Tackling domestic abuse locally: paradigms, ideologies and the political tensions of multi-agency working

The British government’s strategy to tackle violence against women and girls cements an approach seeking to prevent and protect. Within this context, local initiatives to tackle domestic abuse have proliferated. This article draws on an evaluation of an innovative multi-agency tasking and coordination (MATAC) approach to tackling serial perpetrators. Though the evaluation showed positive outcomes, tensions surfaced within this holistic strategy. In reflecting on the shifting economic and political context in which local agenda setting and commissioning is occurring, perceived concerns about victim safety are reported. Where initiatives have a heightened focus on perpetrators, and in the effort to responsibilise, there are tensions around safeguarding and risk. These are discussed with reference to divergent political cultures and translations of the problem of tackling domestic abuse.


Introduction
This article draws on an evaluation of an innovative partnership approach -known as multi-agency tasking and coordination (MATAC) -to tackle serial domestic abuse perpetrators in an English police area. Though the evaluation showed positive outcomes, tensions surfaced within this holistic strategy designed to prevent violence and protect from it. Perceived concerns about victim safety, alongside the heightened focus on perpetrators, prompts this critical reflection on the shifting economic and political context in which local agenda setting and commissioning is occurring. A brief overview of how violence against women and girls (VAWG) is being tackled at a global through to local level leads into an account of the developments in the multiagency approach to tackling such abuse. This contextual backdrop serves to frame the MATAC process and the theory of change that seeks to tackle serial domestic abuse perpetrators at the same time as working to protect victims and prevent future victimisation. It also outlines the dominant ideological underpinnings of domestic abuse policy. The MATAC is then described. The remainder of the article reports on findings and discusses tensions around safeguarding and risk in the effort to responsibilise perpetrators. The long-standing tensions inherent in multi-agency partnership working to tackle VAWG surfaced in the context of the MATAC which has a heightened focus on perpetrators. These tensions serve as healthy reminders of the divergent paradigms, ideologies, politics and working cultures at stake in multiagency partnerships.

The global to local context of multi-agency working to tackle VAWG
Although focusing on a specific local approach, the issues raised in this paper are in no sense parochial. There are wider resonances. VAWG, recognised as a global issue, has been on the agenda of the United Nations for over 20 years and in 2016, member states of the World Health Organisation adopted a plan of action to tackle it. Broader international obligations derive from human rights protections enforceable through the European Court of Human Rights. Other international provisions include the 2011 Council of Europe Convention on preventing and combating violence against women and domestic violence (the 'Istanbul Convention'). Article 16 relates to treatment programmes for perpetrators and the Convention requires signatories to provide legislative or other measures to support prevention together with specialist support for victims.
A fundamental recent change affecting the 'policing' of domestic abuse across many countries is the widespread recognition that domestic abuse is an issue of power and control and that it should be understood as a pattern of behaviours that can be physical, emotional, economic and sexual in nature. In the UK it is now recognised that the dynamics of domestic abuse are connected to the concept of coercive control, which captures both the psychological and physical aspects and on-going nature of the behaviour and the extent to which the actions of the perpetrator control the victim through isolation, intimidation, degradation and micro-regulation of everyday life (HMIC, 2015). The potential to enhance women's access to justice through changes to legislation that reflect the coercive nature of such abuse is the subject of much current debate (see Fitz-Gibbon et al, 2018). Police and their multi-agency partners are simultaneously grappling with the implications of such legislation in practical terms.

Multi-agency working and domestic abuse
Partnership approaches were identified early in the new era of community safety as a way of tackling domestic abuse (Barton and Velero-Silva, 2012). Since the mid-to late 1980s, there has been increasing reliance on such partnerships to prevent abuse and protect from it. This tradition is well established in England and Wales. Prior to this, single agency responses were typical and there was very little information sharing, particularly between statutory and voluntary agencies. Operating largely within the confines of a traditional criminal justice paradigm which seeks to hold perpetrators to account through legal sanctions and mandated rehabilitation solutions, domestic violence forums proliferated in the 1990s, inspired by the 'Duluth approach' in Minnesota, USA. During this decade government leadership on domestic violence saw national action plans emerge. By the turn of the twenty-first century, prompted by a combination of Home Office guidance and legislative requirements to form partnerships to tackle crime and disorder, information sharing in England and Wales became more routinised (Westmarland, 2012).
There has been significant economic and political change in the period since multiagency working became the dominant approach to tackling domestic abuse, which have had an impact on partnership working in many areas of social policy and have affected local agenda setting and commissioning. At the same time, significant victimfocused policy reform has occurred. Policies applied by the 2010-2015 Coalition government, and continued under the Conservative administration, have changed the way victim support is managed. Featherstone and colleagues (2012, 177) call this a period of 'austerity localism' which, in the context of domestic abuse, where it interfaces with women's safety, is problematic (Vacchelli, 2015). Collaboration between statutory agencies and local women's networks have been compromised and local feminist inspired women's groups perceive that they have lost out, resulting in the further marginalisation of domestic abuse victims (Buser, 2013;Clayton et al, 2016;Vacchelli, 2015;Westwood, 2011). From 2014 to 2015, provision of services for many victims have rested with Police and Crime Commissioners, who are also responsible for establishing local policing priorities. The current national strategy 2016-2020 (Home Office, 2016), promotes a coordinated response, within which regional and local initiatives have proliferated. It also coincides with the new offence of 'controlling or coercive behaviour in an intimate or family relationship' under section 76 of the Serious Crime Act 2015. Walklate and colleagues (2017) note that the implementation of this offence in England and Wales has so far been patchy, and, based on their gendered analysis of coercive control, they conclude that more law will not improve responses to intimate partner violence. Others are also wary of the wholesale adoption of coercive control as an approach that seeks to explain variations in domestic violence (Walby and Towers, 2018). In Brisbane, Australia, Douglas (2017) reports that legal engagement can be an opportunity to extend an intimate partner's coercive control. Coining the phrase 'legal systems abuse', Douglas cites survivors' comments as evidence of how the legal system continues to be harnessed by perpetrators as a tool to extend coercive control beyond separation.
Legislative changes to the definition of domestic abuse in England and Wales, followed two decades of policy reform directed towards an integrated strategy to tackle VAWG (HMIC, 2014;. There are many criminal and civil intervention options in England and Wales, some of which see prospective victims provided with information about their partner's previous violent behaviour, advocating preventive ideologies. For example, legally enforceable short-term protective measures include Protection Notices (DVPNs) and Protection Orders (DVPOs), introduced via the Crime and Security Act (2010). DVPOs resemble the 'barring orders' operating elsewhere in Europe: the Austrian Protection against Domestic Violence Act 1996 and the German Protection from Violence Act 2002 (Bessant, 2015). Additionally, the Domestic Violence Disclosure Scheme 2014 (DVDS or Clare's Law) provides a framework for members of the public to ask about a person's history of domestic abuse or intimate partner violence. The latter are part of a recent shift occurring internationally, of campaigns targeting primary prevention at men (Cismaru and Lavack, 2011). Despite such commitments too many women are victims, with an estimated 1.2 million in England and Wales experiencing such abuse in the year ending March 2017 (ONS, 2018a) and, on average, two women are killed each week by a current or former partner (ONS, 2018b). Despite considerable reforms to provide a more effective response to victims of domestic abuse, the challenge to effect change remains in the lap of local stakeholders.

Ideological underpinnings
The rise of the women's refuge movement in the 1970s brought the issue of domestic abuse to the attention of policymakers. In the last three decades, feminist inspired responses to the provisioning of support has ensured that the complex social structural and interpersonal dynamics of domestic abuse are more widely understood. Recently, there has been a marked shift towards the targeting of perpetrators. Holding perpetrators to account and efforts to reduce re-offending are now part of a comprehensive strategy (Devaney, 2014;Griffiths, 2013[[not in references, is 2015]]; Donovan and Hester, 2014;Featherstone and Fraser, 2012). This orthodoxy suggests that a coordinated and holistic response is most likely to be effective (Dobash et al, 2000;Gondolf, 2002;Kelly and Westmarland, 2015;Rajagopalan et al, 2008;Rivett, 2010). However, there are nuanced ideological differences and political sensitivities regarding what should be done and how it should be done.
Several feminist scholars have long maintained that explanations of violence against women should centre on gendered social arrangements and power (Dobash et al, 1992;Yllo, 1993) and the 'gendered nature' of such abuse is clear (Myhill, 2017). However, different feminist perspectives have subtly distinct allegiances to the concepts of male domination, sexual inequality, gender hierarchies, dominance and power arrangements. The view that violence against women is one of many outgrowths of patriarchal systems remains a strong legacy from early radical feminist theorising, when scholars first promoted the idea that patriarchy could explain male violence against women (Brownmiller, 1975;Caputi, 1989;Russell, 1975). However, as Hunnicutt (2009) has observed, there are hotly contested varieties of patriarchy and fierce critiques of the dominant feminist framework within which policy and practice has developed. The continuing effort to tackle violence against women and domestic abuse thus produces tensions. Dixon and colleagues (2012) for example, writing in the context of domestic violence perpetrator programmes (DVPPs), refute the patriarchal ideologies underpinning much current practice and evaluation of partner violence. A variety of views exist too about working with abusive men and fathers Griffiths, 2013[[not in references, is 2015]]; Featherstone, 2014;Featherstone and Peckover, 2007;Featherstone and Fraser, 2012). As noted above, the wide-ranging approach inherent in the new formulation of domestic abuse as coercive control is controversial and this too is perhaps only heightening the tensions about what should be done and how. Robinson and colleagues have reported on the extent to which the work of practitioners is informed by a sound understanding of coercive control and find that the absence of a clear understanding when making judgements about victims and perpetrators has serious implications for the efficacy of current approaches to domestic abuse (Robinson et al, 2017). As will be shown, these tensions come to the fore in local multi-agency partnerships that are tackling domestic abuse.

Multi-agency tasking and coordination (MATAC)
Developed in a police force area in Northern England and driven by the regional VAWG strategy, the MATAC process to tackle serial perpetrators of domestic abuse commenced in November 2015. The objectives of this project, funded by Home Office innovation funding, were to improve victim's safety; criminal justice system outcomes; offender behaviour and partnership engagement. The approach is designed to: prevent domestic abuse occurring in the first place/limit its re-occurrence; ensure victims receive prompt and comprehensive wraparound support where violence occurs; and ensure that those who perpetrate domestic abuse are held to account. The project in its entirety comprised (i) a tool to identify the recency, frequency and gravity (RFG) of offending, (ii) the MATAC process, (iii) a domestic abuse toolkit, (iv) a domestic abuse proximity notification system (GPS tagging units), (v) force wide availability of voluntary DVPPs and (vi) work with housing providers to focus on domestic abuse perpetrators. The MATAC is now part of core police business.
The RFG scored each perpetrator from 0 to 100. Scoring was based on weighted criteria, taking into account serial offending and multiple victims. The list this generates was drawn upon to select perpetrators for discussion at the monthly MATAC meeting. Agency representatives (including local government, third sector, health, criminal justice and housing) were also able to refer perpetrators into the process as part of their information sharing. The RFG aided in identifying individuals whose profile resembles that of a known domestic abuse perpetrator, yet these men are not monitored for their violent inter-personal behaviour (Donovan and Hester, 2014) or categorised as domestic abuse perpetrators (Featherstone and Peckover, 2007). The model builds on other work that highlights the importance of empirically-validated risk assessment tools for identifying high-risk perpetrators (Juodis et al, 2014a;2014b).
At MATAC meetings actions were determined to manage perpetrators (see Figure  1: MATAC Workflow Chart). The toolkit was designed to facilitate prevention, diversion, disruption and enforcement according to the assessment of an engaging/ non-engaging perpetrator. A pathway for 'engaging perpetrators' ('green options') was for those who recognise that their behaviour is problematic and want to change. A pathway for 'non-engaging perpetrators' ('red options') was for those who are unwilling to recognise their behaviour is abusive. Perpetrators could be managed in ways that cut across both pathways. Perpetrators were served a warning letter by the police or most appropriate MATAC stakeholder. The process was risk assessed to ensure all safeguarding precautions were attended to via a harm reduction plan to negate potential escalation of abuse. The neighbourhood police team were responsible for the development, management and review of this plan. The letter explained that the perpetrator is in the MATAC, why they are in it and what this means. It gave information about services and support in the area in which the perpetrator resided and it gave warnings of actions that may be taken if behaviour remains unchanged. The serving of the letter was always accompanied by robust wraparound support for potential victims.
An important intended outcome of MATAC was the shift towards prevention and early intervention via a focus on perpetrators not previously known to police as presenting a significant risk to women. The combined use of the RFG tool and the MATAC principle of sharing information saw previously un-reported offending exposed. Serial perpetrators were to be processed through MATAC, dealt with appropriately while women and children would be protected. Additional DVPP capacity and the opportunity for voluntary participation in them enabled men to take responsibility -as men -and/or fathers.

Methodology
A two-year evaluation of the MATAC ended in March 2017. It utilised a mixedmethods approach, comprising four elements: analysis of perpetrator monitoring, case studies, an online survey and semi-structured interviews with MATAC stakeholders, victims and perpetrators. The research was approved by a university ethics committee. All participants (partner agency respondents surveyed, victims/survivors, perpetrators and police analysts) were provided with written information explaining the purpose of the research. Those interviewed also had a verbal explanation. Participants' consent was obtained based on this knowledge and understanding of the research and of how we envisaged using the information. Victims/survivors and perpetrators could choose to have someone sit with them during the interview. All were free to refuse to answer questions and participants could pull out at any time, without giving a reason and without negative consequences. Discussions were audio-recorded. Where respondents preferred to talk off the record this was facilitated and information gleaned in this way has not been used in the analysis. Individual case studies were compiled following MATAC actions and were analysed alongside the corresponding RFG statistical data. Victims/survivors and perpetrators selected for interview were not matched to individuals undergoing MATAC actions and RFG score monitoring.
Quantitative data, provided by the police allowed us to analyse perpetrator related behaviour while subject to the MATAC. Data analysis included descriptive analysis of RFG scores, offending history, information provided by MATAC on interventions undertaken and outcomes. Perpetrator case studies complemented this statistical data allowing us to illustrate how offending histories and wider issues have an impact on abusive behaviours, MATAC interventions and outcomes, as well as the complex issues faced when dealing with perpetrators of domestic abuse. An online partner agency survey explored understanding of the project's aims, time and resource implications, perceptions of actions and decisions made at MATAC meetings, and the management of harm/risk/safety. The survey was administered using Bristol Online Survey software. It was sent to 182 representatives of partner agencies, via an email explaining the purpose of the survey (N=26). The survey was administered again to 141 individuals at the end of the second year (n=24) to identify any change. Not all were then actively participating in the MATAC. The response rate of those actively involved is likely to have been higher than the apparent 17 per cent.
The fourth component was semi-structured interviews with perpetrators, victims/ survivors and MATAC stakeholders. The latter, which we report on here, involved semi-structured interviews (N=18) with representatives of organisations managing, delivering and participating in the project. The interviews explored understandings of the project, its perceived 'fit' with partner agencies' policies and practices and wider approaches to victim safeguarding, membership and operation of the project including impacts and emerging issues. Findings discussed here draw mainly on qualitative semistructured interviews with the 18 stakeholders and responses to the online surveys. All interviews were fully transcribed and analysed thematically.

Findings
The MATAC has a preventative and forward-looking strand to it, reflecting global, national and local ambitions to protect future and current victims. The police lead commented on how the MATAC approach comprised these complementary strands and how it sought to achieve the overall ambition to reduce domestic abuse: the process is essential. For too long we've focused on picking up the pieces after perpetrators. Having a targeted early intervention focus on perpetrators as well as protecting victims sends a clear message that their behaviour won't be tolerated and that by working together we can disrupt it.
The evaluation reported positive outcomes from the non-statutory MATAC process in a challenging environment (Davies and Biddle, 2018) at the same time as it exposed contentious issues. Outcomes signifying success include reductions in overall offending of perpetrators after MATAC intervention as well as reductions in domestic abuserelated offending. Measures resulted in the re-housing of serial perpetrators, voluntary enrolments on DVPPs and greater use of Criminal Behaviour Orders and Warning Notices. Perpetrators' alcohol misuse was addressed, and, in various combinations, these interventions saw reduced RFG scores. However, anxieties emerged in relation to this holistic strategy. Victim safeguarding, safety and risk and the idea of 'responsibilising' serial perpetrators were where anxieties were evident. These are discussed below using data from the online surveys and anonymised verbatim extracts from interviews. The discussion reflects on this within the context of the MATAC and wider VAWG strategy.

Victim safeguarding
The online partner survey revealed increased levels of satisfaction (from 2016 to 2017) among stakeholders about the actions arising out of MATAC meetings and with the MATAC management of risk and safety (see Table 1).
Police are enthusiastic about the value of DVPNs and DVPOs. One informant told us that such tools 'offer a huge amount of protection for the victim'. Police are also clear about how wraparound support is provided, when and by whom: The ideal time to serve the letter is when he's been excluded through the DVPN so he's not at the address. At that point the victim should be getting support from the IDVAs to make choices around the future, counselling, advice and support.
However, a vocal minority of local authority stakeholders were concerned that the MATAC was not sufficiently prioritising victims and safety.

The warning letter
While police describe the letter as part of a strategy that: sends a clear message -and are clear about when and how to serve the perpetrator letter, a domestic abuse coordinator twice used the example of the serving of the letter to illustrate their concerns about it being used as a trigger for notification of inclusion in the MATAC: I do have an issue about how that's approached, with police going knocking on doors and asking if they fancied a perpetrator programme.
Furthermore, the first time we administered the survey (2016), one respondent provided the following free text comment: No evidence of consideration for the immediate risk posed to the victim at the time the perpetrator is approached.
Another respondent wrote: Police should be the agency to approach the perpetrator not other agencies -other agencies can support the police, as they may know the person.
This may suggest that partners are not yet ready to assume a role requiring them to be in closer proximity to the perpetrator. This resonates with findings from an evaluation of voluntary perpetrator programmes (VPPs) conducted by Donovan and Griffiths (2015). They report that work with perpetrators was outside the remit of some partners and, where it was part of their remit, they operated through a criminal justice lens which prioritised detection, prosecution and mandated sentencing. They also report that female practitioners felt unsafe about engaging with perpetrators, especially when in the domestic setting. Furthermore, they suggest that motivating men to engage in programmes designed to change behaviour, and getting them 'treatment ready', is worthwhile, 'but it is not clear in the case of VPPs who would do this work ' (2015, 1159). The serving of the MATAC warning letter might be considered a 'pre-commencement' strategy for treatment readiness. It remains unclear who should do this type of work and which agency should have this role. A minority of MATAC partners struggled to see working with perpetrators as part of their agenda. These valid concerns are worrying and it is important that they were addressed in the early part of the roll out of the MATAC. No such strong views were repeated in the 2017 survey, perhaps indicating that their views had changed as a result of the MATAC process. For police, the letter is intended to convey a clear message to perpetrators: that their behaviour won't be tolerated. The stakeholder views noted above were from two partners who had never attended a MATAC meeting, yet they were outspoken about their concerns for victims' safety, suggesting: you need to be able to have a very frank and open discussion about case management…There's a tension here; on many levels it's really not safe.
Such case management discussions are precisely the business of the MATAC and these concerns raise two important issues concerning tackling domestic abuse locally. First, they serve as a constant reminder about victim safety at the point of the intervention. Second, they remind us of the continued importance of multi-agency communications outside and beyond the MATAC partners to wider stakeholders.

RFG scores, safety and risk
Police and survey data suggest that victims' safety is improved. There was an overall force-wide decline in the rate of domestic abuse offences committed by perpetrators subject to MATAC interventions. The reduction between pre-admission and point of discharge was -39 per cent. Furthermore, the average RFG scores of perpetrators subject to MATAC interventions declined by 1.07 from their point of engagement in the MATAC to their point of discharge, and by 4.56 from point of engagement to March 2017, though there were varying levels of decline across each locality. The percentage of respondents to the survey who were satisfied with the way that the MATAC managed harm/risk/safety increased from a mediocre 54.1 per cent to almost 88 per cent. This coincided with a sharp increase in agreement that information was shared effectively between MATAC partners (from 63 per cent to 96 per cent).
We found no evidence from the survey or interview data that victims of targeted perpetrators were unprotected, rather, victims reported feeling well protected. Nevertheless, a minority of multi-agency stakeholders perceived that victim safety may be compromised. Knowing whether these are valid remains a challenge for any multi-agency partnership and evaluation team.

Responsibilising serial perpetrators of domestic abuse
The idea of holding perpetrators to account is not new. Successive UK governments and several jurisdictions across the globe have long subscribed to the position that men should be accountable for their domestically abusive behaviour through a strong criminal justice response. Increasingly this is juxtaposed with the new preventive consciousness, where infrastructures are oriented towards prevention, security and fear-reduction (Garland, 2001). The latter priorities differ from the traditional criminal justice goals which have continued to predominate in the policing of domestic abuse. Here, accountability has long been synonymous with legal sanctions and being held to account by the state, rather than men taking responsibility for their behaviour (Devaney, 2014). MATAC has the capacity to facilitate state and personal accountability and allows perpetrators to be managed in ways that include an appropriate mix such that engagement with them is as domestically violent men and, sometimes, as fathers.
The majority of stakeholders were very positive about these aspects of the MATAC. One probation partner reported that the real benefits of engaging with the MATAC are that it established links with other professionals who are otherwise more difficult to contact. The example given concerns partners from health: anything with multi-agency is really good. Just because it gets different people on the table and different points across. Each agency has different priorities…Even if MATAC isn't the right choice, they can then be referred… for example to MAPPA [Multi-agency public protection arrangements].
Overall, we found widespread optimism about the approach. A housing provider stated: MATAC have enabled services to place the onus of change/responsibility on the perpetrator rather than place all the pressure on the already vulnerable victims.

A police participant said:
The MATAC process is an important part of the work we do with regard to working with repeat domestic violence perpetrators, work which is aimed at reducing their risk of reoffending and harm. The MATAC process provides us with an arena to share the work we do and also take into account the views and information from other local agencies. MATAC without doubt assists in enhancing the lines of communication between those agencies involved in the work to protect victims of domestic violence and abuse within our community.
A perpetrator programme coordinator declared: We have found the MATAC process to be invaluable. Ensuring clear and safe pathways into behaviour change programmes which have positive outcomes in relation to the safety of women and children.
Nevertheless, some stakeholders introduced caution and these voices acted as checks and balances as the MATAC processed perpetrators. One partner asked: At which point do we decide what route we're going down, who makes that decision? There are tensions and ethics and dilemmas, how are we gonna do this?
One coordinator captures the tensions inherent in the MATAC approach when reflecting on the setting up of the partnership: It was ensuring compliance, interrupting someone's behaviour; I noticed increasingly there's a tension here. I've noticed the language used is the softer stuff around support. I think trying to do both. Potentially there's a lot for support agencies to do.

The same respondent continued:
It's so hard to go down a criminal justice process as well as doing this as well…Unless we've got mental health very much on board some of the 'green options' are going to have a gap.
There were specific concerns about MATAC's perpetrator focus: The MATAC is perpetrator focused, the MARAC is victim-focused and I think it would probably have been better to find some kind of way to put those both together. Slight concern that if we are not considering the needs of the victim, the risks to the victim and their children and wider family members then we could end up with a situation where you might put remedies in place to manage the risk of the perpetrator in isolation of his family or families and that could potentially escalate the risk to the victim. If we are being perpetrator focused then are we really going to have that opportunity to consider the victim and if we are considering the victim and their needs as well then why aren't we doing that through the MARAC?
As stressed, the traditional criminal justice paradigm holds perpetrators to account through mandated legal sanctions. MATAC retains these mechanisms but introduces an additional form of accountability for those ready to address their offending behaviour. The toolkit thus included 'therapeutic' or 'adjunct' interventions (Bessant, 2015). This avenue facilitated men holding themselves to account. The 'engaging' perpetrator might tackle their mental health or substance misuse issues, take up housing support and voluntarily embark on a domestic abuse perpetrator programme. This approach is rooted in a rather different belief about the potential for personal change than the traditional approach to holding perpetrators to account. As such, it opens up the space to work with motivation rather than compulsion (Devaney, 2014).
The traditional criminal justice approach of imposing restrictions and mandated sentencing is relatively uncontroversial when offenders have been convicted of offences. However, prior to conviction, when screening by practitioners and child protection assessments results in considerable numbers of referrals or notifications from the police either not being acted upon or receiving a perfunctory response, all resulting in no services, women are left to manage violent men's behaviour and its consequences. The MATAC offers a more rigorous screening and tailored support for women during this very risky period.

Discussion
Three points are worth returning to as part of an extended discussion: first, safety planning, second, the relationship between RFG scores, DVPPS and responsibilising perpetrators and third, the synergy between the MATAC approach and the wider VAWG strategy. The tensions discussed above illustrate the need for continuous information sharing and communication to ensure that all stakeholders are aware of the safeguarding protections surrounding victims. The importance of information sharing and communication frequently arises in multi-agency working (Davies and Biddle, 2018). The above also underscores the importance of safety planning for victims, and this too partly concerns improved communication among stakeholders. Juodis and colleagues (2014b) have highlighted this same issue when discussing what can be done about high-risk perpetrators of domestic violence.

Safety planning and risk
The question of whether or not concerns about victims' safety and risk are well founded is a very pertinent one. The imbalance of power in abusive relationships is the crux of the problem. The coercively controlling behaviours of perpetrators severely restrict the opportunities for women to make choices that are most likely to lead to a cessation of violence and abuse (Clarke and Wydall, 2013). Where stakeholders empathise very strongly with this position, yet remain wedded to the familiar and traditional criminal justice paradigm, they will have little faith in the short-term remedies of DVPOs and DVPNs as disruptive and preventive measures. In isolation, these measures do not empower victims to take decisions to protect themselves. With no reassurance that the temporary removal of the perpetrator from the family home is likely to become more permanent, victims will be dubious about their own and their children's safety. Given victims are in greatest danger at the point of separation, when the perpetrator senses a loss of control (Fleury et al, 2000), wraparound support at this point is crucial. Separating from violent men is no guarantee of safety and the promise of long-term changes in perpetrating behaviours through mandated or voluntary perpetrator programmes compounds the problem for some sceptics. Lack of access to other forms of support has become more difficult by austerity cuts that have removed many local services. Just as women who are coercively controlled need to feel fully supported and protected, stakeholders too need convincing, or they will perceive there to be escalated risks to victims' safety.
Seemingly dissenting views about the capacity of MATAC to prevent and protect are healthy reminders of how highly volatile, threatening and risky domestic situations can be and how women, who are separated from their violent partners, are at risk of post separation fatal violence. Cautionary notes are emerging from others about increasing the risk of fatal violence. As Duggan (2018) has noted, the DVDS is predicated on preventing future victimisation and this 'right to ask' (RtA) route to disclosure about an individual's history of abuse now supplements the existing 'right to know' (RtK) route. Under the latter, those working in the statutory sector can initiate disclosures on a safeguarding or public protection basis as a result of the risk to a person's safety. These high-risk victims are, in theory, well protected, whereas those engaging with the scheme via the 'right to ask' (RtA) route may find their risk level elevates as a result of receiving information. RtA applicants may require support from the point of contact through to aftercare, following the outcome of a disclosure, to ensure that risk levels do not increase, given a person's involvement in the scheme (Duggan, 2018;Fitz-Gibbon and Walklate, 2016). Responsibility for such support and care is likely to fall to organisations that already have anxieties about victims' safety, resources having been stretched under the form of localism practised since 2010 (Davies and Biddle, 2018).

RFG scores, DVPPs and responsibilising perpetrators
Though reduced RFG scores are heartening, a reduction in reported offending cannot necessarily be treated as a reduction in actual offending. Reasons for the under-reporting of domestic abuse are well known and these reductions may be otherwise explained. In part, reductions may reflect victims feeling responsible for their partner's wellbeing. Observed changes could also result from men exchanging physical violence for more subtle and coercive forms of abuse, worsening the situation for the victims/survivors and their children. An apparent reduction in offending might therefore be due to increased victim intimidation by an abuser who blames their partner for the increased surveillance and/or interventions which they are now experiencing. It could also be due to increased self-censorship by a victim who feels responsible for 'bringing trouble' to their partner's door, or for bringing their family to the attention of the police.
Voluntary DVPPs exist for men who have not yet been convicted of domestic abuse offences. Such programmes sit comfortably within the recent shift to responsibilise domestically violent men -as men -and as fathers, and offer possibilities for men to change and develop non-violent parenting and partnering relationship patterns (Featherstone, 2014;Featherstone and Peckover, 2007). They may in part explain reduced RFG scores. There is, however, controversy about engaging with male perpetrators, whether categorised as domestic abusers or not. The policy and practice paradigm which posits power, control and patriarchy as explanatory factors in the context of domestic abuse, has a strong hold. A rather different ideology about behavioural change, one that may sit less comfortably in the activist paradigm, recognises other factors as significant, including psychological vulnerabilities, situational factors and relationship dynamics (Devaney, 2014).
Research into the pros and cons of mandated versus voluntary programmes, and the merits or otherwise of stand-alone programmes, reveals mixed results. Identifying the 'success' of perpetrator programmes is challenging given the complex dynamics of domestic abuse and relationships between victims, perpetrators, their families and the wider community (Clarke and Wydall, 2013;Kelly and Westmarland, 2015). Evidence suggests that their impacts can be modest but more effective when tailored to the individual needs of the perpetrator (Babcock et al, 2004), and that victims find breathing space when partners engage in programmes (Kelly and Westmarland, 2015). Critics suggest programmes fail to tackle under-reporting, are insufficiently funded to deliver effectively, excuse and fail to criminalise abusive behaviours and have high drop-out rates (Devaney, 2014;Featherstone and Fraser, 2012;Hester and Lilley, 2017). Some applaud those programmes that address wider issues having an impact on perpetrator behaviour and tackle substance misuse, mental health problems, parenting and emotional issues among perpetrators (Juodis et al, 2014a;2014b;Rivett, 2010).
Foregrounding the needs of male perpetrators is a particularly controversial issue (Clarke and Wydall, 2013;Hester and Westmarland 2006). Some are concerned that perpetrator programmes incorrectly raise the expectations of victims/survivors and their children about the perpetrator's behaviour change, putting them at increased risk (Donovan and Griffiths, 2015). These authors also raise concerns that statutory provision for male perpetrators may be funded at the expense of services providing for the safety of women. While Clarke and Wydall (2013, 402) report unanimous support for the multi-agency 'Making Safe' scheme, they found that 'occasionally there was friction between practitioners who worked with perpetrators and those who worked with victims'. They observed tensions when service provision for victims and perpetrators is compared. 'Careful management was required, not only when allocating resources, but also in ensuring that a "them" and "us" mentality did not develop amongst practitioners and undermine the overall ethos in its joint aims of empowering victims and tackling perpetrators'. With austerity measures designed to reduce public expenditure continuing, similar tensions are reported here; indeed, it seems that these tensions might only have increased as the competition for scarce resources digs deeper, with ideological perspectives on violence against women becoming more deeply divided. Given the level of commitment practitioners' display towards their particular client groups, working together closely does not always make for a comfortable alliance (Clarke and Wydall, 2013). In the MATAC, additional DVPP capacity for men to engage voluntarily came from a police innovation fund. This funding would never have otherwise gone to services supporting victims. Partners in the MATAC did not all appreciate this. Against a harsh backdrop of cutbacks, competition for resources is intense.

MATAC and VAWG
Reflecting on the VAWG strategy, governance at both national and transnational levels is shifting, such that incident-focused, reactive criminal justice approaches no longer hold sway. Holding perpetrators to account is no longer exclusively managed through criminal justice sanctions. This paradigm has failed to reduce levels of recidivism or improve the safety of women and children (Devaney, 2014, 480). Furthermore, due to under-reporting, much domestic abuse never comes to the attention of agencies. Many victims remain without protection, and perpetrators remain hidden and outwith the criminal justice system. This failure of the criminal justice paradigm to tackle domestic abuse is compounded by ignoring those serial perpetrators whose domestically abusive behaviour is below the radar and by failing to protect multiple victims from them. As discussed, the RFG tool aids the identification of these serial perpetrators. The MATAC approach provides a potentially coordinated and integrated response to domestic abuse by focusing on both victims and perpetrators. It identifies and addresses the needs of victims and perpetrators and embraces the key aims of protection and prevention and, as such, it sits comfortably within the national VAWG strategy and also the Istanbul Convention. It shows how multiple stakeholders, with various strengths, expertise and ideological values, are needed to prevent and manage the danger posed by serial domestic abuse perpetrators, and ensure the safety of women and children in our local communities.
The roll-out and operationalisation of MATAC raised tensions common in multiagency partnerships. The current embodiment of localism has also created difficulties, due to its association with austerity. The safety of victims is paramount and the central aim of a domestic violence reduction strategy. The prevent strand of such a strategy is operationalised alongside the support and service provision or protect strand. The former and the latter are designed to do two things simultaneously -tackle perpetrators and support victims. This 'holistic' strategy can present challenges for multi-agency approaches to domestic abuse and this was evident in the MATAC partnership.

Conclusion
The policing of domestic abuse and strategies designed to tackle and prevent continue to rely on multi-agency partnerships in local communities. Longstanding tensions are inherent in such partnerships working to tackle VAWG. The MATAC is an innovative way to tackle serial domestic abuse perpetrators at the same time as working to protect victims. Tensions are evident within MATAC, notably from members whose organisational and/or personal priority is ideologically and historically more clearly wedded to prioritising energy and funding towards only one aspect of this overall strategy.
The discussion highlights the crucial importance of information sharing and communication at the local level. The tensions discussed illustrate the underlying politics of community safety and the heightened and tense politicised climate surrounding it. Context therefore remains crucial including local political, economic and cultural histories (Edwards, 2013;Hughes and Edwards, 2002;Wilson and Ross, 2015). The surfacing of scepticism and undercurrents of concern are rooted in real anxieties that can make partnership working difficult. Tensions are likely to become more widespread in a climate of continued austerity and economic uncertainty (Featherstone and Fraser, 2012), and in times that threaten services, especially if they are in competition for evermore limited funding. Indeed, there are limits to 'how open a partnership can be to the complex milieu of interests in a locality without compromising its capacity to act' (Hughes and Edwards, 2007, 27). However, in relation to MATAC at least, it is a 'red herring' to suggest funding for programmes for violent men diverts attention away from services to women and children. In this context it is unhelpful to couch the problem in such zero-sum terms where a focus on victims is pitched against and versus a focus on perpetrators.
A minority of concerned stakeholders retain a strong and healthy scepticism about MATAC and, given that the MATAC process is now part of core police business it is important that the MATAC meetings are a multi-agency workplace for professional reflection, information sharing and genuine collaboration. A focus on perpetrators has traditionally not been 'everyone's business', and, as partners grasp the importance of recognising mutual interests, tensions between key partner agencies is likely to subside. The majority of stakeholders see the capacity of the MATAC to prevent and protect, such that women and children are safer, and men are held to account by preemptively addressing their complex needs, such that they are 'change ready'. Effective multi-agency information sharing and hard partnership work includes addressing undercurrents of concern, as well as of overt conflict. The concept of patriarchy is useful in that it keeps the gaze directed toward social contexts rather than toward individual men motivated to dominate (Hunnicutt, 2009). It may be that with regard to domestic abuse, there are also other significant factors to be taken into account if effective solutions are to be found via local partnerships, even in times of austerity.