‘It was a challenge to look at things from a perpetrator perspective’. The problem of holding domestically abusive men to account in multi-agency partnership work

Abstract

One of the most challenging aspects of multi-agency partnership work aimed at reducing domestic abuse is the problem of holding perpetrators to account. Drawing on findings from our recent mixed methods evaluation of a revised approach to multi-agency policing of domestic abuse, this article explores this problem from the perspective of the multi-agency practitioner stakeholders. We present evidence attesting to the challenge these stakeholders are experiencing as they strive to ensure they play their part in ensuring the perpetrator is accountable for their behaviour. We review what is known about male perpetrators and perpetrator programmes and examine our findings in the context of the wider challenges of engaging perpetrators in behavioural change. We argue that the perpetrator is the elusive stakeholder in multi-agency partnership work to reduce domestic abuse and that the web of accountability requires strengthening if domestically abusive men are to be held to account.

Introduction

In recent years, there has been an upsurge in prevention measures targeted at male domestic abuse perpetrators. These preventive approaches have arguably made a discernible impact manifesting in tangible measures and strategies, as well as legislative changes providing legal options and court proceedings, such that perpetrators can be held to account and ‘responsibilised’ for their abusive behaviour. Specific developments include closer monitoring of patterns of behaviour, and enhanced police surveillance of perpetrators, and the sharing of perpetrator related behaviour through disclosure schemes and the collection of criminal activity and broader intelligence data (Salter, 2014). Other tools for change have been proposed in the criminal and family courts in Australia (Spencer, 2016). Each of these examples signal a joined-up pincer movement, whereby informal and formal services, support providers and networks together hold perpetrators to account. The ideology behind such developments, as well as the efficacy of disclosure schemes in general, have been subjected to various levels of critique (see Barlow et al., 2023; Duggan, 2018; Duggan and Grace, 2018; Fitz-Gibbon and Walklate, 2017; Hadjimatheou and Grace, 2021). Nevertheless, preventive ideologies continue to gain momentum as part of multi-agency and partnership work to tackle the problem of domestic abuse. Indeed, it is widely recognised that such measures and schemes are part of a wider toolkit and ambition for a holistic strategy to tackle violence against women and girls more broadly.

Drawing in the main on research exploring one recent innovative approach, and bringing forth additional insights gained from our previous evaluations of multi-agency working to tackle domestic abuse, this article explores the problem of holding male domestic abuse perpetrators to account within the context of multi-agency working. First, we set the context in terms of the dominant way of operationalising policing of domestic abuse through multi-agency partnership approaches. Second, we explore the context of what we know about the management of male perpetrators and perpetrator programmes. We then introduce Spencer’s (2016) concept of the ‘web of accountability’, before outlining our study and methodological approach. A thematic presentation of our findings on perpetrator issues from the practitioner stakeholder perspective follows, and the final substantive section engages the reader in a reflective discussion on the question of holding domestically abusive men to account, arguing that this is the most challenging aspect of multi-agency working to tackle domestic abuse in the 2020s.

Multi-agency policing of domestic abuse

Multi-agency partnership work continues to dominate the landscape for policing domestic abuse at international, national and local levels. However, despite over four decades of ‘policing’ domestic abuse in England and Wales through the multi-agency partnership approach, domestic abuse remains a prevalent and stubborn issue. Prickly questions have recently been posed about our continued faith in the multi-agency approach to reduce domestic abuse, to protect and support survivors, and to hold perpetrators to account. The present climate of financial instability and austerity and the longer term direction for multi-agency and partnership working looks set to remain precarious and concerning (Davies, 2022). The imperative to make a sustained positive impact to reduce the risk to victimisation cannot be overstate.

Multi-Agency Risk Assessment Conferences (MARACs) were established in the wake of the 2004 Domestic Violence, Crime and Victims Act as a way to form a co-ordinated response to address high-risk cases of domestic abuse. MARACs now number around 300 in England and Wales, yet are struggling to deal with the increasing number of cases referred to them. Despite the concerns about maintaining faith in multi-agency approaches, and the increasing number of high-risk cases coming to MARACs, multi-agency partnership working is not on the wane and innovations are plentiful in global, national and local level efforts to prioritise this widespread problem. Examples of such innovations include MARAC developments to enhance the targeting of serial perpetrators and high-risk cases of domestic abuse.

Additional layers of multi-agency coordination interventions have been pioneered in recent years, where abusive behaviours are monitored and where the dynamic and sometimes fast changing level of risk is captured, so that information can be swiftly shared and acted upon. Examples include a Multi-Agency Tasking and Co-ordination (MATAC) approach, as pioneered in the North of England and new MARAC processes in several other jurisdictions in England and Wales (Davies et al., 2023). These innovative partnerships tend to have a strong shared vision among all stakeholders and aim to promote a whole system approach to tackling domestic abuse. This ‘holistic’ strategy can present challenges for domestic abuse partnership work (Davies, 2018), and one of the most challenging aspects of this partnership work is the problem of holding domestically abusive men to account.

Focussing on perpetrators

Having established our primary focus is on the problem of holding perpetrators to account and the faith placed in the efforts of multi-agency and other more informal players to change men’s abusive patterns of behaviour, we want to further contextualise this focus. Despite the various ways of capturing information about the experience of domestic abuse, understanding its prevalence remains difficult. Police recorded cases are unreliable as the sole or primary measure, and police recorded crime data are not designated as National Statistics. In England and Wales, approximately four in five, or 80% of victims do not report their experience to the police (Office for National Statistics (ONS), 2019), though police do respond to around 1 million domestic abuse incidents every year in the United Kingdom. Of all crimes recorded by the police in the year ending March 2021, 18% were domestic abuse-related (Office for National Statistics, 2021). It is well documented from a variety of sources, including statistical data from the police, the Office for National Statistics (ONS), Ministry of Justice (MoJ), courts, crime and victimisation surveys such as the Crime Survey for England and Wales (CSEW) and charitable organisations, that the volume of incidents of domestic abuse are staggeringly high and these extensive rates hold fast over time and across jurisdictions around the globe. The most recent CSEW estimates that 2.4 million people experienced domestic abuse in the last year. This is a prevalence rate of approximately 5 in 100 adults. Violence against the person had the highest proportion of offences identified as domestic abuse-related (34.4%; ONS, 2022a). In the same period, there were 234,045 domestic abuse-related stalking and harassment offences accounting for a quarter of all domestic abuse-related crimes (ONS, 2022a).

These patterns of abuse are gendered with the overwhelming percentage of cases showing a pattern of men abusing women. An estimated 6.9% of women (1.7 million) and 3.0% of men (699,000) experienced domestic abuse in the last year (ONS, 2022b). The victim was female in 74.1% of domestic abuse-related crimes. The nature of the abuse is all too frequently fatal. As is the case in England and Wales on average, at least one woman a week is killed by a partner or former partner. In terms of domestic homicide, between the year ending March 2019 and the year ending March 2021, 373 domestic homicides were recorded by the police, representing about 1 in 5 of all homicides where the victim was 16 or above. In that period, 72.1% of victims were women compared with 12.3% of victims of non-domestic homicide.

The pool of perpetrators is vast and though offenders share abusive patterns of behaviour, the nuanced data suggest they are not a homogeneous group. Mounting evidence from Sherman et al. (2016) and more recently others indicates that a small sub-set of serial, highly prolific perpetrators and repeat family violence abusers are responsible for the majority of the known harm reported (Barnham et al., 2017; Bland and Ariel, 2015; Davies and Biddle, 2018; Robinson and Clancy, 2021; Sherman et al. 2016). We would like to note that Sherman and colleagues’ characterisation of the felonious few versus the miscreant many is for us an odd way of differentiating these sub-sets of offenders, in that it carries forth the implication that victims of the miscreant population of perpetrators experience less harm. The influential ‘Power and Control Wheel’ emerging from the 1980s ‘Duluth approach’ (originating in Minnesota, the United States) is always a useful reminder for understanding that abusive behaviour is patterned, it is repeated and it is gendered in nature. Research further reveals that many perpetrators repeat their violence in future relationships (Hester and Westmarland, 2005) and multi-perpetrator domestic abuse is not uncommon, that is, domestic violence against women in large proportion involves multiple perpetrators (Salter, 2014).

One innovation recently reported on – also referred to above – was the MATAC, where a sub-set of perpetrators is regularly extracted according to a definition of serial perpetration. A serial perpetrator of domestic abuse for the purposes of that approach is a person who

• Is listed as a domestic abuse suspect or offender;

• Has more than one domestic abuse incident;

• Has offended against two or more different domestic abuse victims (this includes partner/ex-partners, familial relationships, non-violent incidents and nominals aged 16 and above within the last 2 years; see Davies, 2018).

As much research increasingly supports the proposition that perpetrators are a non-homogeneous group and reports on different perpetrator populations, it is important to foreground the commonalities to centralise the dynamic nature of risk and the potentially fatal outcomes. We pick up on this thread in our consideration of perpetrator programmes later in our discussion. Systemic and relentless pursuance strategies and focussed deterrence strategies are increasingly being mobilised by police and multi-agency partnerships who are searching for safe, plausible, and defensible ways of allocating scarce protection and support resources. These partnerships are refining the tools they use to identify domestically abusive behaviours and assess risk while re-assessing the adequacy of their focus on perpetrators. Although not commonplace, there is a history of the use of domestic violence perpetrator programmes (DVPPs), which we will explore next.

Domestic violence perpetrator programmes (DVPPs)

DVPPs were popularised in the 1980–1990s and proliferated across North America, Europe and elsewhere (Phillips et al., 2013). These programmes are for male perpetrators and vary according to where in the world they are developed. Some follow the pioneering Duluth model, while others provide more therapeutic and cognitive behavioural approaches. Furthermore, there are variations in how perpetrators become enrolled on such programmes. Some perpetrators are directed towards the programmes through a court mandate and thus attendance is forced, while other participants on the same or different programmes may embark on the programme as a voluntary service user (Renehan, 2021). These variations all make evaluation of ‘what works’ complex, and comparison of studies problematic. There is also an accompanying ideological critique of such programmes, especially those that veer towards the purer cognitive form of anger management (Renehan, 2023). Nevertheless, large numbers of such interventions exist which broadly aim for lasting behavioural change, and such programmes remain the primary service provision offer for men who are violent and abusive towards women. Programmes are premised on the belief that men can change and unlearn their behaviour and the effort to engage men in this way is part of the current holistic package that aims to reduce and prevent victimisation (Dobash et al., 1999; Dobash et al., 2000).

In the wake of their reporting of widespread scepticism about the effectiveness of DVPP programmes in 2012, Kelly and Westmarland later reported on their own longitudinal programme of research – Project Mirabal – which investigated whether DVPPs offered in the non-governmental organization (NGO) sector work to reduce men’s violence and abuse and increase the freedom of women and children (Kelly and Westmarland, 2015). The quantitative and qualitative data from this study showed steps towards change for the vast majority of men attending DVPPs. Physical and sexual violence was not just reduced but ended for the majority of women in this research. Everyday abuse and harassment and coercive and controlling behaviour was more difficult to curtail. A key concluding argument of Kelly and Westmarland’s (2015) research is that DVPPs are not considered a panacea. They identified a continuum of change among the men, with some taking minimal steps towards change, and others – from the accounts of their partners – moving a considerable way.

Since this work, NGOs have continued to be affected by austerity measures. Precarity in this sector leaves us knowing less about the current offer of support and the effectiveness of it. Evaluations of provisions more recently tend to be provider specific, based on short-medium term commissioning of providers all of which obscures the longer term continuities and discontinuities, as well as the wider wraparound problem this article addresses. There have also been further significant policy developments in the United Kingdom, such as the creation of the Domestic Abuse Act (2021), with perpetrator work being acknowledged as a significant step to tackling domestic abuse. Holding more perpetrators to account and supporting them to reduce their domestic abuse is certainly necessary. Building Better Relationships (BBR), developed as part of the Transforming Rehabilitation agenda, aimed to achieve such outcomes though to date there has been no co-ordinated evaluation. BBR marked a shift away from feminist informed practice towards a General Aggression Model (purportedly) informed by desistance-focused research; emphasising the importance of the therapeutic alliance in the client/practitioner relationship and adopting a strengths-based approach. BBR has never been evaluated bar Renehan’s (2021) in-depth ethnography that highlighted a number of issues with the efficacy of this recent, and widely used, perpetrator programme, which deviates from many of the approaches outlined in the previously discussed Project Mirabal study. Renehan (2021, 2023) found that BBR was not responsive to a diverse perpetrator population and not necessarily redressing male perpetrators’ reasons for violence. The programme focussed on teaching men to manage their reaction to conflicts, irrespective of the many complex needs, personal and emotional vulnerabilities, and neurodivergence factors that many of the men faced. Facilitators were attuned to the reality that they were working with traumatising but also traumatised men, but they said they were less able to support them with complex needs or the difficult feelings they raised.

In sum, to achieve the maximum benefits from perpetrator programmes, a gendered, individualised and needs-based approach should be adopted, recognising the trauma, mental and emotional health issues that perpetrators themselves also may experience (Hughes, 2023; Renehan, 2023). Although it is well documented that working with domestically violent men is challenging work – not least because of their tendency to deny, minimise, and blame (Renehan, 2021), the needs-based requirement demands we acknowledge that many men who are violent will also have lived experience of various forms of harms and victimisation/criminalisation as children. Nevertheless, perpetrator accountability and responsibility is key to addressing men’s abusive behaviour. Spencer’s (2016) ‘web of accountability’ is a useful framework within which to situate such efforts.

The web of accountability

As outlined so far, the state of knowledge about ‘what works’ in the area of domestic abuse perpetrator intervention is evolving. It has long been recognised that perpetrator accountability is key to victim safety. One such way of conceptualising this approach is Spencer’s ‘web of accountability’. According to Spencer (2016),

A web of accountability comprises various strands including the actions of legal systems (criminal, civil, child protection and family law), service systems and informal networks of victims, families and communities that together hold the perpetrator to account by intervening and monitoring ongoing behaviour. Women are much more able to assert themselves to hold men accountable for abuse when a ‘web of accountability’ provides both informal and formal support. (p. 225)

Key components to effectively managing perpetrators within this context include: continuous risk assessment and safety planning; consistent support for victims with a focus on safety; early identification and referral of perpetrators into evidence-based, individualised and needs-led DVPPs with regular monitoring of engagement; timely information sharing about risks and needs and effective case and domestic abuse knowledge. Spencer (2016) suggests that these constitute key ‘tools for change’, ensuring that perpetrator accountability extends to both formal and informal networks in the community.

The research study

An adapted MARAC approach initiated in the North of England aimed to include and centralise the perpetrator in the MARAC process as part of an innovative, whole systems approach to managing risk and safety in high-risk domestic abuse cases. Our evaluation examined the impact and findings of this outcome-focussed approach and our methodology is described below. In this article, we focus on one particularly prominent theme from our findings: the perpetrator as stakeholder. The research team have been engaged in numerous evaluations and research projects over several years and the findings from this study caused us to reflect more broadly on this work. We observed that a common feature in a number of these previous evaluations of multi-agency partnership approaches and other developments designed to tackle the problem of domestic abuse, was the limited focus and paucity of detail on factors specifically related to the perpetrator. This includes identifying specific abusive behaviours, tracking behaviours and activities over time, capturing evidential patterns, and supporting male perpetrators to desist.

Our evaluation examined a streamlined way of managing and coordinating the MARAC, which, in this article, we call ‘the new process’, implemented in one region in England. The aim is to provide holistic support for three parties: victims, children, and perpetrators. In this article, we refer to this as ‘the whole family’. The new process does not rely on a single meeting as with the traditional MARAC process, but rather includes the following four steps: gathering and assessing information, analysing risks and needs for each member of the family, identifying solutions, and finally completing the case. The process is primarily outcome-focussed and crucially, involves co-location of all of the key agencies involved in tackling DA (such as the police, child and adult safeguarding, health, specialist DA services, and probation). The aim of this co-location approach is to improve the speed and effectiveness of information sharing. The focus on victims, perpetrators and children ensures a holistic family approach. As noted, it is the perpetrator as stakeholder we will be focussing on here.

Methods

Our evaluation involved a mixed methods approach, capturing a mix of qualitative and quantitative data as well as a social return on investment (SROI) assessment. The new MARAC process had been implemented in one clearly defined administrative area within the police force boundary for a period of 6 weeks by the time we commenced real time evaluation. The aim of the evaluation was to gather perspectives on the new system, harness continual learning and progress improvement. During the evaluation period, 97 cases were administered.

We carried out qualitative interviews with key stakeholders about the implementation process, and focused observations of MARAC steering and working group meetings over a 2-year period. We also distributed online surveys to core MARAC team members. Ethical approval was granted by the University of Central Lancashire Ethics Committee prior to data collection. Steps to ethical adherence involved informing participants about the purpose of the evaluation, how their anonymised responses would be used, and their right to withdraw without explanation.

Semi-structured interviews took place with 15 stakeholders who were involved in the development of the principles underlying the new MARAC process. Participants spanned a range of areas including policing (varying role and rank; n = 7), third sector (including DA services; n = 4), health (n = 2) and probation (n = 2). Recruitment for these interviews involved an email sent to all stakeholders involved in the development of the MARAC principles, followed by the option of an informal discussion with the researchers for those who sought further information as they considered their involvement in our evaluation. Practitioners who wanted to take part contacted the researchers directly through email.

In addition to the ongoing attendance at working group and steering group meetings and interviews, we administered a detailed anonymous online survey to all of those who were involved with the implementation and roll-out of the new process. The survey included questions garnering opinions on how the implementation had gone in the early weeks, perceived benefits, barriers encountered and recommendations for further improvement of the new process. Respondents to this survey (n = 8) included representatives from police (n = 2), health (n = 1), adult and child safeguarding (n = 3), and probation (n = 2). We also circulated surveys to all adult victim-survivors, and a third to perpetrators, thus all stakeholders experiencing the new process. We received no responses from victim-survivors or perpetrators in the short period of time we had left to complete our evaluation and we reflect on this in our discussion. This article therefore draws on the perspectives of practitioners who completed our online survey and those who took part in interviews.

The interview, survey and observation data were coded and analysed using thematic analysis (Braun and Clarke, 2006) to identify overarching themes in the data. Two researchers performed this analytic stage where themes were independently identified within the data and then compared and discussed to reach a thematic consensus. Themes were then applied throughout the data including the interviews, survey and observation analysis. In what follows, we discuss our findings related to the major theme of perpetrators and their management. We focus upon two aspects of the perpetrator stakeholder theme, namely the challenges of holding perpetrators to account and the provision of services and support for perpetrators.

Challenges of holding perpetrators to account

All participants in both the interviews and survey noted the importance of including perpetrators in the new MARAC multi-agency process. This was also discussed in the stakeholder meetings we observed. One participant remarked that ‘engaging perpetrators and managing their behaviour and abuse is key’ (Interviewee 4, Health Professional). Furthermore, one example of a more positive engagement with a perpetrator was discussed:

I did observe a telephone call with a perpetrator and that amazed me that they were asking the perpetrator what they wanted to change and it turned out that the perpetrator had been sent to services that weren’t helping him for many years, but he wanted a different service. I don’t know what the long term outcome was but at the time it seemed like a lightbulb moment, because everyone had been running around and it had been repeat after repeat but they hadn’t actually asked the person what they want. I thought that was the way we should work and then I just kind of saw what they were doing. (Interviewee 4, Health Professional)

However, despite an awareness of the need to engage perpetrators and this one positive example, participants overwhelmingly reflected on this being a key challenge in effectively adopting a whole systems approach. In particular, participants noted the ways in which the new MARAC process did not adequately engage with perpetrators. The new MARAC process aimed to include victim-survivors, children and perpetrators, yet perpetrators were perceived as the ‘missing stakeholder’ (Interviewee 2, Probation Officer) in the process. This is exemplified in the three quotes below:

Trying to engage perpetrators safely, in a timely manner and in line with the co-ordinated safety plan remains a challenge. We have a significant gap in this area of delivery. We have had some limited success engaging perpetrators within the custody suites, and also if open to Probation Services or Children’s / Adults Social Care. Though if they are outside of these services, it is increasingly difficult to find a way or the right person to make the initial contact and seek to engage within the multi-agency process. We are currently exploring options on how best to address this across our multi-agency partnership. We need to understand the Perpetrators’ needs, views and root causes to be able to effectively offer support, and reduce risk with the aim of preventing repeat referrals. (Survey respondent 8, Police Representative)

Perpetrators are missing in our service delivery at the moment. They are part of the process on paper, but not really in practice. This needs to change if we’re going to get this right (Interview 5, Police)

Working with perpetrators is difficult, but they are the key piece of the puzzle that we haven’t got right yet (Interviewee 1, Police)

These extracts show that participants felt that there were gaps in knowledge and a lack of experience in working with perpetrators, which meant that they faced difficulties engaging this stakeholder group. However, the latter quote in particular highlights that despite the challenges of working with perpetrators, to effectively shift accountability to the perpetrator and remove responsibility away from victim-survivors, engagement and monitoring of perpetrators is key. This is also captured by the following quote:

[The new process] was about forming a team and making sure we were no longer working in silos, and working with not only victim, but also perpetrator and children. It was a challenge, I have worked with victims since 2003. I had been going to MARACS since 2007. So it was a challenge to look at things from a perpetrator perspective. But for years I had been saying if we don’t work with perpetrators, these victims are going to go round and round. And if they get free, that perpetrator without intervention will move on to someone else. But speaking to perpetrators was a challenge. It was hard to speak to him with a different hat on if that makes sense? But it is necessary. (Interviewee 3, IDVA)

A key aspect of whole systems multi-agency working is engaging with all parties involved in domestic abuse (victim-survivors, children, perpetrators and the wider community). The above quote is from an Independent Domestic Violence Advisor (IDVA). IDVAs play a vital role in supporting survivors of domestic abuse, acting as a single point of contact providing emotional support, and guidance around criminal justice procedures (Taylor-Dunn and Enrol, 2021). They may address wider needs such as housing and healthcare. IDVAs also provide support during pre-trial visits, the day of trial, and liaise with the police and other professionals throughout a case. These respondents’ observations highlight how ‘silo-working’ can prevent the opportunity to work with all parties. However, despite agencies being co-located within the new process to minimise ‘silo-working’, the previous quotes from colleagues highlight co-location alone is not enough to ensure perpetrators are a central stakeholder in multi-agency responses. We have headlined a shorter extract from the interview with the same IDVA respondent in the title to this article. Her words most concisely capture the multi-agency stakeholders’ most commonly articulated challenge – ‘to look at things from a perpetrator perspective’.

Participants also reflected on the difficulties of sharing information on perpetrators across agencies. Many of the meetings we observed focused on this issue, with inadequate data recording of perpetrator information being flagged as a particular barrier. For example, ‘We often don’t have full histories of perpetrators, so we need more robust agreements on this’ (meeting observation notes). Another participant similarly reflected,

We often focus a lot on the victim risk, but we don’t always have all of the information we need on perpetrators to effectively share between us. We need to keep working on this (Interview 12, Police)

Information sharing is a persistent barrier to effective multi-agency working within the context of policing domestic abuse. It is regularly cited as a key issue in domestic homicide reviews and serious child abuse case reviews (Olszowy et al., 2020) and we have discussed this issue in relation to this evaluation elsewhere (Davies et al., 2023). Sharing information in a timely manner and ensuring that information is complete, of good quality, and regularly reviewed and updated, is highly relevant to holding perpetrators to account and ensuring their behaviour is managed effectively.

An important challenge discussed by participants is the difficulty with effectively identifying and managing the risks of perpetrators whose behaviour is on a sliding risk continuum. Two stakeholders from third sector organisations were especially attuned to the problem of the dynamics of risk. For example,

[An important issue is that] risk is changing all the time. You could have a case that comes in where a police officer grades it as standard, but it isn’t standard because when you look at the history of the perpetrator, you know they are not standard risk. Cases move up and down all the time. (Interviewee 2, third sector)

Another third sector colleague talks about the complexity of the problem in this respect. Doing safeguarding in a risk paradigm is ‘not easy’:

I think there is an ongoing piece of work to do around how do you take a trauma involved approach to this because naturally you slip into covering your back, checking that you have done everything and doing your basic safeguarding thing and that’s the easy thing to do. I think the skill of this work is how do you bring all of that together so that you have all the complexity at the centre but you still must hold risk at the centre, because this isn’t about being kind to people for kindness sake, you need to be kind to people at the moment you are engaging with them, whoever they are, whatever they have done, because it is kindness and compassion that moves people forward. But doing that within a risk framework – it’s not easy. (Interviewee 5, third sector)

These quotes suggest there is a disjuncture between assessing risk, as required within any MARAC process, but doing so while using a holistic, trauma-informed approach. Risk is a structurally neutral concept that does not account for the ways in which intersectional constraints (such as gender and ethnicity) can affect behaviour (Mythen, 2014). Adopting a risk-focused approach while attempting to be trauma-informed creates a fundamental tension (see inter alia Hannah-Moffatt, 2015). These conundrums have been exposed by others at the global level. Hudson, for example, highlights that the pursuit of safety in the risk society threatens to compromise justice (Hudson, 2003). We see this playing out in risk assessment processes within the context of DA at the local level. For example, it is often unclear what risk is being assessed, that is, is it the current risk, future/predictive risk, or risk of the perpetrator/victim (Barlow and Walklate, 2021; Medina et al, 2016). It is victim-survivors who are asked questions using the frontline police risk assessment tool (often DASH or increasingly DARA in England and Wales), which then goes on to inform the level of support they are provided with. In thinking about perpetrator accountability and risk assessment, the risk indicators relating to the perpetrator need to be more closely weaved in. Knowledge and intelligence about perpetrator activity and behaviours comes from a wide web of sources and a wide net. Capturing and harnessing this in such a way that feeds into the overall assessment of the threat to women is essential for effectively managing and monitoring the risk perpetrators pose.

Provision of services and support for perpetrators

As well as identifying difficulties with engaging with perpetrators as part of the new MARAC process, stakeholders also noted that there was a lack of perpetrator services available to appropriately manage and support them. This is illustrated by the following quotes:

There’s always a level of support for victims and children but support for perpetrators is poor. Social workers are not confident or well equipped. There are few courses for perpetrators to go on and a low take up. (Interviewee 7, Police)

The support locally for victims/survivors and their children is well established in our area. There was a gap in service for a dedicated perpetrator worker to engage into the process where it is safe to do so. This is an opportunity to work creatively with perpetrators and funding needs to be considered to improve the process. (Survey respondent 2)

The paucity of perpetrator services was noted specifically in relation to a lack of DVPP’s. Stakeholders from police and third sector organisations commented on this as illustrated below:

There’s a massive issue with the perpetrator programmes at the minute. The [ANON] centre do a really good job, but they are only a small agency so we are going to have to massively expand perpetrator services across the region. It might well be that our measures, when we come to analysing them, show that there should be a bigger provision for them (Interviewee 8, Police)

There isn’t really any accountability and that’s why we have so many repeat cases. There aren’t many opportunities for perpetrators to go on effective perpetrator programmes, because of a lack of funding, so what can we do to manage them? (Interviewee 5, third sector)

The latter quote highlights two key issues. First, a lack of funding and investment in perpetrator programmes means that there is not enough supply (i.e. availability) to meet the demand. This is not peculiar to the particular area we conducted this study in. It is a wider national issue that warrants geographical mapping, particularly if repeat, harm high perpetrators are being targeted to desist (Robinson and Clancy, 2021). Second, the lack of support provision available for perpetrators means that their needs are not effectively supported and they are often not held responsible for their abusive behaviour. Prosecutions for domestic abuse are consistently low and there is extensive evidence which highlights that criminal justice outcomes are not always favoured by victim-survivors for many and varied reasons. Criminal justice outcomes are therefore not always wanted, viable or available. Perpetrator programmes, which are evidence based and provide individualised, needs-based trauma-informed support, have the potential to support perpetrators to confront their abusive behaviour and place responsibility on them to change their behaviour. This is particularly important as perpetrators often have complex needs which are too frequently not addressed in current interventions (Renehan, 2021), leading to high levels of repeat offending and a lack of appropriate services available to address this.

Opportunities to divert or place change-ready perpetrators swiftly on to suitable programmes ought to be made more uniform.

Discussion

This article has so far explored some of the problems faced in holding perpetrators to account within the context of multi-agency responses to domestic abuse. The challenges stakeholders face include a lack of engagement with perpetrators, issues with information sharing, understanding and managing the risks associated with perpetrators, and a lack of effective perpetrator programmes and support services available to effectively hold them to account. While there are widespread assumptions that a needs-based approach to perpetrator support is necessary, and this aligns with the ‘lived experience’ agenda, this is beginning to sit rather uncomfortably alongside the ‘relentless pursuit’ policy agenda. Relentlessly pursuing perpetrators is becoming a key trope in the effort to reduce domestic abuse as evidenced in recent strategy and funding call headlines. One of three pillars in the VAWG Framework for England and Wales focusses on this very theme (Home Office, 2021). Furthermore, Operation Soteria, a Home Office funded project exploring police responses to rape and sexual offences, also similarly places emphasis on perpetrator responsibility. This focus is evident in other jurisdictions across the globe (such as Australia and some states in the United States). There are also extensive laws and policies in the United Kingdom which aim to support multi-agency responses to perpetrators, including Protections Orders and the Domestic Violence Disclosure Scheme. However, in spite of this, the scale of domestic abuse and the number of women being killed by a current or former partner in the United Kingdom and internationally is significant (ONS, 2021).

Spencer’s (2016) web of accountability is one way of thinking conceptually about how to hold perpetrators accountable for their abusive behaviour. Such conceptualisation fits well with multi-agency approaches that have the key stakeholders centre stage. Whole systems approaches comprise legal (criminal, civil, family law), service (police, adult and child safeguarding, health, third sector) and informal networks and communities. The core stakeholders are the focal point for this wider web of participants. For the partner area, we worked with on this evaluation, although some positive steps were taken to move towards this approach, key tools for change were not being utilised to their full potential.

First, the new approach to the MARAC faced issues with routinely engaging in continuous risk assessment and supporting victim-survivors with effective safety planning. Representatives of all agencies interviewed for this study reflected on the barriers to effecting change. Managing the risk of perpetrators, particularly those who engage in abusive behaviours repeatedly and have complex needs, feature high on stakeholders list of concerns (Robinson and Clancy, 2021). This was compounded by persistent issues with information sharing, particularly due to inconsistent data recording of perpetrator information. There were also limited understandings of the fluctuating nature of risk in domestic abuse cases. For example, many domestic homicide victims may never have been risk assessed at all if they had not come to the attention of the police or other services (Monckton-Smith, 2021). Current multi-agency approaches to policing domestic abuse focus on assessing the risk to the victim-survivor, usually through a frontline risk assessment tool such as the DASH. The emphasis on the safety of the victim-survivor (emphasising this includes safeguarding for her and her children) is only partial. Focussing on the perpetrators risks in parallel may provide an opportunity to more effectively hold perpetrators accountable without compromising risks to women and children. Although perpetrators are risk assessed as they work their way through the criminal justice process, this does not always happen at the initial frontline response. If a case leads to ‘no further action’ (which many domestic abuse cases do), the perpetrator themselves may not face any kind of risk assessment process. Many of the risks that victim-survivors experience are as a direct result of the perpetrators abuse. A pincered and combined emphasis of victim explored safety needs with those of the dynamic behaviours of perpetrators may provide a holistic and comprehensive wrap around approach. This would prompt a better support plan for perpetrators, considering any needs they themselves may have (such as mental health or substance misuse) to help to reduce the potential for further offending.

The second ‘tool for change’ (Spencer, 2016) that was not utilised to its full potential is the availability and use of individualised and needs-led DVPPs, with regular monitoring of engagement. Perpetrator programmes were expensive, limited in availability, and some lacked a robust evidence base in the partner force area for this study. Individualised, gendered and trauma-informed DVPPs offer an opportunity to more effectively hold perpetrators to account, while simultaneously ensuring they get the needs-led support that they need to help to address their abusive behaviour (Downes et al., 2019). Two of our participants reflected on the ways in which perpetrators often leave one relationship and then move on to another and engage in the same (if not more violent) offending, continuing the pattern of abuse. Ensuring that support is in place to effectively try to tackle and hold perpetrators responsible for their abuse is potentially key in breaking this cycle. It is crucial that DVPPs are widely available, adequately resourced, responsive to a diverse perpetrator population, and delivered by appropriately qualified, experienced, and supported staff. Indeed, it is well documented that working with domestically violent men is challenging work – not least because of the tendency to deny, minimise, and blame (Renehan, 2023).

Collectively, stimulating perpetrator accountability and engaging them effectively in the multi-agency policing process is crucial in the balance between safety planning and risk. Perpetrators continue to be the ‘elusive stakeholder’ in multi-agency work to reduce domestic abuse. Many of the tools for change already sit within reach of those within the web of accountability, including legal and service systems and informal networks. All of these could be further exploited and capitalised upon to ensure perpetrators are held responsible for their abusive behaviour.

Conclusion

Balancing risk while addressing the needs of victims, children and perpetrators, is critical for providing holistic support in cases of domestic abuse. Despite advancements in multi-agency ways of working, case numbers remain high, and stakeholder practitioners mention lack of engagement with perpetrators as a major hindrance. Although there are limitations to the study we report on (low-sample sizes, lack of victim or offender participants) which may affect on the findings reported, we have found that attempts to improve the MARAC system can work for victims and children, yet challenges to engaging perpetrators remain. These challenges include issues with information sharing, understanding and assessing changing risk levels, and a lack of effective perpetrator programmes and support services available to effectively hold them to account and to support them with their needs (e.g. trauma, mental health, addiction). We recommend future work to address this missing link to provide support that takes into account the diversity and complex needs of perpetrators (Hilton et al., 2019).