Introduction

Engagement with fathers who use violence has increased in numerous jurisdictions since the relatively recent shift toward a whole-of family-approach in statutory child protection (CP) agencies (Humphreys & Campo, 2017; Humphreys et al., 2019; Mandel & Wright, 2019). CP responses to families experiencing domestic and family violence (DFV) also increasingly utilize multi-agency integrated responses, which mobilize statutory and non-statutory organizations to intervene in high-risk cases (O'Leary et al., 2018; Humphreys et al., 2018). These practice developments bring higher numbers of workers from disparate agencies into closer contact with perpetrators of violence, which necessitates greater attention to worker safety as practitioners can be the targets of client-initiated intimidation, threats, verbal abuse and physical violence (Strolin-Goltzman et al., 2016). This is of particular concern for CP practitioners, but also for others working with victim/survivors and perpetrators of DFV (Button & Payne, 2009; Stover & Morgos, 2013; Strolin-Goltzman et al., 2016).

Despite the risk to workers, limited research investigates worker safety in Australian DFV cases (Humphreys & Campo, 2017; Strolin-Goltzman et al., 2016), including how workers experience client directed violence in the child welfare sector (Broadley & Paterson, 2020). There is also a gap in the literature on what workers need from their organizations to support them in their practice (Humphreys et al., 2019), or the range or adequacy of organizational responses (Broadley & Paterson, 2020). Further, the available literature makes generic reference to worker safety, with little consideration of whether it is likely to result in significant harm, or how the perceived level of risk may influence practice and outcomes for child and/or adult victim/survivors or workers. This article explores worker safety at the intersection of high-risk CP and DFV practice. Addressing a significant gap in the literature, the study examines the findings from a participatory action research project undertaken in Queensland, Australia. Fifteen senior practitioners and team leaders from CP, DFV, family support and justice services (police and probation and parole) came together to form a community of practice (COP) guided by the U.S. based Safe & Together Institute.

Background Literature

Violence against workers includes verbal abuse, physical assault and psychological harm. This section explores different types of violence against CP and DFV workers, the impact of client initiated abuse on workers’ physical and emotional safety, and how risks to workers are experienced and responded to when working with perpetrators of DFV. Following this, effective practice in CP and DFV cases is discussed.

Verbal abuse is the most prevalent form of aggression towards workers. The commonality of verbal abuse can result in it being de-emphasised, normalised and underreported. It can be subtle, insidious, and extensive, and therefore difficult to address (Robson et al., 2014). Psychological harm is, nevertheless, noted across studies on workers’ experiences of client violence and can include fear of future violence (Broadley & Paterson, 2020). While physical violence is less prevalent, when it does occur, it can have severe and ongoing consequences, resulting in physical and emotional harm (Littlechild et al., 2016). As Button and Payne (2009) note, working with domestic violence can be dangerous (p.365). In conjunction with the risk of client-initiated verbal and physical aggression, work in CP and DFV involves dealing with traumatic circumstances, which carries the risk of secondary trauma (Ben-Porat, 2017; Choi, 2017; Kanno & Newhill, 2009; Slattery & Goodman, 2009). Secondary trauma is characterised by a sense of helplessness, decreased confidence and lowered competence. It can impair knowledge retrieval and problem-solving abilities and decrease workers’ sense of workplace control, and thus adversely affect their practice (Ben-Porat, 2017).

Most violence towards DFV workers in women’s shelters is from clients’ partners (Kanno & Newhill, 2009). For their part, CP workers often experience risk when they attend homes where violent men may be present, which potentially exposes them to greater harm than office-bound workers. Unsurprisingly, CP workers often have reservations about engaging with families experiencing DFV due to the complexity, unpredictability, and potential for harm associated with co-occurring problematic substance use and mental health issues (Fusco, 2013). Although they express feeling inadequately trained for this work (Humphreys et al., 2019), child protection workers are reportedly more knowledgeable about DFV compared with workers in other social services (Button & Payne, 2009). However, inadequate knowledge on communicating the risk of lethality, managing their own safety, and working with perpetrators of DFV is also reported (Button & Payne, 2009).

Not knowing how to intervene safely with dangerous men, and failing to assess or communicate the potential for lethality, puts both workers and families at risk (Button & Payne, 2009). When workers feel unsafe, their decision-making may be affected – fear can reduce engagement with the family and compromise focus on the safety of women and children (Button & Payne, 2009; Broadley & Paterson, 2020; Littlechild et al., 2016; Mandel, n.d.; Robson et al., 2014). Additionally, workplace violence decreases efficiency, lowers morale and increases absenteeism, which further impacts the service received by those in need (Robson et al., 2014). While it is accepted that CP practitioners need more training on DFV perpetrator interventions (Fusco, 2013; Humphreys et al., 2019; Kanno & Newhill, 2009; Strolin-Goltzman et al., 2016), large caseloads and high turnover of staff can lead to chronic lack of training in this difficult and specialized area of practice (Button & Payne, 2009).

Further, workers rely on police for added protection; however, police often lack DFV specific training and fail to understand the dynamics in DFV (Fusco, 2013). For example, risk in DFV occurs on a continuum and often escalates in DFV cases where children are present and the mother and child(ren) have ceased co-residing with the perpetrator, or are seeking to leave (Thornton, 2017). At this critical time, the use of dangerous patterns of coercive control, which increases the risk of femicide (Thornton, 2017), is poorly considered in police risk assessment or responses (Wire & Myhill, 2018). Coercive control is used not only against women, but also against workers and systems to manipulate or distort information about women and children, particularly when men engage legal institutions, such as the family court (Douglas, 2018). This raises questions about the ability of police to predict serious incidents of DFV or lethality (Thornton, 2017). Concern has also been raised about the significant number of police who perpetrate DFV and how this may influence their interventions (Heward-Belle et al., 2018).

Working Effectively in DFV Cases

To work effectively in DFV, practitioners need to make an accurate assessment of risk in relation to family members and themselves (Stover & Morgos, 2013; Strolin-Goltzman et al., 2016). As perpetrators of DFV are a heterogeneous group, there are no universal solutions to addressing the risk they pose (Baynes & Holland, 2012). DFV risk assessment tools are often used to structure professional judgements and to provide guidance about risk factors; however, these tools are not predictive of future violence (Lamb et al., 2022).

A dynamic approach to assessing safety is therefore needed when working with men who use violence, requiring case-by-case judgment, with some structured guidance and supervision. This is enabled by recognizing threats and behaviour that can be a precursor to violence, determining perpetrators’ patterns of behaviour, undertaking evidence-based risk assessments, listening to the survivors’ own assessments of the dangerousness of the perpetrator (Stanley & Humphreys, 2014; Strolin-Goltzman et al., 2016), and recording all abusive incidents to aid identification of patterns of behaviour (Strolin-Goltzman et al., 2016). Most of these factors are embedded within the Safe & Together model, which is increasingly used to inform child-focused DFV interventions in the U.S. the U.K., Canada and Australia (Healey et al., 2018). The following section outlines how the Safe & Together model supported an action research project that explored responses to DFV and CP cases while simultaneously building the capacity of the workforce to intervene effectively with men who use violence.

Method

The present study formed a component of the larger Invisible Practices: Intervention with fathers who use violence project, funded by the Australian National Research Organisation for Women’s Safety (ANROWS) (Healey et al., 2018). This multi-site research project aimed to develop the domestic violence competency of workers engaging fathers who use violence. The research questions guiding the project were:

  • What do practitioners require from their organizations and/or other organizations to support them in working with fathers who use violence?

  • What evidence is there that the capacity building of communities of practice (COPs) supported by coaching and supervision by the Safe & Together Institute, provide increased experiences of safety and support for practitioners? (Healey et al., 2018, p. 12)

The project utilised a participatory action research framework, allowing participants to learn collaboratively from peers, facilitators and researchers, and to engage in a cycle of knowledge creation that further develops practice. In this process, dynamic situations that arise during the research are addressed through an iterative cycle of reflection and review (Wicks et al., 2008). Data were collected across four Australian states (Victoria, New South Wales, Western Australia and Queensland) in 2018. In each state, senior practitioners, team leaders and managers participated in 6 two-hour monthly COP workshops. Focus groups, which lasted between 45–90 min, were held at the end of the COPs. Across the states, 65 practitioners and two consultants from the Safe & Together Institute contributed to the research.

During the COPs, participants presented de-identified cases from their caseloads. These were discussed with Safe & Together consultants who provided input on best practice based on the Safe & Together model (Healey et al., 2018). Prior to the COPs, participants received online training plus two full days of face-to-face training in the Safe & Together model, which aimed to ensure a shared understanding of DFV. The research teams were present for the entirety of the training. The Safe & Together model is based on understanding perpetrators’ patterns of abuse and includes:

  • Assessing perpetrator behaviours that harm children;

  • Assessing multiple causes of harm to children;

  • Holding men to high standards as fathers; and,

  • Focusing on men’s parenting choices (Mandel & Wright, 2019, p. 121).

There are three core principles underpinning the model:

  • Keeping children safe and together with the non-offending parent;

  • Partnering with the non-offending parent; and,

  • Intervening with perpetrators to reduce risk and harm to the child (Mandel, 2014).

The COPs and focus groups were recorded and transcribed. Ethnographic notes were taken during the COPs. As the research was conducted across multiple sites, a note-taking template was devised. The template provided a theoretical and practical basis for the research and ensured consistency in data collection and analysis (Stake, 2013). Data were analyzed deductively to foster generalizability across the four research sites. The overall interpretation and synthesis of data was undertaken by the project’s senior researcher, with cross-checking by other team members. The University of Melbourne Human Research Ethics Committee (HREC) approved the study.

This paper reports specifically on data coded to worker safety from the ethnographic notes, COP discussions and focus group transcripts from the Queensland site, and does not include data from the other states. The decision to focus exclusively on the Queensland site was due to the unique make-up of the COP, which included 12 women and three men representing the following services: CP (n = 8); DFV (n = 2); family support (n = 2); and justice (n = 3). Queensland was the only state to include justice services comprised of Queensland Police and Probation and Parole (these services are grouped to protect participant anonymity). While participants in each state were asked to bring de-identified representative cases from practice to the COPs for discussion, the presence of justice services in the Queensland site led to an overrepresentation of high-risk cases, as noted among the wider research team. Participants were keen to avail themselves not only of the expertise of the consultant from the Safe & Together Institute, but also of Police and Probation and Parole officers. Further, cases seen by justice services tend to involve perpetrators who pose a greater threat of harm to women, children and workers, which influenced discussions in the COP.

As there is no consistent definition of high-risk in DFV cases (Lamb et al., 2022), a formal assessment tool was not applied in the study. Participants and the research team considered the cases discussed in this paper to be high-risk due to perpetrator histories of extreme violence towards adult victim/survivors, the risk of lethality for adult and/or child victim/survivors, and high levels of coercive control. These factors conform with the definition of high-risk in DFV cases outlined by the United Kingdom’s Home Office (2002): “A risk which is life threatening and/or traumatic, and from which recovery, whether physical or psychological, can be expected to be difficult or impossible” (Thornton, 2017, p. 69).

After data from the COPs and focus groups were coded to the theme of worker safety using the coding template, an experienced practitioner/senior researcher from the Queensland site reviewed the data inductively for sub-themes. This researcher read each line carefully before applying Braun and Clarke’s (2006) thematic analysis approach to the data. Codes were inductively determined without use of the pre-determined template and grouped to form themes. Prior to final analysis, each theme was cross-checked with all members of the Queensland research team and with senior researchers from the other states.

Findings

The language used in this section and in the discussion is gendered as all perpetrators discussed in each workshop across states were male and all adult victim/survivors were female. One of the six workshops explicitly focused on worker safety as a practice topic. However, worker safety was identified as a theme in all COP workshops and in the reflective focus groups. Five key themes were identified in the data: targeting of individual workers; the physical work environment; risk management involving police and the judicial system; the parallel between victim/survivors’ and workers’ physical and psychological safety; and strategies for improving worker safety.

Targeting of Individual Workers

Participants noted that dangerous men target individual workers from across the service spectrum. Some workers were at risk through their work in supporting children:

We’ve got high-risk perpetrators who would single out the Child Safety (statutory child protection service) Officer who’s trying to work with them around their children. (child protection worker)

Other workers were seen as an impediment to the controlling behaviours of violent men when they supported adult victim/survivors:

Because you are teaching the aggrieved how to get some empowerment or self-esteem. They don’t like that one little bit. (justice worker)

Threats were not confined to workers who directly engage with men. A worker from a women’s DFV service, who worked only with women and children, nevertheless also noted being an impediment to the power and control wielded by men who use violence, and the subsequent threat of retribution:

I don’t work with them, but I am exposed to them. I get multiple death threats, and the list goes on. Even though you don’t work with them, you become a threat, impediment to them.

The risk to safety extended to individual police officers. A child protection worker recounted:

She’s absolutely terrified of him…she’s that worried about him and in the conversation saying, “He’s a complete psychopath, we should all be really worried about him.” So, she’s police, and she’s scared. So even police have that worker safety, they’re the same…she’s got a gun and she’s frightened, and we don’t have guns and, now, he’s got our names on the card we left behind.

The Physical Work Environment

While all workers were potentially at risk, CP and family services workers who attend client homes were at heightened risk compared with workers who conduct meetings with clients in office-based settings:

Child protection workers, in general, have more of a focus on perpetrators, perpetrator accountability, working with them…people at Probation have done it for a long time, having done it in very well, as much as you can. Safe little offices, you know, whereas Child Safety are, of course, going…out to see these men at home, you know, or to see the mothers at home and the men are there hiding in the background. (justice worker)

While office settings were safer workplaces than family homes, they were not without risk. A child protection worker described inter-sectoral safety planning with the partner of a man with an extensive criminal history that included violence towards individuals beyond the family. The perpetrator’s use of technology to stalk victims and interrupt safety planning, even within the parameters of a child protection office, is evident in the following quote:

The collaboration was bringing in a DV service that works with the police, and getting them to come in with us to have a session, which was [held] outside. We couldn’t have it inside [the office] due to the bugging [by the perpetrator], and then, working with them to actually get her into a refuge space…we’ve got a photo of him so we can identify him…He was outside the office with a phone and he tracked [mother’s] phones…we’ve got great relationships with the police.

The quote below from a CP worker illustrates the additional danger of delayed professional communication when perpetrators and workers access the same physical space in a judicial setting. The worker’s comments also highlight the value of integrated responses structured to facilitate interagency decision-making and to enable swift action.

When we have these cases that are so hard that, it’s almost like we need that integrated response to happen, that people are on board to actually move very quickly, that we have a meeting and have everybody there. I think the other issue is the worker safety issue. Like, the other CSO (Child Safety Officer/CP worker) and I that went to do the statement and, then, knowing that he’s going to go to court tomorrow, but not knowing when he’s coming out of court.

Even the prison setting was seen as inadequate in providing safety for workers when high-risk perpetrators were likely to target them. A child protection worker reported:

He’s too dangerous and aggressive so I couldn’t sit in prison with him. Prison doesn’t actually, amazingly, doesn’t have great security. How it works is you sit in a little room with him and by the time I reached the buzzer, I’d be dead. And he’s very angry. Probably out of all of ours, he probably tops them all, this guy…He wants to get revenge on probably us and a lot of other people.

Risk Management—Police and the Judicial System

Participants highlighted that systems designed to protect workers are inadequate. The effective management of high-risk cases required conveying of information to police used to prosecute and convict men, which made individual workers potential targets for retaliatory actions. The account below highlights the danger in gaps in communication between CP workers and police, and the risks to individual workers when collaboration is poor and police delay action.

The Child Safety Officer, and myself, went to the home and we could see that mum had bruising on her neck and we had a notification to say that the daughter saw the father strangling her…There was a bail condition that Dad couldn’t attend the home…We told the police what happened…The next day he went to court for breaching the bail, and we thought he would be locked up, and he wasn’t…We didn’t know that, so that if I’m walking outside, that he just comes straight over. We had no worker safety and he knew that I was one of the workers and the other worker that breached him and he was really upset, as you can imagine, about that, and very angry. (child protection worker)

CP workers reported simultaneously needing to communicate the depth of legal complexity to police when men have issues pending in more than one court, along with the level of threat they pose, as indicated in the following quote:

They (police) actually had no idea that there was Children’s Court matters and there was actually Supreme Court bail conditions. They were just simply going in to hear the DVO (domestic violence order). So, we’ve had to try to explain this quite quickly: “No, no, no; you really need to be aware that there’s all these other things pending, and he’s a very dangerous man.”

Workers attributed such gaps in information to high workloads for police and the courts, and to lack of communication between various courts. A justice worker reported:

It’s (DV cases), so many, the prosecutor’s going in with a pile so high and it’s literally one after another…It’s the Supreme Court, we don’t have that information on our systems.

It was noted that the gap in communication between police and other service providers could be a deterrent in future reporting as workers may prioritise their own safety ahead of that of women and children, as noted by a specialist DFV worker:

We would now think about whether we would ever do that again in regards to breaching him using our statements.

While police responses were central to managing risk, at least two men discussed in the workshops were members of “motorcycle gangs” and had criminal connections that infiltrated police. At least one of these men also had connections in the federal government’s social security agency, which added further complexity to casework. Managing these cases involved moving from state to federal police and up the organisational ladder in the police force until workers believed they reached a level above corruption in police ranks. As reported by a child protection worker:

…once it got to police and corruption in the police force, I had to work around different police to get it to the right person in Federal to avoid all of that [corruption]. Each time anything was local, it would just get wiped, every time I would get it to a place and it would get wiped because of the corruption within the police force.

To manage cases with men whose criminality had infiltrated police, workers also escalated actions to more senior personnel within the statutory CP agency. A CP worker commented:

I’ve spoken out. The regional director knows about this case. It needs to be a higher up response than our response.

Parallels Between Victim/Survivors’ and Workers’ Physical and Psychological Safety

A parallel process between women managing their own and their children’s safety, and workers managing their own safety, while trying to ensure the safety of adult and child victim/survivors, was evident in the study. A family support worker explicitly drew attention to the parallel process that can develop when mothers and their associated workers deal with highly aggressive men, and the way in which each party is affected:

I shared with the team about the parallel process that women must have felt really exhausted to have been controlled like that, and how the workers too would have felt really exhausted.

The quote below from specialist DFV workers lends support to the notion of a parallel process in which workers’ emotions and actions mirror those of victim/survivors:

I’m mentally managing it, like her, and I think we all are. In this case, we try and give him some hope, you know, and we are part of that plan with her, you know, in some sort of collusion, in a sense, to try and dampen him down.

Participants stated that intervention, when not guided by women, potentially increases risk for mothers. They discussed changes to the well-established practice of removing children from homes without consultation with the mother, and described leaving them with their mothers while they worked to resolve safety for adult and child victim/survivors by partnering with mothers.

An interesting thing that was said in the integrated response to DV meeting last week is that the Child Safety officer said that this is the first case in a while where they’re actually taking her lead. They said that they should have removed the kids potentially a while ago, but they didn’t because they took the mum’s lead on what that risk was like. If they did remove the kids she was like, “Well, I’ll be dead, so you’ll just be signing that.” (child protection worker)

While they expressed conviction that partnering with mothers to improve safety for women and children improved practice, child protection workers noted that delaying removal of children in high-risk situations affected their own psychological safety. Some workers held lingering fears for women and children’s safety after cases closed, and discomfort at leaving women in homes with violent men after children were removed. A Child Safety Officer stated:

It makes us freaked out around closure and what’s going to happen.

Taking the woman’s lead could involve not alerting police, which left workers to manage high-risk cases in ways that significantly raised their anxiety. The following quote about a child left with a perpetrator of extreme violence illustrates the difficulty in managing cases in which men have infiltrated police, and the fear child protection workers experience in relation to the consequences of their decision-making:

We were following her safety plan…Child Safety was thinking, “Well, we should contact the police, but she’s saying to us, “Do not contact police about where I’m going because he’s got connections.” So we didn’t...but, that will be something, if somebody dies, that they will come back to us and say, “Why didn’t you deal with that? Why didn’t you contact who was supposed to deal with that?”

Workers also drew attention to divisions within the police that reduced collaborative efforts with high-risk offenders, along with poor communication within police, and between police and other workers.

In the police, it’s almost like a hierarchy, you know; like, really important policemen can deal with the gangs and the murders and those things, and because this guy is broaching both, it means that some of our communications with our normal people maybe isn’t hitting the mark or it is, but we just don’t know it is.

Improving Worker Safety

Participants made recommendations for improving worker safety, noting that less experienced workers were the least aware of risks in practice. A family services worker considered a junior colleague’s enthusiasm for promoting inclusive practice with dangerous men ill considered:

Our worker says, “Let’s ring dad and see if we can get him to come along”, and I’m thinking, “No, let’s not”…keen workers, excited workers, wanting to do the right thing.

A family services worker’s comments highlight threat when engaging men in difficult conversations:

You have workers…going into the lions’ den and talking about accountability, and pushing buttons, and sometimes not knowing what they are doing. The risk to safety is quite high…When you are at the pointy end, you have two people going in to talk about some really hard stuff to someone who is king of their domain. So, we’ve had him get really angry, and the worry is that something might happen to a worker just because they are having a conversation…We are just going to go in and have those conversations and somehow that is going to be safe? I don’t know if it is.

Participants outlined various strategies, some internal to their organization, and others external, they employ to increase safety. For example, child protection workers sent texts to colleagues to warn them when a high-risk perpetrator was in the building and, when space was available, parked their cars in the secure staff car park rather than on the street. Workers from child protection and family support services commented on not always knowing what the perpetrator of violence looks like:

We have a lot of cases where the Child Safety Officer will get a lot of harassing phone calls but not know what he looks like; then, they are out on the street thinking, ‘Is he looking at me?’ (child protection worker)

Workers wanted to see interagency collaboration extended to include a focus on worker psychological safety. The quote below from a child protection worker demonstrates a siloed response to critical incidents such as a child’s death when cases are reviewed internally only. The comments highlight that individual agencies supporting the psychological wellbeing of their own staff can restrict open discussion and facilitate a culture of blame, which hampers exploration and learning and, by association, practice.

You tend to see those things [mental health support] as internal to your own agency. It would push us towards saying, “We collaborate towards our own safety and our own wellbeing.” Because when something goes wrong and, say, a child dies…then everyone goes off and [debriefs] in their own agencies. Because, one of the risks of doing it in your own agency, is you can say the reason it went wrong was because of [another agency]; whereas, when you are all in the room together, it is processed properly.

A justice officer pointed out that safely managing risk in “those really pointy cases” is time-intensive and requires a team-effort, but that workers can be directed by management to process cases more quickly, which potentially increases risk for victim/survivors and workers:

...practice takes a lot of thought; it takes a lot of planning. With this one, the team leader was like, “You can’t spend any more time on this case, we’ve got all these other jobs,” but…the workers’ safety, the client’s safety…We spent ages talking about it, they’re working out the, you know, you need some of those devil’s advocates saying, “What about this” and, “What about that”, and there’s just no time.

Discussion

The present study was a participatory action research project designed to investigate and develop workforce capacity at the intersection of DFV and CP concerns. The purpose of this paper was to explore worker safety in a subset of data from the Queensland site, as this was the only site to include police and probation and parole officers, which resulted in a disproportionate number of high-risk cases. The key questions guiding the overall research were:

  • What do practitioners require from their organisations and/or other organisations to support them in working with fathers who use violence?

  • What evidence is there that the capacity building of communities of practice (COPs) supported by coaching and supervision by the Safe & Together Institute, provide increased experiences of safety and support for practitioners? (Healey et al., 2018, p. 12)

Initially, the findings answer the second research question, which is concerned with workers’ experiences of implementation of the Safe & Together model in the context of their practice. The first question, which focuses on workers’ requirements for safety, is discussed after workers’ accounts of their safety and support are presented.

Practitioners’ Experiences of Safety and Support in High-risk DFV and CP Cases

The workers in the present study were operating in an occupational environment that required them to alter practices by adopting the key principles of the Safe & Together model, which includes keeping children safe and together with the non-offending parent, where possible. Abiding by this tenet required workers to allow children to remain longer in homes with dangerous men while they partnered with mothers to manage the safety of adult and child victim/survivors. Previous practices would have seen children removed from dangerous situations earlier. Working in a timely, but slower, pace to accommodate both the mother’s safety, as well as the child’s, and to keep them together in the process of securing a safer outcome, is a more nuanced practice. However, the practice affected workers’ own psychological safety through concerns that their actions or inactions might result in serious harm to women and children, as also noted in Fusco’s (2013) study of child welfare workers’ experiences of intervention in cases of intimate partner violence.

The literature indicates that fear of perpetrators (Stover & Morgos, 2013) shows up in subtle and not so subtle ways such as lacklustre attempts at engagement with perpetrators of violence or avoidance of home visits (Hunt et al., 2016 cited in Broadley & Paterson, 2020; Mandel & Wright, 2019). Keeping children safe and together with the non-offending parent increased visits to homes where angry and revengeful men were resident. While workers considered these men to be high-risk perpetrators of violence, there was no indication of them avoiding home visits in order to prioritize their own safety and wellbeing over that of women and children. This finding aligns with Baynes and Holland’s (2012) child protection case file review, which found that violent men were as likely to be visited as non-violent men. During these visits, workers could be unsure of who was present, or what the perpetrator of violence looked like, while leaving evidence of their own identity in the form of businesses cards. Home visits also gave workers access to information used to convict perpetrators. Given they reported that perpetrators are likely to target them personally, their anxiety needs to be acknowledged and respected as the response of highly committed professionals operating in an occupational environment prone to risk.

To manage risk, workers took the woman’s lead, which shifted practice away from well-established organizational norms whereby workers dictate actions to women experiencing DFV (Mandel & Wright, 2019). Through careful attention to the experiences and strategies of victim/survivors, participants deepened their own understanding of managing risk in DFV, and used this knowledge to inform and improve their own safety. Coercive control often featured in these situations for both the victim/survivor and the worker. The Safe & Together Model places emphasis on patterns of perpetrator behaviour that help to expose coercively controlling actions. Mandel (n.d.) states, “Beyond avoidance, unaddressed worker safety concerns can manifest as collusive or placating behaviour” (p.3). The findings from the present study indicate that what can appear to be collusive or placating behaviour on the part of experienced workers can be a purposeful strategy undertaken in a parallel process with the actions of the victim/survivor. In the present study, this parallel process extended to workers’ experiences with the police and the judicial system. The findings indicate that police and the judiciary need to know that coercive control is also used against workers when men who use violence manipulate legal systems.

The Safe & Together model directs workers to pivot to the perpetrator of violence. The principle refers to making perpetrators visible, rather than necessarily engaging them directly, which may be ill advised with men who have committed extreme acts of DFV, as some of the men discussed in the present study had. When risk is high, women experiencing violence (Douglas, 2019), and CP workers (Fusco, 2013), rely on police for protection. Yet, a number of reasons, including fear of retaliation from the perpetrator, and the belief that they are not necessarily in physical danger, prevent women from calling police (Douglas, 2019). The findings from the present study show that in high-risk cases, when perpetrators have infiltrated police and public institutions such as social security departments, workers can also be reluctant to call police. While there were reports in the present study of “great” working relationships with police, the findings nevertheless indicate that police responses in high-risk cases can further jeopardise the safety of workers called upon to provide information leading to conviction and subsequently left without protection (Broadley & Paterson, 2020). Understandably, workers, like women, can find themselves unprepared to call police when they fear that police cannot be relied upon (Douglas, 2019), or trusted (Heward-Belle et al., 2018), again indicating a parallel process between women’s and workers’ experiences.

Practitioner Requirements for Support in Working with High-risk Perpetrators of DFV

Guaranteeing worker safety in DFV and CP cases, particularly in high-risk situations, is an impossibility; nevertheless, actions can be taken to mitigate risk and improve safety. The discussion below initially focuses on intra-organizational issues in addressing worker safety prior to turning attention to inter-organizational and systemic factors. The parallel process between women managing their safety and workers managing their own safety suggests there is a lot that service providers can learn from the strategies women employ to protect themselves and their children. For this to meaningfully happen, it is essential that workers allay the mother’s fears that she will be accused of not being a good mother and partner with her to understand the risk she faces and the ways in which she keeps herself and her children safe (Mandel & Wright, 2019). This information could inform practitioner training and the education of social work and human services students. However, there is the caveat that women’s own assessments are not always a reliable gauge of their safety, as demonstrated by Campbell’s (2004) finding that only about half of the women killed, or almost killed, in DFV incidents accurately predicted the level of risk they faced.

Several studies have examined how workers’ personal experiences of DFV can undermine practice and have identified the need for workers to be supported in forming healthy coping strategies in order that they can work competently and safely with families (Choi, 2017; Fusco, 2013; Slattery & Goodman, 2009). Studies have also called for training for child welfare and DFV workers (Fusco, 2013; Humphreys et al., 2019; Kanno & Newhill, 2009; Robson et al., 2014) and their managers (Littlechild et al., 2016). Stover and Morgos (2013) recommend training and supervision on personal risk assessment and management, mitigating escalated situations, and self-defence, should workers become unsafe. As senior workers, the participants in the present study were acutely aware of their own safety – and of the limits to that safety. However, there were instances when discussion turned to the behaviours of inexperienced workers less able to assess their safety. The latter group of workers require training and supervision not only on working with men as parents (Mandel & Wright, 2019), but also on the hallmarks of high-risk perpetrators and in knowing when not to attempt to engage fathers. Further training could be targeted to additional gaps in worker knowledge identified in the literature: mental health problems associated with DFV for both victims and perpetrators and the warning signs of perpetrator lethality (Button & Payne, 2009). As evidenced in the present study, perpetrators’ use of technology has brought new challenges for worker safety; it is, therefore, also an important avenue for further education and training. Beyond training, the literature recommends clinical supervision to increase practitioner competence and confidence in working with families in which there is DFV (Choi, 2017) and the provision of counselling and support (Broadley & Paterson, 2020).

While no doubt useful, the factors mentioned above largely maintain the onus for safety on individual workers. The literature also identifies a number of strategies for addressing organizational responsibility for worker safety. Intra-organizationally, an open culture (Choi, 2017), supportive colleagues, and an environment of respect, equality, trust and empowerment have been found to mitigate the risks of secondary trauma for practitioners working with domestic violence (Slattery & Goodman, 2009). These organizational qualities are critical for workers living with the psychological toll of delaying the removal of children when partnering with the non-offending parent to secure safety for child and adult victim/survivors.

DFV and CP workplaces typically have workplace safety programs in place that rely on management working closely with practitioners to ensure open communication and action on safety concerns (Strolin-Goltzman et al., 2016). The present study indicates that some policies and practices need to be reconsidered; for example, the inclusion of workers’ full names on email addresses and business cards. Organizations also need to account for the time taken to manage high-risk cases safely, which requires smaller caseloads. While allocating high-risk cases to specialist teams means that some workers will disproportionately carry risk, the complexity of high-risk CP and DFV cases requires interdisciplinary inter-organizational practice, as discussed below.

Most Australian jurisdictions have multi-disciplinary high-risks teams (variously known by their acronyms, RAMPS, SAMS, MARACS). These teams are comprised of multidisciplinary staff from the various agencies that intersect when CP and DFV concerns are present in families – men’s and women’s DFV services, CP and law enforcement, as practiced in integrated responses (O'Leary et al., 2018). These teams use standard risk assessment tools to assess risk; however, the accuracy of assessing high-risk perpetrators using standard risk assessment tools has been questioned, and the tools do not attend to worker safety (Lamb et al., 2022; Thornton, 2017), or address the risk of involvement with systems operating to protect women and children (Heward-Belle et al., 2018).

Recent developments that extend beyond incident-based assessment to include perpetrators’ patterns of behaviours and the use of coercive control could be applied to worker safety (Mandel & Wright, 2019; Wire & Myhill, 2018), and conducted collaboratively across agencies. According to Mandel and Wright (2019), “By examining a perpetrator’s past responses to outside interventions by police and others, child welfare can better anticipate and plan for any danger a perpetrator may represent to their own staff” (p. 134). There is also potential to leverage current work in Australia to better identify and act in situations of coercive control. Several jurisdictions are considering legislation to address the insidious nature of coercive control due to its association with lethal outcomes. Queensland plans to introduce such legislation in 2023.

Perpetrator mapping in multi-disciplinary and multi-agency teams could beneficially conceal the source of information used to convict high-risk offenders thereby making it harder for them to identify and target individual workers. In reviewing the literature on violence towards workers in the child welfare sector, Broadley and Paterson (2020) note that, “When the organization takes responsibility for reporting incidents to the police, it takes the burden off individual workers who may feel reluctant to press charges; for example, because they are fearful of retaliation from the client” (p. 23). Well-functioning teams within integrated responses increase safety for women and children (O'Leary et al., 2018) and may reduce the propensity for blame when harm comes to child and/or adult victim/survivors. A reduction in actual or perceived blame would likely improve the psychological wellbeing of workers. For multidisciplinary teamwork to effectively support child and adult victim/survivors, and prevent workplace violence towards practitioners, strengthened collaboration (Broadley & Paterson, 2020) and leadership (Humphreys et al., 2019; Strolin-Goltzman et al., 2016) at the interface of DFV and CP practice is required.

However, practice at the intersection of DFV and CP is embedded in complex legal systems that have been described as lacking in safety for both women and children (Fusco, 2013; Heward-Belle et al., 2018). While dissatisfaction among child welfare workers with how police intervene in DFV is reported in the literature, this concern is in relation to the safety of victim/survivors (Douglas, 2019; Fusco, 2013; Heward-Belle et al., 2018), rather than workers themselves. Literature from the United Sates (Fusco, 2013), the United Kingdom (Thornton, 2017; Wire & Myhill, 2018), and Australia (Douglas, 2019) supports the notion that police lack training on DFV and do not know how to effectively intervene. The findings from the present study indicate that international calls for training and education for police (Douglas, 2019; Fusco, 2013; Thornton, 2017; Wire & Myhill, 2018) should extend beyond the use of risk assessment tools focused on the safety of victim/survivors and include attention on the safety and wellbeing of workers across disciplines. Discussion in the COPs demonstrates that assessment of high-risk perpetrators needs to include history of criminality and, if identified, that assumptions should not be made that police involvement will be universally benign or helpful.

Further, the present study indicates that in high-risk cases the problem can be wider than lack of training or education for police, or the need for cultural change within the police force (Douglas, 2019), and may include corruption. In the present study, communication with police was both extensive and guarded. When attempting to circumvent police corruption, CP workers moved from state to federal police and escalated cases up organizational ladders. The present study also illustrates increased risk for workers when perpetrators are involved with multiple courts at state and national levels. Communication between courts was limited, a situation that facilitates and exacerbates coercive control against women, children (Douglas, 2018), and their associated workers. Inter-organizational worker safety guidelines and protocols when multiple services (Humphreys & Campo, 2017) and different courts are dealing with high-risk violent perpetrators would help to mitigate abusive behaviours. Addressing police corruption requires political will, investigation, and accountability at the highest levels of policing and the judiciary.

Limitations and Strengths

There are a number of limitations in this exploratory study. The study draws on a small sample size in one jurisdiction in Australia; it is therefore not possible to generalize the findings. The participatory action-research methodology may have resulted in reticence by participants to share details about some aspects of practice. For example, shame may have prevented workers from disclosing avoiding home visits. To overcome the limitations of data drawn from a COP and focus groups, future research could include a larger sample size with different methodologies such as the use of surveys. Individual interviews could also enable deeper exploration of some of the key findings from the present study. Further, this research did not include the perspectives of victim/survivors; as such, it is not possible to know how clients perceived workers’ concerns or actions. Additionally, an intersectional lens was not included. Despite these limitations, the study helps to fill a gap in the literature. There is minimal research documenting women’s experiences of police intervention in DFV (Douglas, 2019). There is even less research documenting workers’ experiences of police actions in DFV cases (Humphreys et al., 2018). There is also little attention to public institutions as settings of heightened risk and anxiety for workers. The present study illustrates issues and practices in high-risk DFV and CP cases involving multiple professions and agencies across different practice settings. This information is important for guiding interventions and designing workplace practices that support safety (Humphreys et al., 2018, 2019).

Conclusion

The present study used participatory action research methodology to investigate and develop cross-sectoral workforce capacity at the intersection of DFV and CP concerns. This paper reports on worker safety among practitioners involved in high-risk cases in the Queensland COP. The study identified risks to worker safety associated with the physical environment, advances in perpetrators’ use of technology, and failings in responses from police and the judicial system. In managing risk, a parallel process between workers’ responses to threats to their physical and psychological safety, and those of women experiencing DFV, was evident in the study.

The value of building workforce capacity through training, resources, and coaching from the Safe & Together Institute was borne out in the study. While implementation of the model brought challenges for the physical and psychological safety of workers, it also introduced strategies for addressing worker safety, including learning from women, and using perpetrator mapping as a shared task across agencies and sectors. Perpetrator mapping and information sharing take time, which needs to be factored into workloads (Humphreys et al, 2019). Policy and practice guidelines need to be adapted accordingly, with intra-organizational change alongside inter-organizational change.

The literature generally discusses worker safety at the intersection of DFV and CP in an aggregated way. There is little, therefore, to guide interventions with high-risk perpetrators specifically. The study findings indicate that high-risk cases of DFV and CP need to be managed differently with escalation up organizational structures across core agencies, particularly when perpetrators have criminal connections and may have infiltrated police and other government services. Reports that information sharing is not a universally positive or safe practice in relation to high-risk perpetrators of DFV is an important contribution to the literature. Further research could usefully identify other examples of interventions with high-risk perpetrators in other jurisdictions to examine how they are managed.