018) further looked at the employment level and income of victims of IPH, and found that these factors did not impact a woman’s risk of IPH victimisation. Similarly, in a study of 4,931 female victims of IPH in the United States between 2003 and 2015, Carmichael and colleagues (2019) found that women may become victims of IPH regardless of their occupation and education level. Female victims of IPH have also been shown to have a significantly lower likelihood of having any criminal history than female victims of non-intimate partner homicides (23.9% vs. 49.1%; Caman et al., 2016a). Further, this criminal history was much less likely to be for violent offences (0.0% vs. 24.8% of non-IPH victims; Caman et al., 2016a). Similarly, Bridger and colleagues (2017) found that 19.1 per cent of female victims of IPH in their sample had a criminal conviction, however the breakdown of offence type for female victims was not provided.

Pregnancy or living with children from a previous relationship may also increase the risk of IPH victimisation. Previous literature reviews have established that the onset or escalation of IPV may coincide with pregnancy (Garcia et al. 2007), and that women who experience violence during their pregnancy are at nearly three times the risk of serious injury and homicide compared with women who report a cessation of IPV during the pregnancy (Campbell et al., 2007). Previous reviews of the literature have found that pregnant women and women who have recently given birth may be two or three times more likely to be killed compared to non-pregnant women (Campbell et al., 2007; Matias et al., 2019). Further, 23 to 40 per cent of homicides of pregnant women are perpetrated by a current partner, and up to 86 per cent are perpetrated by a current or former partner (Cliffe et al., 2019). Conversely, analysis of 162 incidents of male-perpetrated IPH of a female partner in England and Wales between 2011 and 2013 suggested that there were lower rates of pregnancy among victims of IPH compared with the general population of similarly aged women (Bridger et al., 2017), indicating that pregnancy did not increase risk of victimisation.

In relation to having children from a previous relationship, Sebire (2017) found that more than half of women killed by an intimate partner in London in 1998–2009 had children from a previous relationship. Indeed, the likelihood of IPH victimisation more than doubles when women have children from prior relationships living with them, when compared to non-fatal IPV victimisation (OR=2.23; Campbell et al., 2003; OR=2.29; Spencer & Stith, 2018). This risk factor may be unique to IPH as a meta-analytic review of IPV studies indicated the presence of children has little effect on non-fatal IPV victimisation (r=0.06; Stith et al., 2004; see also Spencer & Stith, 2018).

The relationship setting

The relationship setting comprises the circumstances and context for IPH (Dobash et al., 2007). This may include the nature of the relationship, the use of violence and coercive control, and the emotional states that engender these responses.

Relationship status and state

There is mixed evidence about whether the status of a relationship (that is, whether a couple are married, de facto or dating) is a risk factor for IPH. Garcia and colleagues’ (2007) review of IPH literature from the United States found that IPH primarily occurs between spouses; approximately 53 per cent of female victims were killed by their spouse (Garcia et al., 2007). Conversely, Aldridge and Browne (2003) found that IPH was more likely to occur in a de facto or cohabiting relationship than in a marriage. Bridger and colleagues (2017) similarly found that most (55.3%) homicides in England and Wales that occurred between 2011 and 2013 involved a cohabiting (or de facto) couple.

Other evidence suggests that IPH may be more likely to occur within dating relationships. In an analysis of 207 IPH incidents in London between 1998 and 2009, Sebire (2017) found that males were slightly more likely to murder their girlfriends than their wives (55.5% vs. 44.5%). This led the author to suggest that marriage may actually be a protective factor against IPH. Further, in a comparison of incidents involving non-fatal IPV and IPH, Dobash and colleagues (2007) found that women in non-cohabitating dating relationships (including engaged couples) had eight times the odds of being murdered compared with married or other types of cohabiting relationships (OR=8.0; Dobash et al., 2007). This may reflect the more tenuous nature of dating relationships which may involve less commitment between partners, greater conflict and fewer social support systems, thus making them more vulnerable to IPH (Dobash et al., 2007).

Research into the association between IPH and the victim–perpetrator relationship has also examined the state of the relationship, i.e. whether the couple is together, in the process of separating or separated. There is significant evidence to suggest that previous or ongoing separation places women at greater risk of IPH victimisation (Aldridge & Browne, 2003; Campbell et al., 2007; Dobash et al., 2004; Dobash & Dobash, 2011; Garcia et al., 2007; Kivisto, 2015; Spencer & Stith, 2018; Wilson & Daly, 1993), irrespective of the individual characteristics, circumstances and status of the relationship (Dobash et al., 2007). Previous reviews of the literature have found that recent separation was present in 20–57 per cent of IPH incidents (Campbell et al., 2007; Kivisto, 2015). This rate may be higher when recent reconciliations are taken into consideration, where the woman left and returned to the relationship during the year prior to the homicide (Campbell et al., 2003; Campbell et al., 2007). Further, in a comparison of IPH and non-fatal IPV, women who were separated from their partner were twice as likely to be murdered by their intimate partner (OR=2.33; Spencer & Stith, 2018). The risk associated with separation appears to be immediate. In the context of separation, most IPHs occur in the first three months of separation (Aldridge & Browne, 2003; Campbell et al., 2007; Dobash et al., 2007; Spencer & Stith, 2018; Wilson & Daly, 1993), and approximately 90 per cent within the first year (Kivisto, 2015).

Among women who have separated from their partner, two distinct risk factors have emerged. First, compared with non-fatal IPV victimisation, women who have separated from an intimate partner after living together have been found to have three times the odds of IPH victimisation (OR=3.64; Campbell et al., 2003). Second, legal separation and associated disputes (such as child custody arrangements) may heighten the risk associated with physical separation (Campbell et al., 2007; Dobash et al., 2007; Ellis, 2017; Harden et al., 2019; Sheehan et al., 2015).

Jealousy

Jealousy and possessiveness are often characteristics of the perpetrator–victim relationship, and have also been identified as the motive for some male-perpetrated IPHs (Aldridge & Browne, 2003; Caman et al., 2016b; Campbell et al., 2003; Kivisto, 2015; Matias et al., 2019). Jealousy may result in men using violence to “enforce rigid standards based on their beliefs about relationships between intimate partners”, in which homicide represents the ultimate act of violence (Dobash & Dobash, 2011, p. 130). Jealousy has been identified in 20–39 per cent of IPH incidents (Kivisto, 2015) and, relative to non-fatal IPV, approximately triples the risk of IPH (OR=2.58; Spencer & Stith, 2018; OR=3.31; Matias et al. 2019). Dobash and colleagues (2007) further found that male possessiveness and jealousy were five times as likely to be the cause of the conflict preceding an IPH compared to non-fatal IPV (OR=5.08). Studies on the impact of jealousy may be “limited by a lack of definitional clarity and leave open the possibility that both jealousy and related affects are common in IPH” (Kivisto, 2015, p. 308).

IPH may be precipitated by the male perpetrator’s jealousy over his partner’s new relationship (OR=4.9; Campbell et al., 2003) or actual or perceived infidelity (Aldridge & Browne, 2003). In fact, almost 19 per cent of IPH incidents analysed by Dobash and Dobash (2011) were motivated by an ongoing dispute around the female partner’s actual or perceived infidelity. This finding is echoed by Sheehan and colleagues’ (2015) analysis of interviews with 14 co-victims of IPH (family members or close friends of IPH victims), which revealed that jealousy over suspected infidelity and ex-partners initiating new relationships were significant motivators of the homicides in their sample. The risk of IPH associated with jealous partners may be increased in the context of a separation (Dobash & Dobash, 2011; Dobash et al., 2007).

Intimate partner violence and coercive control

Previous IPV against the victim is a significant risk factor for IPH (Aldridge & Brown, 2003; Bridger et al., 2017; Campbell et al., 2007; Dobash & Dobash, 2011; Dobash et al., 2004; Garcia et al., 2007; Matias et al., 2019; Sebire, 2017; Vatnar et al., 2017). Literature reviews have identified prevalence estimates for prior IPV ranging from approximately 22–77 per cent (Aldridge & Browne, 2003; Campbell et al., 2007; Kivisto, 2015). Kivisto (2015) has argued that the variation in estimates may be due to inconsistencies in the definition of IPV across multiple studies. For example, studies may include only those convicted of an IPV-related offence, or all those self-reporting previous IPV perpetration.

Victims of IPH are likely to have experienced multiple types of IPV, including physical, psychological and sexual violence and coercive control. For example, in an analysis of 177 IPH incidents in Norway in 1990–2013, Vatnar et al. (2017) found that 71 per cent of cases presented with previous IPV, of which 87 per cent involved physical violence, 79 per cent involved psychological violence and 20 per cent involved sexual violence. Arrests for previous instances of IPV such as these have been shown to be protective against IPH (Campbell et al., 2003). Further, male perpetrators of IPH appear to specialise in and diversify the violence used against their female partner. For example, 62 per cent of male IPH perpetrators with no self-reported history of IPV used coercive control (such as controlling and proprietary behaviour, sexual jealousy, psychological abuse and/or stalking) against their intimate partner compared with 88 per cent of perpetrators with a self-reported history of IPV (Johnson et al., 2017). Men who use coercive control towards their partners have been shown to be much more likely to murder an intimate partner (OR=5.60; Spencer & Stith, 2018), particularly in the context of a separation after living together (OR=8.98; Campbell et al., 2003).

Four specific forms of IPV have been found to be particularly relevant risk factors for IPH: threats of violence, non-fatal strangulation, sexual assault and stalking.

Threats to the victim

Threats to the victim can include threats with a weapon, threatening harm and threats to kill. Two previous meta-analyses have found that threats with a weapon may increase the likelihood of IPH between seven (OR=7.36; Spencer & Stith, 2018) and 18 times (OR=18.54; Matias et al., 2019) compared with non-fatal IPV. Further, an earlier multisite case control study in the United States undertaken by Campbell and colleagues (2003) found that previous threats with a weapon increased the odds of IPH by four times (OR=4.08).

Each of these studies also examined the risk of IPH associated with threatening harm and death threats. In their meta-analysis, Spencer and Stith (2018) found that perpetrators who had previously threatened to harm the victim had five times the odds of murdering their partner, compared with IPV perpetrators (OR=4.83; Spencer & Stith, 2018). A more recent meta-analysis found that generally threatening behaviour and death threats increased the likelihood of IPH by approximately 11 times (OR=11.36 and OR=10.57 respectively; Matias et al., 2019). Finally, Campbell and colleagues (2003) found that if a perpetrator had threatened to kill the victim, the likelihood of IPH occurring increased by approximately three times (OR=2.60; Campbell et al., 2003).

Non-fatal strangulation

Previous non-fatal strangulation has been shown to increase the likelihood of IPH by approximately seven times (OR=7.48; Glass et al., 2008; OR=7.23; Spencer & Stith, 2018), and is significantly more common in this population than among abused women more generally (Campbell et al., 2003; Glass et al., 2008). In their analysis of 310 IPH incidents in the United States in 1994–2000, Glass and colleagues (2008) found that previous non-fatal strangulation was present in 43 per cent of cases in their sample, compared with 10 per cent of women in an abused (non-fatal IPV) control group. Similarly, in an analysis of 220 female IPH victims in the United States over the same time period, Campbell and colleagues (2003) found that these women were significantly more likely to have experienced prior non-fatal strangulation compared with an abused control group (56.4% and 9.9% respectively).

Previous sexual assault within the relationship

In their meta-analysis, Spencer and Stith (2018) found that perpetration of sexual assault within the domestic relationship increased the likelihood of IPH by more than five times (OR=5.45). Campbell and colleagues (2003) further found that sexual assault against the victim was significantly more likely to occur within relationships resulting in IPH compared with an abused control group (57.1% and 14.9% respectively; Campbell et al., 2003). Perpetrators of sexual assault against an intimate partner are likely to be highly controlling (Harden et al., 2019) and particularly violent (Campbell et al., 2007), each of which may ultimately increase the risk of IPH.

Stalking

Stalking is a type of coercive control that often emerges or increases in frequency if termination of the relationship is suggested by the woman, or in the context of a recent separation (Aldridge & Browne, 2003; Bridger et al., 2017; Campbell et al., 2007; Kivisto, 2015; Matias et al., 2019; Sheehan et al., 2015; Spencer & Stith, 2018). Stalking victimisation is significantly associated with life-threatening IPV victimisation (OR=1.81; Brady & Hayes, 2018), and women who experience stalking have 3 to 4 times the odds of being murdered by their current or former partner (Matias et al., 2019; Spencer & Stith, 2018). Further, stalking is estimated to occur in between 23 and 90 per cent of IPH incidents (Campbell et al., 2007; Kivisto, 2015). This variation may be due to different definitions of stalking used in each study. Female victims of IPH are likely to have experienced an average of four types of stalking, most commonly being followed or spied on, the perpetrator sitting in a car outside the victim’s home or work, and the victim receiving unwanted phone calls (McFarlane et al., 1999).

Typologies of intimate partner homicide

The considerable literature reviewed above describes the “constellation” of factors that characterise and increase the risk of IPH. The literature also demonstrates that while common risk scenarios exist, they are not universal. Typologies are often developed to round out these analyses in order to capture general classifications of behaviour or perpetrator type that can be used to identify risk and treatment. Although numerous IPV typologies have been developed over the past 30 years (see, for example, Holtzworth-Munroe & Stuart, 1994; Jacobson & Gottman, 1994), one of the most dominant and well-cited is Michael P. Johnson’s (1995, 2006).

Johnson (1995, 2006) has argued that IPV (and by extension IPH) is not a unitary phenomenon and prevailing models of intimate partner violence are derivative of the information sourced to characterise offending behaviour. Thus, reported crime data tend to produce a picture of expressive, sometimes mutual violence that may spontaneously erupt into lethal violence. Alternatively, self-report and support service agency data generally show unidirectional violence marked by coercive control, which may culminate in lethal violence (Johnson, 2006). Neither scenario, however, accounts for all intimate partner violence. Instead, Johnson proposed four “types” of domestic violence that illustrate the relationship dynamic and gender symmetry in the perpetration of intimate partner violence (see Table A2).

Table A2: Typology of IPV developed by Johnson (2006)

Group	Partner A	Partner B	Perpetrator of IPV
Intimate or patriarchal terrorism	Violent and controlling	Non-violent or violent and non-controlling	Male partner
Situational control violence	Violent and non-controlling	Non-violent or violent and non-controlling	Male (predominantly) and female partner
Violent resistance	Violent and non-controlling	Violent and non-controlling	Female
Mutual violence	Violent and controlling	Violent and controlling	Equally male and female

Intimate terrorism is almost exclusively perpetrated by the male intimate partner. While the intimate terrorist is characterised as both violent and controlling, the physical violence represents one component of the pervasive use of coercive control within the relationship. Situational control violence contrasts with intimate terrorism by the absence of coercive control. Instead the relationship is characterised by persistent, low-level enmity which situationally intensifies to physical violence. Situational control violence is also largely perpetrated by the male partner. Female violence, however, exists but often in the context of bidirectional violence. Johnson’s (2006) analysis of interview data collated by Frieze and colleagues from women in Pennsylvania in the late 1970s described 97 cases of intimate terrorism and 146 cases of situational control violence. Of the 97 cases of intimate terrorism, the male partner was violent in all 97 cases and the female partner violent in three per cent of cases (ns not specified; Johnson, 2006). Among the 146 cases of situational control violence, the male partner was violent in 56 per cent of cases and the female partner violent in 44 per cent of cases.

Less common forms of violence are violent resistance and mutual violence. Violent resistance is characterised by two violent and non-controlling partners but the female partner’s violence is a protective or defensive response to abuse from their male partner (Johnson, 2006). There were 77 cases of violent resistance identified by Johnson (2006) in which 96 per cent were characterised by a violent female partner. Mutual violence is the least common scenario; both partners are violent and controlling and both partners are equal participants in the perpetration of violence (Johnson, 2006). There were just 10 cases of mutual violence, evenly split between male and female perpetration of violence.

The applicability of IPV typologies to IPH is debatable (Kivisto, 2015), dependent on whether the homicide is treated as the outcome of continued intimate partner violence (Dixon et al., 2008; Monckton Smith, 2019) or a distinct expression of violence (Gelles, 1991). Typologies of IPH have tended to focus on the psychopathology of the perpetrator as the distinctive trait, coupled with the types of behaviour and relationship events preceding and succeeding the homicide (see, for example, Dixon et al., 2008; Elisha et al., 2010; Kivisto, 2015; Liem & Koenraadt, 2008). Campbell et al.’s (2003) 11-city analysis of pre-homicide risk factors categorised incidents of IPH into three clusters, largely differentiated on the age and mental health of the perpetrator, relationship status with their intimate partner and whether the homicide was followed by the perpetrator’s suicide. Each cluster accounted for around a third of the incidents, and included:

• the “non-depressed/non-violent” cluster: men in this cluster were in intact relationships and around half attempted suicide or died by suicide after killing their intimate partner (Campbell et al., 2003)
• the “depressed/non-violent” cluster: the suicide rate was the same for men in this cluster as it was for men in the “non-depressed/non-violent” cluster, but a larger proportion were separated from their partner (40% vs. 21%) at the time of the homicide
• the “non-depressed/violent” cluster: these men were younger than men in the other two clusters (mid-30s compared with early to mid-40s) and less likely to suicide, and most were separated from their partner.

The pathway to intimate partner violence

Moving beyond the examination of static risk factors and sub-groups of offenders, more recent IPH research has emphasised the fluidity of violence or the “influential and continuous interaction between individuals and the various situations they encounter” (Vatnar et al., 2017, p. 395). It is the specific sequence of encounters and events that can direct the fatal course of an intimate partner relationship (Dixon & Graham-Kevan, 2011; Vatnar & Bjørkly, 2008).

As described previously in this report, such sequencing of IPH was recently undertaken by Monckton Smith (2019). The study involved two stages of data collection and analysis: first a review of 372 cases of male-perpetrated IPH included in the United Kingdom’s Counting Dead Women database for the period 2012-2015, and then a more detailed analysis of 25 cases IPH from the author’s previous research. Monckton Smith (2019) identified eight stages that led to and culminated in the homicide. These eight stages were associated with three relationship phases – pre-relationship, relationship and post-relationship – and the known risk factors associated with each phase.

Research gap

There has been substantial research focus on IPH, in particular male-perpetrated IPH, over the last decade, and findings from the relevant literature have been cited above. These describe the underlying motivations for IPH, the combination of behaviours and emotional states of male IPH perpetrators, and the relationship events understood to initiate the homicidal response. The current knowledge records tangible risk factors which identify when lethal violence is a possibility. Yet these studies effectively represent static representations of IPH – they capture the critical elements but not the patterning or convergence of these elements in the lead-up to and commission of IPH. Monckton Smith’s (2019) research provides a blueprint for IPH, but it is preliminary at best. What is needed is a complementary examination of this blueprint (alongside the IPH literature) to establish if IPH follows a scheduled pathway and how and when the interactions of pathway factors contribute to the lethal outcome.

Of equal relevance is the generalisability of the documented research findings to the Australian context, particularly in regard to Aboriginal and/or Torres Strait Islander women who are absent from these analyses. This research intends to build on and complement studies from state and territory domestic and family violence death reviews, the Australian Institute of Criminology’s Homicide in Australia series and the Australian academic literature to describe the nature and course of intimate partner relationships that end in the homicide of a female partner, and the events and behaviours that could foreshadow fatal outcomes within relationships.