Understanding and responding to family violence risks to children: Evidence-based risk assessment for children and the importance of gender

Abstract

This article responds to recent calls to better understand and respond to family violence risks to children. Drawing on the findings of a wider research project on family violence risk which engaged with over 1000 members of Victoria’s family violence system through a survey, focus groups and in-depth interviews, this article examines practitioners’ views on current practices and future needs for reform to improve family violence risk assessment practices for children. The findings have implications both nationally and internationally, given the dearth of evidence-based family violence risks assessment tools. Key findings reinforce the importance of interagency collaboration and a shared responsibility for children impacted by family violence across services and the importance of specialised training in this area. Caution, however, is raised about ongoing patterns of blame for mothers affected by family violence: we conclude that the need to address children’s risk in family violence is critical but ongoing attention to how gendered patterns structure family violence and social responses is also essential.

Keywords

Child victims family violence family violence risk assessment gender risk management

In February 2014, the ‘premeditated’ killing of 11-year-old Luke Batty by his father shone national attention on the impact and risk of family violence for children in Australia (Gray, 2015, p. 1). At the time of his death, multiple agencies in Victoria’s integrated family violence and criminal justice systems were alive to the dangers facing Luke and his mother, Rosemary Batty. Careful to note that no organisation, ‘singularly or in combination’, directly contributed to the death of Luke Batty, the 2015 Coronial Inquest found that there was ‘no validated risk assessment tool’ to predict the likelihood of filicide. Documenting the series of ‘missed opportunities’ to intervene in the case (Le Grand, 2015), the Inquest Findings made a series of recommendations to improve risk assessment, coordinated risk management and safety planning to ensure family violence risks to a child are identified, shared and responded to more effectively (Gray, 2015, pp. 3–4).

In the three years since Luke Batty’s death, reports and action plans at the state and national level have recognised the severity of family violence impacts on children (see inter alia Council of Australian Governments (COAG) Advisory Panel, 2016; Special Taskforce on Domestic and Family Violence in Queensland, 2015; Victorian Royal Commission into Family Violence (RCFV), 2016). Numerous national and international studies have highlighted the extent of violence in the family lives of children (Australian Bureau of Statistics, 2005; RCFV, 2016). Increasingly, research has attended to children as direct victims of family violence, but also the impact of children’s exposure to domestic violence (see inter alia Humphreys, 2008; Powell & Murray, 2008; Richards, 2011; Stanley & Humphreys, 2017). In Australia, the Personal Safety Survey found that of all women who had experienced partner violence and had children in their care (n = 11,800 women), 59% reported that the children had witnessed violence (Australian Bureau of Statistics, 2005). However, despite acknowledgement of the extent and harms of family violence on children alongside the need for specialised responses and early intervention, there continues to be limited service responses directed specifically at children, and the evidence base on understanding family violence risks to children remains critically underdeveloped.

Most recently, the RCFV (2016) findings highlighted the need to focus attention on child victims of family violence. The first recommendation made by the Commission stated that the current risk assessment and management framework in Victoria should be reviewed and redeveloped, in part aimed at enhancing child risk assessment processes. Taking that recommendation as our focus, this article examines the need to better understand and respond to family violence risks to children and how this may be best achieved. It draws on the findings of a wider research project which engaged with over 1000 members of Victoria’s family violence system. The findings provide unique insight into practitioners’ views on current practices and future needs for reform to improve family violence risk assessment practices for children. While the data here are Victorian, an examination of family violence risk assessment practices for children is pertinent for all Australian state and territory jurisdictions currently reviewing and/or developing new family violence risk assessment and management practices.

This article is structured into four parts. In part one we outline current family violence risk assessment practice for children internationally and nationally with a focus on the Victorian context. Part two provides an overview of the research on which this article draws. Part three acknowledges other research which has stressed the importance of understanding how family violence undermines the relationship between mother and children, and in doing so, examines concerns about assessing risk of family violence for children without attention to how family violence impacts the relationships between mothers and children. In part four, we detail practitioners’ perceptions on how best to assess family violence risk for children, including reflections on the importance of interagency collaboration and a shared framework of responsibility across services, and the importance of specialised training in this area. We conclude that the need to address children’s risk in family violence is critical but ongoing attention to how gendered patterns structure family violence and social responses is also essential.

Family violence risk assessment for children: International and national practice

Within Australia there is no nationally consistent approach to family violence risk assessment and management (COAG Advisory Panel, 2016). Some states have developed a common risk assessment framework to support integration and management of risk across services and agencies, while other state jurisdictions adopt different methods depending on the practitioner area and/or agency involved (COAG Advisory Panel, 2016; Special Taskforce on Domestic and Family Violence in Queensland, 2015). In Victoria, South Australia (SA) and Western Australia, a common risk assessment framework is used; however, the contents, processes and operations differ (McCulloch, Maher, Fitz-Gibbon, Segrave, & Roffee, 2016, pp. 145–147). In no State or Territory instrument, are children’s risks in family violence a focus of risk assessment and management.

Internationally, there are two key approaches mobilising an assessment of family violence risks to children – the Safe and Together Model (Mandel, 2013, 2014) and Barnardo’s Domestic Violence Risk Identification Matrix (DVRIM). The Safe and Together model, developed by David Mandel in the US, represents a child-centred approach to addressing family violence (Mandel, 2013, 2014), since implemented in jurisdictions including England, Scotland, Ireland and Singapore (Mandel, 2013). The model focuses on the man as father and is aimed at service providers and professionals who work with male perpetrators of family violence.

The central aim of the Safe and Together model is to increase practitioners’ understanding of the effects of domestic violence on children and following observation of each family to evaluate risk and create a unique case management plan. The model is based on three key principles:

To keep children ‘safe and together’ with the non-offending parent to ensure safety, stability and nurturance, and to allow for healing from the trauma of domestic violence.

To partner with the non-offending parent to develop a risk management plan that will be efficient, effective and child centred.

To intervene with the perpetrator to reduce the risk and harm to the child through engagement, accountability and the justice system (Mandel, 2013, 2014).

These aims reinforce perpetrator accountability and responsibility and locate the relationship with the non-offending parent, generally the mother, as vital.

Within Australia, the Safe and Together model was recently adopted in SA and is used by Families SA in its work with fathers who have perpetrated domestic violence (Junction Australia, 2015). The Victorian Family Violence Risk Assessment and Risk Management Framework, most often called the Common Risk Assessment Framework (hereafter the CRAF) and practice guides also reference the need for professionals to minimise risk to and vulnerabilities of children, identify the harm of living with family violence and understand the impact of removing children from the family home (Department of Human Services (DHS), 2012a, p. 30).

In addition to the Safe and Together Model, Barnardo’s DVRIM, first introduced in Northern Ireland in 2003, also specifically assesses the risk children face where family violence is present. The DVRIM adopts gendered language that assesses male-to-female violence and identifies risk as age specific and greater for younger children (Bell, n.d.). While it has been used for over a decade in the UK, to date, the DVRIM has not been subjected to validity testing and has not been adopted in Australia.

Family violence risk assessment for children in Victoria

The CRAF has been in use in Victoria since 2007 and is designed to support risk assessment for a range of different professional groups who come into contact with family violence. The RCFV (2016) suggested that the inability of the CRAF to meaningfully consider risks to the child is arguably one of its most significant limitations. These limitations were also recognised in the 2015 Inquest into the death of Luke Batty, when State Coroner Gray (2015) noted that the CRAF framework is narrowly aimed at assessing risk to a woman, rather than a child or the perpetrator.

In 2012, the Victorian government released the Assessing children and young people experiencing family violence framework (DHS, 2012b). Colloquially called the ‘Children’s CRAF’, the aim was to ‘support family violence professionals to assess the safety, and needs of unborn children, infants, children and young people affected by family violence’ (DHS, 2012b, p. 1). However, the limited training that supported its introduction and questions about usability have significantly impeded its operation (McCulloch et al., 2016, p. 97). Within Victoria’s family violence sector, however, this framework is recognised as a useful starting point for developing and embedding a child-specific family violence risk assessment (McCulloch et al., 2016).

The Victorian RCFV devoted significant attention both to children and to risk assessment and management practices, concluding that the needs of children and young people experiencing family violence in Victoria have been overlooked (RCFV, 2016; Summary, p. 23). The RCFV provided the background and impetus for this research study.

Research design

This article draws on the findings of a larger research project into the family violence risk assessment and management framework in Victoria. Through a survey (n = 836 respondents), focus groups (FGs) (n = 262 participants in 21 FGs) and a small number of expert interviews (EIs) (n = 10 participants), the research engaged with over 1100 people from 127 organisations and relevant workforces, including specialist family violence services; primary health care; maternal and child health nurses; community support workers; those working in family and children’s services; governmental officers from the departments of Justice, Education, Premier and Cabinet and DHHS; and relevant non-government organisations.¹ A smaller number of individual interviews (n = 5) and FGs (n = 3) were undertaken with 24 women who have directly experienced family violence and worked to achieve their own safety. The specifications of the larger research project provided that Aboriginal and Torres Strait Islander (ATSI) communities would not be engaged with as a separate piece of work would be undertaken to review family violence risk assessment practices in ATSI communities. For this reason, this article does not examine issues of children’s risk assessment as they apply to aboriginal communities.

In this article, we focus on findings from the sector-wide survey, FGs and EIs, addressing processes of risk assessment for children. The sector-wide survey achieved a good response rate (n = 64, or 11% of responses) from workers in child and youth services and organisations. Among survey respondents who work directly with victims (87% of survey respondents), children emerged as the second most common group of clients. However, the responses more generally indicated that those working in specialist and generalist services were mindful of child-specific aspects of risk and the limitations of the CRAF to currently respond or capture this specificity.

The survey was designed to capture participants’ awareness of, and engagement with the CRAF, as well their utilisation of other risk assessment tools and any challenges with and opportunities to improve the existing CRAF that they identify. As such it included multiple-choice, multi-answer and open-ended questions. The survey was designed in collaboration with DHHS and sought to capture practice and attitudinal data, reflecting the priorities of the research. The survey was conducted online (using Qualtrics software) and was publically available; however, to encourage engagement potential participants were recruited via directly contacting Victorian specialist and related workforces to increase awareness and encourage participation. The survey response analysis was undertaken using a combination of Qualtrics and Stata software, allowing survey data to be analysed quantitatively and qualitatively.

All interviews involved directly contacting identified key stakeholders, while the FGs varied from being open to interested participants from related workforces, to targeted key groups and/or key populations (e.g. children’s services and LGTBIQ services FGs). Interviews and FGs adopted a semi-structured format with questions based on emerging themes from the literature review, the RCFV recommendations as well as those stipulated by DHHS. All interviews and FGs were audio-recorded and transcribed in full, following which the data were thematically analysed using NVivo qualitative analysis software. All participants were assured anonymity as part of the data collection process. As such, throughout this article where specific quotes from our participants are used they are referenced with an abbreviation according to the point of the study at which they were collected (i.e. EI, FG, Open Survey response (OS) as well as a randomly assigned number).

Concerns about children’s risk assessment and maternal relationships

Children are victims of family violence both directly and indirectly through witnessing violence and through the impacts of victimisation on their mothers. As noted above, the RCFV stressed the importance of this and registered the limitations in current responses to children’s risk. However, whilst children and their safety may be recognised as an important issue, beyond and separate to concerns related to the mother (as the victim/survivor), there is often limited recognition of the entanglement of concerns pertaining to children’s safety and women’s ability to mother and the ways in which this entanglement impacts full and frank assessment of child risk and welfare. As Thiara and Humphreys (2017) observe, a key outcome of family violence for victim/survivors is the persistent undermining of women’s abilities to mother (see also Jaffe & Crooks, 2005; Stanley & Humphreys, 2014, 2017). This raises concerns for women about losing their children when they disclose or report family violence: what Stanley and Humphreys (2017, p. 100) describe as the ‘double whammy’ of family violence and child protection involvement (see also Jeffries, 2016). The Victorian Commission for Children and Young People (2016) Neither Seen, Nor Heard Report into child deaths found that family violence notifications involving children very rarely resulted in mothers being referred to family violence support agencies: instead child protection agencies often considered mothers were responsible for securing their children’s safety. These findings were echoed by participants who indicated their awareness of the potential for further victimising or undermining maternal victims of family violence:

If we start extending it [family violence risk assessment] to children, there is fear of authorities and statutory response around the welfare of children…. So we actually have to be really mindful [of]…the unintended consequence of seeing it through the child’s eyes … I’ve read most women are frightened to actually start to talk about these things for fear of losing their children. (FG14)

As captured here, institutions and authorities such as the Family Court, Child Protection and Family Services have specific responsibilities for reporting risk to children and mothers are acutely aware of this. While the mandatory reporting requirements and legislation governing child protection and domestic violence vary across Australian states and territories (see Jeffries, Field, & Bond, 2015, for an analysis of Australian legal practice in this area), each jurisdiction legislates that health, welfare, education and justice professionals who interact and engage with children must report incidents as well as suspicions of child abuse, including family violence. Research has consistently demonstrated that the concerns some maternal victims of family violence may hold regarding the impact of reporting family violence are not unfounded. These mandatory reports regarding suspected child abuse including family violence have been shown in some circumstances to create unintended adverse consequences if the risks of the mother and her children are assessed separately or distinctly, most particularly in relation to blame apportioned to maternal victims of family violence rather than support for the family (Humphreys, 2008; Richards, 2011; Stanley & Humphreys, 2017).

There was further acknowledgement in our research from those working within the sector of their concerns and observations that women may not disclose family violence if they fear losing their children:

My concern is if you tell a woman or ask a woman, “Have your children witnessed or been exposed to family violence?” some women aren’t going to tell the truth and the reason why is… their concern is the children will be taken from them.… (FG18)

We still hold women accountable for the safety of the children and I think going to that point, if the risk assessment is, are we keeping the women and children safe? (FG16)

Research suggests that such fears are likely exacerbated in Indigenous communities, where the history of the Stolen Generation and ongoing fears surrounding removal of children present significant barriers for women reporting their own domestic violence victimisation or the impact on their children (Humphreys, 2008). There is also emerging research which suggests migrant women without permanent residency may also fear the removal of children and permanent separation if their children are Australian citizens and their mothers are not (Segrave, 2017). In considering how family violence risk assessment and management practices for children can be best developed, attention to the experiences of diverse communities will be critical, as intersections of gender and other forms of difference are likely to intensify unintended adverse outcomes. A children’s risk assessment tool that fails to understand the ‘complex picture … of women who are systematically physically and emotionally disabled becoming poorly placed to respond consistently to their children’s needs’ (Thiara & Humphreys, 2017, p. 140) and assigns blame and negative responsibility to mothers is likely to further victimise mothers and create more damaging effects for children. In the following sections, we consider existing practices of children’s family violence risk assessment, asking how we can maintain a gendered lens that supports children’s access to mothering as well as their security.

The need to develop evidence-based children’s risk factors

The RCFV described children as the ‘silent victims’ of family violence. This sentiment was strongly echoed by all participants, revealing a shared view that all family violence harms children. Practitioners recognised that the harms suffered by children are serious and cumulative and can include direct victimisation, witnessing victimisation and impacts due to compromised maternal capacities to provide care and nurture. All forms of child development can be affected. One FG respondent said:

Risk to children and young people is about more than risk of physical harm. It is also about risk to emotional wellbeing and risk to life trajectory. Family violence impacts all aspects of a child’s life including the establishment of normal developmental pathways, often beginning in utero. The mother-child bond is attacked … Risk to life trajectory including limited early childhood support, moving schools, poor emotional regulation all directly impact a child’s life opportunity as a result of family violence. (EFG)

Despite this shared recognition of the harm of family violence on children, respondents considered that family violence risks to children remain generally unassessed and that the CRAF is focused on adults and IPV. As two FG respondents said:

At no point is anyone actually asking a child what their view is and what their experience has been. To me that’s crucial. (FG20)

For the referrals that come from us, there is no risk assessment or any knowledge in some cases that there are children involved which is something we need to improve on … the only time they’ll have a specific CRAF done is if they’re the primary AFM [Affected Family Member] and the CRAF is adult-based so it’s not even relevant. (FG20)

Many echoed this view, commenting that current risk assessment ‘doesn’t do much for kids’ (FG9) and that children ‘have no voice’ in Victorian family violence risk assessment (FG18).

There was overwhelming support within the integrated family violence system for improved child-specific risk assessment and management practices. In the data there were a number of key issues that were identified across the family violence sector as critical to building and developing effective family violence risk assessment and responses for children. These were modifications of existing risk assessment frameworks to better capture risks to children, the importance of interagency collaboration and a shared framework of responsibility, the need to develop pathways and referrals to support children’s risk assessment and the need for specialised training to support workers in this space.

(a) Modifying universal practice to better capture family violence risks to children

Despite recognising the limitations of the CRAF in capturing risks for children, in lieu of an alternative risk assessment, the majority (63%) of survey respondents from child and youth specific organisations acknowledged using the CRAF in their professional work. However, many had implemented modifications in order to better and more directly assess family violence risks to children. This finding was supported by FG and stakeholder input.

Modifications of the CRAF generally are not uncommon nor unique to professionals working with children: 15% of all survey respondents (n = 125) using the CRAF indicated that they use a modified version of the tool. Within that group, children-related modifications were cited as the most common reason for modifying the universal tool, with 35% citing this as the primary focus of the modifications made. As one FG respondent explained:

I think it would be nice to incorporate questions that are specific to children. I know in family services, we’ve tweaked it a little bit where we do have a little bit just specific questions towards children and harm or relationship between the child and her. (FG15)

Modification of the CRAF to incorporate some aspects of the Children’s CRAF (DHHS, 2012b) framework emerged as common practice among a small number of respondents in the interviews. Despite concerns about training and roll-out, respondents identified value in this child-focused framework and urged its use as a platform in the development of new risk assessment frameworks to effectively assess the risks specific to children. One expert interviewee stated:

The thing about the children’s guide [CRAF] is because everybody was so stretched and there was a sense of well we can’t ask women and the men’s services to do another thing … I don’t think it’s really be taken up. So I think it has been a little bit influential but not very. I think it’s largely a resource that sat on the shelf to some extent. (EI)

Another respondent explained: ‘I find the Assessing children and young people experiencing family violence much more useful [than the CRAF] – actually I have invented a combination of the two tools which sort of works for me’ (OS response).

In considering what specific modifications are required to ensure that the CRAF, as a universal risk assessment framework, can better assess risks to children, respondents noted that presently such risk is only captured through the lens of risks to the mother. This approach was flagged as inadequate by many respondents who emphasised the importance of distinguishing between the risks to the mother and to the child. The survey offered a number of opportunities for respondents to provide detailed, open-ended responses and the need for the CRAF to be more child specific was the most consistent (i.e. of 71 qualitative responses, 38 respondents, or 54%) feedback in the open-ended questions asking respondents about issues that need to be considered in relation to children, risk and the current CRAF in Victoria. While most respondents simply made statements such as the CRAF captures ‘nothing on children’s risk’ there were a few more detailed responses. This respondent’s statement captures the broad concerns raised:

The CRAF combines women and children however the risks for a woman and a child are and will be different. I strongly believe that if we are to reduce the number of child deaths from family violence we need to separate the CRAF for women and children and view a child as an independent victim. This will also align with Child protection and ensure less hostility between agencies and more focus on keeping all victims safe. (OS response)

The importance of recognising children independently as well as co-dependently with their victim/survivor parent (most often, the child’s mother) was consistently raised across the research. Practitioners in FGs concurred with the need to assess family violence risks to the child beyond the lens of the mother:

You run the risk that children are always dependent upon their mother’s disclosure before they can seek assistance … You’ve got to allow children to step out of the secrecy of all of this. (FG14)

To see them as clients in and of themselves as opposed to a client of the person connected to the protective parent. (FG7)

Risk to children should be assessed separately from the mother, and where possible, the child should be interviewed separately from their mother … Children may not communicate all of what they have experienced to their mothers, especially if the perpetrator is seeking to diminish the mother–child bond. (FG6)

Several FG respondents also emphasised the importance of assessing risk for each child within a family individually. Risk may differ among siblings as well as between the mother and child. As explained by two FG respondents:

The risks are different and you need to ensure that each individual child has their own risk assessment because the risks are not the same as mums for starters, and they’re not the same as their siblings sometimes. You need to make sure that you look at every risk for the child … there is significant risks for some kids, and if you’re going to make children safe – which has got to be our duty of care – we need to identify what the risks are to put the supports in for mum to be able to look after the child in a way that builds safety. (FG1)

Children aren’t the same. I think that in a family with three kids, you would have children reacting quite differently to the same situation. Some will have different needs to the others. Some might then become – be also bullying kids at school and being violent themselves, and others might be totally withdrawn and depressed. I don’t like that language, and I think we should see them as individuals, not as part of the mum. (FG12)

These views contribute to a growing body of scholarship that argues family violence risks for an adult and child may be ‘linked but separate’, requiring more targeted and possibly distinct assessment and response (Stanley & Humphreys, 2014, p. 79). This body of research indicates that the traditional view of children as passive victims of domestic violence should be replaced by an acknowledgement that children carry their own individual risks and experiences which may be distinct from those held by the adults within their familial environment. The findings here also raise the question of the number of risk assessment tools in use across Victoria that depart from standardised practice. While there is no validated risk instrument included in the CRAF at present, the impact of modifications in practice would need to be taken into consideration in the development of any future iterations of the framework.

(b) The importance of interagency collaboration and a shared framework of responsibility

Recent reviews and inquiries, including the Batty Coronial Inquest (Gray, 2015), and the Neither Seen, Nor Heard report (2016) reinforce that children’s risk in family violence contexts may be heightened when this is limited multiagency collaboration. This has been stark in cases where the children have died. These reviews have noted the range and number of agencies that typically have contact with, and potential view of, a child prior to the child’s death (Victorian Commission for Children and Young People, 2016; Gray, 2015; Office of the State Coroner, 2014; Queensland Family and Child Commission, 2017). This finding is supported internationally, including in a recent Canadian study of children’s risk for homicide in the context of family violence, which found that families experiencing family violence where children were present had a high number of agencies involved but did not routinely undertake risk assessment or safety plans (Hazel, Hamilton, Jaffe, & Campbell, 2013, see also Jaffe, Campbell, Olszowy, & Hamilton, 2014).

Several respondents noted the importance of integrated responses in this area and the need to move away from the current siloed approach. As one interview respondent explained:

The family violence service system has been designed with the needs of women in mind. The Child Protection system has been designed with the interests of children in mind, but [the] integrated family system tries to kind of bring all those in together and doesn’t do that very successfully either. So, we are working in silos, but the reality is we’ve got to stop. (EI1)

Another respondent argued for an integrated approach to ensure that children are referred to specialist services where needed:

I think we need the system to be fluid enough so that if there are potential children at risk in this situation I don’t have to be the actual expert protective worker but can raise a red flag or a referral or something where there’s a more integrated agency connection. (FG4)

Developing an integrated approach to children’s risk assessment was perceived by many as an opportunity to clarify lines of communication and responsibility to ensure that risks to the child are kept within view. The value of multidisciplinary approaches in responding to children exposed to family violence has been previously emphasised in research and in the findings of state and national inquiries (see, e.g. ALRC/NSWLRC, 2010; RCFV, 2016; Richards, 2011; Special Taskforce on Domestic and Family Violence in Queensland, 2015).

There was a shared perception among respondents that the current lack of integration contributed to a blurring of lines of responsibility to the detriment of children. This view is captured in the following FG excerpts:

A significant barrier … is the lack of clarity regarding where the responsibility lies. Children may move through multiple services before a specific risk assessment is completed for them individually. Alternatively, children will have multiple risk assessments but collaboration and information sharing between services is insufficient to identify this. (EFG)

I think the idea of who is responsible is really unclear … we’re often seeing people when they’ve had a lot of contact with other services but we very rarely have a risk assessment for children when we’re taking on the referral, so the family have often had lots of contact … we’re often shocked that they haven’t been [assessed]. (FG3)

This supports previous research by Stanley and Humphreys (2014, p. 78), finding that agencies responding to children experiencing family violence, including police, child protection and specialist domestic violence services, have often ‘struggled to harmonise and co-ordinate’ their communications, collaborations and responses. The outcome may be a failure among agencies to adequately assess and respond to the risk present. In their Canadian review of children killed in the context of family violence, for example Hazel et al. (2013, p. 184) found that despite the high number of agencies involved with each family, risk assessments were completed in less than 15% of the cases under study. They concluded that while an increase in the number of agencies involved with a family should improve the degree to which risk is assessed and kept in view, no benefits were observed and indeed ‘several missed opportunities for intervention and prevention strategies by professionals’ were evident in cases where a higher number of agencies were involved (Hazel et al., 2013, p. 185).

Another key benefit identified by respondents of integrating service responses in this area was the opportunity to create a shared language surrounding family violence risks to the child. The 2010 Report by the Australian Law Reform Commission with the NSW Law Reform Commission noted the importance of a shared contextual understanding of family violence across state and territory jurisdictions. The creation of a shared language and understanding of women’s family violence risk has been cited as one of the key achievements of the CRAF framework (McCulloch et al., 2016), but this has not yet extended to assessing children’s risk. The need for common understandings emerged clearly throughout the research among practitioners:

particularly relating to the risk factors for children the CRAF implementation really does offer an opportunity to bring people together and actually develop that shared language and understanding so that it can be implemented in a way that provides a more holistic response. (FG6)

Multiple sectors support children who experience family violence. In order to begin to identify key risk indicators there must be a unified definition of what constitutes risk to children and young people in family violence situations. (EFG)

The challenge of creating this shared understanding and language in the context of limited knowledge about children’s risk assessment was clearly identified by one FG respondent:

there’s no common understanding of what risk means and we’re working from different underpinnings … how can we implement a CRAF around child specific risk factors? We have to actually have a common understanding of what risk means in the context of family violence for children. (FG3)

The high level of engagement of child specialist services and practitioners in this research indicates sector willingness to work to develop child-specific family violence risk assessment and management practices.

There was a shared concern among respondents involved in all stages of our research that developing effective and robust referral pathways is essential to creating an effective risk-sensitive response to children experiencing family violence. Respondents reflected on the lack of referral options currently available to children impacted by family violence where the child was not listed as the primary victim. While the past decade has seen the emergence of a small number of child-specific early intervention programs and services, specialised responses to children experiencing family violence across Australia, including in Victoria, remain scarce and typically inaccessible for diverse communities and rural and regional areas (Powell & Murray, 2008; RCFV, 2016; Richards, 2011).

For many practitioners, the lack of referral options was a barrier to assessing risks to children in the first instance for fear that nothing could be done once that risk was identified. As Stanley and Humphreys (2014, p. 83) have previously noted, without risk management processes, risk assessment alone ‘does not support families, achieve change or provide interventions for children’. Many participants identified the lack of services for children as a critical influence on how and whether children’s risk is assessed. Some, including regional participants, suggested that worker reluctance to specifically assess children’s risk arises from an unwillingness to do further harm where services are not available. As explained by two FG practitioners:

[A] lot of their workers were not asking that question because there was no mechanism for them to work with children once they had identified risk and so the fear of doing harm by having children articulate that they weren’t safe and then they’re still not safe. (FG14)

Let’s be absolutely clear … they [practitioners] don’t want to go there because then they’re going to find out that they’ve got this incredibly traumatised bunch of kids and the mother knows that and they’re going to say, “Well sorry, we’ve got nowhere to take them. Goodbye, finish.” (FG3)

Another practitioner explained the difficulty of receiving information about a child’s risk where there were no perceived options for referral or management and where the child is not central to the remit of the service provider:

What do we do with the information? Because we don’t know where that person’s going … so we can’t actually put anything in place, and the staffing team don’t feel confident to actually do that because again, they’re coming from the perspective of we’re [a] family violence service, we work with women, we’re not skilled to do that [assess children]; it’s not our role. (FG3)

Several respondents linked this perceived lack of child-specific services and referral pathways to a lack of case funding for children associated with family violence cases. As two FG respondents explained:

There are certainly a lot of workers who are very skilled and very passionate about including children and assessing children, but if they’re embedded in services that don’t see children as their core business, who see adults as their core business … you still have services that say, “Our focus is the adult, we’re not funded to work with children,” … it’s very, very difficult. (FG3)

We’re not funded to see children, we don’t assess children, there’s too much risk in us actually doing risk assessments for children, so there’s a fundamental resistance. (FG16)

This raises a broader question about how funding for adult and child clients impacts and defines the availability and provision of services. This study illuminated the unintended consequences of service responses to family violence funded solely with male perpetrators of heterosexual intimate partner violence in mind. While intimate partner violence against women occurs at rates higher than any other form of family violence, the impacts on, and risks for, children need to be considered as part of the diverse contexts of family violence.

In addition to the lack of a clear referral pathway system, many respondents reflected on cases where children may present through different pathways, and where privacy concerns and/or other factors may mean that existing knowledge about the risks to children is not shared between organisations. The need to ensure information can be shared within and between multiple agencies engaged with individual families echoes prior research which documents variability in information gathered by different agencies working with families (Stanley & Humphreys, 2014, p. 80).

The importance of greater information sharing pathways and protocols have also emerged as a key finding in recent Australian reviews of responses to family violence (see inter alia ALRC/NSWLRC, 2010; Gray, 2015; RCFV, 2016). The COAG Advisory Panel (2016), for example, concluded that a critical element of an integrated response to family violence is information sharing between agencies and organisations across government and non-government sectors. Without this, there is a real risk that the harm to a child (or adult person) can be underestimated (COAG Advisory Panel, 2016, p. 106). The COAG Advisory Panel (2016) found that there are a number of factors which impede effective information sharing in family violence cases, including cultural barriers, financial consideration, human resource barriers and legal barriers (including privacy laws and legislation). Each of these barriers is relevant in building knowledge about family violence risks to children.

(d) The need for specialised training to support workers

In Victoria, training focused on child-specific risk assessment and management has not been available. Many in the sector felt that a lack of training has diminished workforce confidence and capability in this area of critical need. Again, this echoes other research reinforcing the benefits of different types of specialised family violence training, including improving knowledge, informing attitudes and enhancing practice (Turner et al., 2017, p. 35; see also McCulloch et al., 2016, pp. 83–85).

Alongside acknowledging that assessing risks to a child is ‘a specialised thing’ (EI), FG respondents identified the lack of specialised training as a key barrier.

We’ve got varying degrees of level of comfort that our staff have, and experience, in actually assessing and meeting with the kids, the children … so we’ve got some staff who are great and brilliant and that’s their interest and they will do because there is an expectation in family services in our work that we also take on board the voice of the child, so the children are sighted, the children are meant to be part of a family action plan … I think there’s a lot of discomfort and people say, “I don’t want to make things worse,” or, “I’m not sure”. (FG3)

I think part of that is capability, skilling up and part of it is the tools and what you can bring in, the tools and the prompts and the framework that supports workers to be able to hold those kinds of different lenses. (FG19)

In their reflections on how this could be improved moving forward, the inclusion of child-specific risk indicators within the common risk assessment framework would, in and of itself, serve to increase confidence in professional practice. Participants explained that it would be educative for practitioners while also providing them with a tool to aid them in articulating concerns surrounding children’s risk. As one regional practitioner respondent explained:

I think that would be really useful, only from a perspective where that would help you to articulate your concerns to child protection, because sometimes it’s what the children are saying, and what the children are doing. And the concerns you’ve got for them, but we can’t really articulate that very well to child protection. (FG13)

Systematic specialised training and new approaches to information sharing, interagency collaboration and referral pathways were identified as critical in advancing responses to children’s risk assessment and management. Yet, as these shared sector views suggest, building towards family violence risk assessment and management for children will require a range of changes including modifications of risk assessment tools to recognise children as linked to their mothers in terms of risk, but also facing specific individual risks. Given ongoing challenges in effective responses to the family violence experienced by women, extension to, and inclusion of, children, will require attention to the ways in which gender and dependency interact in this context.

Conclusion: Creating a better understanding of family violence risks to children

This is a major gap in the CRAF and should be progressed as a matter of priority. (EFG)

As captured in the above FG quotation, participants in this research were strongly supportive of enhancing family violence risk assessment and risk management practices for children. That support stems from shared recognition that not enough is currently being done to assess and manage child-specific risk. Specific issues identified were interagency collaboration, information sharing protocols, specialised training and a clear framework to inform and support risk management and referral once risks to the child are identified. These views support the recent findings of inquests and reviews into the deaths of children in the context of family violence, where missed opportunities for intervention and identification of risk has emerged as a key failing of integrated family violence systems across Australia and internationally (see inter alia Hazel et al., 2013; Jaffe et al., 2014).

This gap in current practice is not unique to Victoria nor Australia. Recent international reviews increasingly recognise the need to bring family violence risks to children to the fore. While a small number of risk assessment models aimed specifically at children exist in international practice, to date these have not been validated and a robust evidence base for child-specific risk factors is not yet available. For this reason, the introduction of risk assessment tools for children will need to occur as part of a well-designed, long-term framework that includes review and evaluation. This study found both strong support, and emergent practices, that provide a valuable starting point to establish clearer practice and guidance in this area of critical need.

The issue of how we integrate the development of specific risk factors for children in the context of the gendered nature of family violence risk and family violence more generally is complex. Pre-existing myths about family violence that have blamed and responsibilised women for their own experiences and for damaging their children prevail. Many participants here expressed concerns about how to develop specific risk factors for children given an ongoing tendency to blame or marginalise mothers, especially in relation to child protection and family court processes. While these concerns were not viewed as insurmountable hurdles to developing family violence risk assessment for children, they do point to the need to ensure that any reform in this area attends to the persistent operation of gender in the structures and institutions that address family violence.

Footnotes

Acknowledgements

The data for the wider study were initially published in . We would like to acknowledge all the colleagues who worked with us on that report.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding for the wider study upon which this article draws was provided by the Victorian Department of Health and Human Services.

Note

References

Australian Bureau of Statistics. (2005). Personal Safety Survey, cat no. 4906.0. Canberra: Australian Bureau of Statistics.

Australian Law Reform Commission & New South Wales Law Reform Commission. (ALRC/NSWLRC). (2010). Family violence – A national legal response. Canberra, Australia: Australian Law Reform Commission.

Bell

(n.d). Guidance for Trust’s Gateway Staff on how to use Barnardo’s multi-agency risk identification matrix (DVRIM) in assessing risks to children and young people from domestic violence and abuse. Barnardo’s Northern Ireland. Retrieved from http://onlinetraining.learningtoendabuse.ca/sites/default/files/resource_files/Gatewaytguidancesmall.pdf

COAG Advisory Panel. (2016). COAG advisory panel on reducing violence against women and their Children: Final report. Canberra: Commonwealth of Australia.

Commission for Children and Young People. (2016). Neither seen nor heard – Inquiry into issues of family violence in child deaths. Melbourne: Commission for Children and Young People.

Department of Human Services. (2012a). Family violence: Risk assessment and risk management framework and practice guides 1–3. Melbourne, Australia: Victorian Government – Department of Human Services.

Department of Human Services. (2012b). Assessing children and young people experiencing family violence: A practice guide for family violence practitioners. Melbourne, Australia: Victorian Government – Department of Human Services.

Gray

(2015). Finding – Inquest into the death of Luke Geoffrey Batty. Victoria, Australia: Coroners Court of Victoria.

Hazel

Hamilton

Jaffe

P. G.

Campbell

(2013). Assessing children’s risk for homicide in the context of domestic violence. Journal of Family Violence, 28, 179–189.

10.

Humphreys

(2008). Problems in the system of mandatory reporting of children living with domestic violence. Journal of Family Studies, 14(2/3), 228–239.

11.

Jaffe

P. G.

Campbell

Olszowy

Hamilton

L. H. A.

(2014). Paternal filicide in the context of domestic violence: Challenges in risk assessment and risk management for community and justice professionals. Child Abuse Review, 23, 142–153.

12.

Jaffe

P. G.

Crooks

C. V.

(2005). Understanding women’s experiences parenting in the context of domestic violence: Implications for Community and Court related service providers. Violence against Women Online Resources, http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.192.3729&rep=rep1&type=pdf (accessed 26 February 2018).

13.

Jeffries

(2016). In the best interests of the abuser: Coercive control, child custody proceedings and the “expert” assessments that guide judicial determinations. Laws, 5(1), 14.

14.

Jeffries

Field

Bond

C. E. W.

(2015). Protecting Australia’s children: A cross-jurisdictional review of domestic violence protection order legislation. Psychiatry, Psychology and Law, 22(6), 800–813.

15.

Junction Australia. (2015). Submission to the social development committee of SA parliament inquiry into domestic and family violence. Retrieved from http://junctionaustralia.org.au/images/pages/JA%20-%20Submission%20to%20Social%20Development%20Committee%20of%20SA%20Parliament%20Inquiry%20into%20Domestic%20and%20Family%20Violence.pdf

16.

Le Grand

(2015). Luke Batty finding Victorian system a failure. The Australian, 3 October.

17.

Mandel

(2013). The safe and together model: Making good decisions for children impacted by domestic and family violence. Retrieved from http://www.noviolence.com.au/sites/default/files/ResearchSeminarAdditionalFiles/davidmandel.pdf

18.

Mandel

(2014). Safe and together model: Overview and evaluation data briefing. Retrieved from http://ntv.org.au/conference/wp-content/uploads/2015-dm-Safe-and-Together-Model-Overview-and-Evidence-Briefing-October-2014.pdf

19.

McCulloch

Maher

J. M. M.

Fitz-Gibbon

Segrave

Roffee

(2016). Review of the Common Risk Assessment and Management Framework (CRAF): Final report. Victoria, Australia: Department of Health and Human Services, Victoria.

20.

Office of the State Coroner. (2014). Inquest into the death of Noelene Marie Beutel. Brisbane, Australia: Coroner’s Court Brisbane, Queensland Courts.

21.

Powell

Murray

(2008). Children and domestic violence: Constructing a policy problem in Australia and New Zealand. Social and Legal Studies, 17(4), 453–473.

22.

Queensland Family and Children Commission. (2017). A systems review of individual agency findings following the death of a child. Brisbane, Australia: Queensland Government.

23.

Richards

(2011). Children’s exposure to domestic violence in Australia. Trends and Issues in Crime and Criminal Justice, no 419. Canberra: Australian Institute of Criminology.

24.

Royal Commission into Family Violence. (2016). Report and recommendations. Victoria, Australia: Royal Commission into Family Violence.

25.

Segrave, M. (2017). Temporary migration and family violence: An analysis of victimisation, vulnerability and support. Melbourne: School of Social Sciences, Monash University.

26.

Special Taskforce on Domestic and Family Violence in Queensland. (2015). Not now, not ever: Putting an end to domestic and family violence in Queensland. Queensland, Australia: Queensland Government.

27.

Stanley

Humphreys

(2014). Multi-agency risk assessment and management for children and families experiencing domestic violence. Children and Youth Services Review, 47(1), 78–85.

28.

Stanley

Humphreys

(2017). Identifying the key components of a ‘whole family’ intervention for families experiencing domestic violence and abuse. Journal of Gender-Based Violence, 1, 99–115.

29.

Thiara

Humphreys

(2017). Absent presence: The ongoing impact of men’s violence on the mother–child relationship. Child & Family Social Work, 22, 137–145.

30.

Turner

Broad

Drinkwater

Firth

Hester

Stanley

, … Feder

(2017). Interventions to improve the response of professionals to children exposed to domestic violence and abuse: A systematic review. Child Abuse Review, 26, 19–39.