UNDER THE RADAR: THE WIDESPREAD USE OF ‘OUT OF COURT RESOLUTIONS’ IN POLICING DOMESTIC VIOLENCE AND ABUSE IN THE UNITED KINGDOM

The suitability of ‘out of court resolutions’ (restorative justice and community resolutions) in cases of domestic abuse is theoretically contentious and empirically under-researched. This study investigated the nature and extent of out of court resolutions for domestic abuse using the Freedom of Information Act. Out of court resolutions were used by every UK police force except Scotland to respond to over 5,000 domestic abuse incidents (including intimate partner abuse) in 2014. Some of these incidents related to offences with sentencing tariffs up to life imprisonment. Such widespread use has been taking place ‘under the radar’ in stark contrast to police guidance, has immediate implications for policy and practice, and fundamentally shifts the research terrain in this field.


Introduction
The last 20 years has seen a major overhaul in the way domestic violence and abuse is policed in the United Kingdom (hereafter 'domestic abuse' in line with police terminology).Feminist criminologists first highlighted how it was not treated with the seriousness accorded to other types of crime in the late 1980s and early 1990s (e.g.Edwards 1989; Dobash and Dobash 1992).Advances have included the opening of specialist Domestic Violence Units (although most have now closed or been subsumed into 'vulnerabilities'), the introduction of specialist police officers, training for police officers, the introduction of Independent Domestic Violence Advocates (IDVAs), Domestic Homicide Reviews and Clare's Law (the right to ask and the right to disclose the name of domestic violence perpetrators to potential victims).However, there exists an extensive body of research documenting a range of ongoing problems.For many victims, the responses of the police and the criminal justice system more broadly remain deeply problematic.These concerns have been evidenced in academic research (e.g.Hester 2006;Walklate 2008;Bond and Jeffries 2014;Westmarland 2015), as well as by a succession of critical reports from Her Majesty's Inspectorate of Constabulary (HMIC).The 2014 HMIC report, e.g., found significant weaknesses and called for urgent action to be taken (HMIC 2014).The following inspection 12 months on, however, found that although a positive action approach exists on paper, this had still not been translated into effective, equitable practice (HMIC 2015: 17).These problems, plus the ongoing and unresolved debate raised previously in this journal on the role of the police in cases of domestic abuse where victims do not wish for the case to progress through the criminal justice system (Hoyle and Sanders 2000), have led some to consider alternative justice mechanisms.
One alternative justice mechanism available in the United Kingdom is the 'out of court resolution', which we define as consisting of a range of police-led restorative justice practices, approaches or techniques (these terms are used interchangeably) as well as community resolutions (an out of court disposal which has overlaps with street-level restorative justice techniques where appropriate).These differ from other out of court penalties such as cautions and fixed penalties in that they seek a 'resolution' to an individual or community level dispute or conflict.In contrast to other crime types, there exists little knowledge on restorative approaches in relation to policing domestic abuse.Likewise, research on community resolutions is lacking in terms of domestic abuse and is scarce for other crime types also.Despite this empirical vacuum, we were aware through our research partnerships with some police forces that out of court resolutions were being used in cases of domestic abuse.Accordingly, we undertook this study to gain systematically collected, comprehensive, baseline data on the nature and extent of UK police use of out of court resolutions in cases of domestic abuse.What we found was that every police force in England, Wales and Northern Ireland (but not Scotland) used out of court resolutions to respond to domestic abuse in 2014.This is far more widespread than previously imagined: it fundamentally shifts the research terrain in this field, and has immediate implications for policy and practice.

Police Use of 'Out of Court Resolutions' in Domestic Abuse
'Report to court' style studies are complicated but important because they show how attrition throughout the criminal justice system operates and where cases 'drop out' of the system.There are few 'report to court' style studies for domestic abuse.While offences such as rape can be more easily counted and tracked through the criminal justice system, tracking domestic abuse is more complicated because it relies on the police applying a domestic abuse 'flag' to the incident.For crimed domestic abuse incidents, there exists a broad range of offences that domestic abuse can be recorded as-some of the most common being criminal damage, common assault, actual bodily harm, harassment, threats to kill and theft (Hester and Westmarland 2006).Looking from the 'report' stage right through to the 'court' stage allows a more holistic view than police, Crown Prosecution Service or court data on its own gives, and reveals that perpetrators reported to the police for domestic violence are very unlikely to be convicted and punished.Hester's (2006) study remains the most comprehensive, which found that only 31 suspects were convicted out of 869 domestic violence incidents recorded by Northumbria police (4 per cent of all incidents or 14 per cent of incidents where an arrest was made).Only one in two hundred incidents resulted in the suspect being convicted and given a custodial sentence (0.5 per cent or 4 out of 869 incidents).
More up-to-date data are available in terms of outcomes at the police stage.HMIC data show variability between forces in terms of the ratio of crimed incidents that they charge, caution or to which they apply another out of court disposal (HMIC 2014).This has led HMIC to state that they are 'gravely concerned' both at the apparent inconsistencies between forces but also about the appropriateness of high rates of cautioning, taking no further action, and out of court disposals including restorative justice.They found that in some cases officers were more inclined to caution because it counts as a detection and is therefore recorded as a 'success' in terms of force performance management.They did not disaggregate the out of court disposals but did note that, in general, 'forces are following the long established national policy position, that police use of restorative justice interventions in domestic abuse cases concerning intimate partners is inappropriate, ineffective and potentially dangerous' (HMIC 2014: 101).They did report that some forces were using restorative justice interventions routinely, but did not quantify this or give any further detail in terms of crime types, form of restorative justice used etc.Community resolutions were not mentioned at all within the inspection report.
It is important to unpick what the terms 'restorative justice' and 'community resolution' mean in terms of policing-widely different conceptions of 'restorative justice', in particular, has been previously identified as problematic (Miers 2004).The Association of Chief Police Officers (ACPO, now the National Police Chief's Council-NPCC) outline three different 'levels' of restorative justice that officers may use when responding to a reported incident: Level 1: an instant or on-street disposal, where police officers or PCSOs [Police Community Support Officers] use restorative skills … to resolve conflict in minor crimes and incidents … an alternative to a formal criminal justice process.Level 2: measures such as restorative justice conferences, and may involve more participants, risk assessments and seek longer-term solutions.A Level two restorative justice response can occur either as an alternative to criminal justice proceedings, or in addition to criminal justice proceedings, as part of a formal crime disposal.Level 3: resolutions that take place in addition to criminal justice proceedings, mainly post-sentence … for cases that involve serious, complex or sensitive incidents, or where offenders are being monitored by an offender management team and/or are deemed at risk of continued offending.(ACPO 2012) ACPO define a community resolution as the nationally recognized term for the resolution of a less serious offence where an offender has been identified through agreement between the parties, as opposed to via the criminal justice system (ACPO 2012: 1.1.1).In line with the recent emphasis on positive action and discretion in policing generally, community resolutions were introduced to provide a means by which officers could respond to lower level crimes proportionately-where crimes and disputes could be 'resolved' via officer intervention, without resulting in criminal prosecutions.The guidelines suggest that community resolutions are generally to be used for first time offenders, where the victim does not wish to pursue any more formal action and, while they can be used together with restorative approaches, they do not have to meet all the restorative justice preconditions.Importantly for domestic abuse, this means they would not need to meet the precondition of victim agreement.

The Policy Context
As already noted, there has been a dearth of academic attention paid to how the police use out of court resolutions in domestic abuse cases.This is likely to be because they are thought to be seldom used: firstly, because of concerns and a general lack of support from the specialist domestic violence sector; secondly because community resolutions (though not necessarily restorative justice) are designed to resolve less serious offences; and thirdly because a review of policy documents shows that guidance offered to the police gives a strong steer that restorative justice particularly but also community resolutions should not generally be used in cases of domestic abuse.
This review of policy documents shows that the Independent Police Complaints Commission (IPCC), the ACPO, HMIC and the previous Home Secretary (now Prime Minister) Theresa May have all expressed reticence regarding the use of restorative justice and/or community resolutions.In 2011, ACPO stated that it did not support the use of restorative justice in cases of domestic abuse.However, they simultaneously recognized that restorative justice was a 'customer focused methodology' disposal and so could be considered where sought by the victim (ACPO 2011: para 6).Guidance issued the following year on community resolutions (with or without restorative justice) specifically excluded domestic abuse from its remit (ACPO 2012: 1.1.5).In 2013, in its national recommendations following a domestic homicide, the IPCC stated categorically that 'restorative justice should not be used in cases of domestic abuse or domestic assault' (IPCC 2013: 4).During 2014, the HMIC published its report on policing domestic abuse which expressed concern over the use of restorative justice which they stated gives rise to 'unacceptable risk' (HMIC 2014: 15).It specifically stated that the police 'should not use restorative justice in intimate partner domestic abuse cases and should do so with extreme caution in other forms of domestic abuse' (HMIC 2014: 15).By 2015, the College of Policing published revised guidance stating that restorative justice and community resolutions are 'rarely appropriate in domestic abuse cases and not recommended in cases involving intimate partner abuse' (College of Policing 2015: 4.2).Finally, the guidance warns officers that a request for restorative justice from a victim may be indicative of being under the 'influence of controlling or coercive behaviour and making the request to please or appease the perpetrator' (College of Policing 2015: 4.2).In her 2016 speech at the Police Federation's annual conference, Theresa May criticized the police for failing to accord domestic abuse with the urgency it required and specifically rejected the use of restorative justice for intimate partner abuse (Travis 2016).Domestic abuse policing policy over the last five years at least, therefore, has been to avoid the use of out of court resolutions-both in terms of restorative justice and community resolutions-particularly for intimate partner abuse.

Investigating Police Use of Out of Court Resolutions: Research Methods
The aim of this research was to investigate the nature and extent of police use of out of court resolutions in responding to domestic abuse in the United Kingdom.It was not our intention to study their effectiveness but rather to provide a comprehensive empirical base from which further research could be developed.The Freedom of Information Act 2000 (England, Wales and Northern Ireland) and the Freedom of Information (Scotland) Act 2002 were used to issue a series of questions to all police forces.Ethical approval was obtained from the Research Ethics Committee at the School of Applied Social Sciences, Durham University.
Two phases of Freedom of Information (FOI) requests were used: the first to all forces asking about restorative justice practices, and the second asking about community resolutions.The requests asked about: the total number of recorded domestic abuse incidents in 2014; of these, how many incidents involved the use of restorative justice (in the first request) or community resolution (in the second request); the level of restorative justice used (in line with the ACPO definition); and whether the recorded relationship between the offender and the victim was intimate partner or familial.For those cases that involved intimate partner domestic abuse we asked for the additional incident level data: offence type; relationship between victim and perpetrator; the gender of the victim and perpetrator; whether the incident resulted in a prosecution (if a restorative approach); the level of the restorative approach employed; and finally a brief description of the type of restorative approach/community resolution adopted.We therefore generated a data set containing both quantitative and qualitative data.
It is important to note that the terms community resolution and restorative justice were used interchangeably in the responses from 14 forces.One (incorrectly in our view) stated that: 'Since 2013 outcomes which would have been known as restorative justice are known as community resolutions'.Another force replied: 'We see RJ and CR as the same thing' and another responded with data in respect of 'Restorative Justice, i.e.Community Resolution'.In total, we identified 12 different ways of recording these out of court responses including the obvious 'restorative justice' and 'community resolution', but also categories such as 'outcomes equivalent to restorative justice', 'restorative justice (community resolution)' and 'community resolution (with restorative justice)'.Overall, we found no difference between the descriptions of the action taken and whether it was marked as restorative justice, community resolution or one of the other terms identified above.This conflation across force recording systems (and arguably also in terms of each force's use and understanding of the terms) is why we have coined the broader term 'out of court resolution'.It will be important for future researchers to note this merging of approaches, processes and terminology when undertaking research in this field.
In terms of responses, all UK forces replied to at least one of the FOI requests sent and supplied us with some or all of the data requested.A database was created and frequencies were calculated for the quantitative data and thematic analysis was used on the qualitative data (incident or outcome descriptions).The main strength of this method was that it was a relatively low cost way of gaining an original and substantial UK-wide data set.Nonetheless, it must be emphasized that securing a comprehensive data set was time-intensive in terms of frequent liaising with FOI officers to secure accurate and comparable data.
The main limitations were inconsistencies across forces in the way the police record data and the systems they use, making it difficult to draw direct comparisons.This was also found by Bows and Westmarland (2016) in their study of rape of older people using the FOI method.Bows and Westmarland (2016) also point to the need to clearly define terms because of the reliance placed on FOI officers' understandings of the request.This was particularly pertinent in the current study because domestic abuse spans a wide range of offence types (given 'domestic abuse' is not a criminal offence).Several forces said that the use of the domestic abuse 'flag' was not consistently applied, some were not able to include non-crimed incidents even when restorative approaches were used in such cases, some did not consistently record the relationship of the victim and perpetrator and so were unable to separate partner/expartner from familial violence, and there were some individual inconsistencies which are dealt with via missing data exclusions in some of the analysis presented in the next section.As described above, there were also inconsistencies in how the terms 'restorative justice' and 'community resolutions' were used and recorded, and the overlaps were to such a great extent that it was not possible to separate out the practices.We therefore concur with Bows and Westmarland (2016) that FOI requests are a useful research method for criminology research but additionally highlight the complexity that comes with investigating types of crime such as domestic abuse which encompass a very broad range of offence types.

Research Findings
In 2014, out of court resolutions were used by all police forces in England, Wales and Northern Ireland to respond to domestic abuse Every police force in England, Wales and Northern Ireland (but not Scotland)-so 45 out of 46 UK police forces-reported using out of court resolutions-either restorative justice, community resolutions and/or a regional equivalent (in the case of Northern Ireland) to respond to domestic abuse in 2014.This is extremely widespread coverage for police responses that are not nationally recommended in cases of domestic abuse and about which we have little, if any, supporting evidence.The only force that did not report using any form of restorative justice, community resolution or equivalent was Police Scotland.

The proportion of domestic abuse cases dealt with using out of court resolutions varied significantly across the forces
The total number of domestic abuse cases (crimed and non-crimed incidents) dealt with using out of court resolutions across the 45 forces was 5,466.This is the minimum number of cases and should be considered an undercount because of data quality issues (such as not always applying the domestic abuse 'flag' and whether the data referred to incidents or offences, as forces supplied us with different data).
When we look at the 24 forces that provided us with data relating to domestic abuse incidents (everything recorded by the police with a domestic abuse flag, regardless of whether it constituted a criminal offence or not), the proportion of cases dealt with using restorative justice or community resolution ranged from 0.1 per cent to 2.4 per cent of all domestic abuse incidents.The three forces with the lowest use were Northumbria (0.1 per cent, n = 23), North Wales (0.1 per cent, n = 11) and Thames Valley (0.1 per cent, n = 52).The forces with the highest proportions were Derbyshire (1.7 per cent, n = 306), Lancashire (2.4 per cent, n = 430) and West Midlands (2.4 per cent, n = 751).
When we look at the 31 forces that provided us with domestic abuse offences (crimed incidents-only those incidents which were criminal offences), the proportion of cases dealt with using restorative justice or community resolution ranged from 0.3 per cent to 4.5 per cent.The three forces with the lowest use were North Wales (0.3 per cent, n = 11), Dyfed-Powys (0.4 per cent, n = 7) and Avon and Somerset (0.4 per cent, n = 98).The forces with the highest proportions were Northamptonshire (3.9 per cent, n = 188), West Midlands (4.4 per cent, n = 751) and Greater Manchester (4.5 per cent, n = 837).
As previously mentioned, we do have some data quality issues.However, these issues are associated with an undercount rather than an overcount of the number and proportion of cases.The fact that over 5,000 domestic abuse cases in 2014 used out of court resolutions, and that as many as one in forty incidents and one in twenty offences (crimed incidents) were responded to in these ways, means that their deployment is far more widespread than has previously been thought.This means that rather than primarily engaging in theoretical debates about whether out of court disposals should be used, there exists a current practice that researchers can empirically investigate.

Out of court resolutions were not only used in familial domestic abuse cases but also for intimate partner domestic abuse
The Government definition of domestic abuse used by the police and most other organizations in England and Wales includes both intimate partner and familial forms of violence and abuse (Home Office 2013).We were particularly interested in the use of out of court resolutions in intimate partner domestic abuse cases due it being a more controversial and under-researched area.Recording the relationship between victims and offenders in domestic abuse cases only became mandatory for forces in April 2015 (shortly after our data collection time period ended): accordingly, as above, the data we received were variable and we had a large amount of missing data.
In total, 39 forces recorded the relationship between victims and offenders for at least some of the cases.This translated into 1,555 cases that involved intimate partners (either current or ex-partners).Due to the missing data, we are unable to present this as a proportion of the overall sample of cases because we do not know whether the missing data refer to intimate partner or familial cases.The key finding here is that out of court resolutions were not restricted to familial forms of domestic abuse but were also being used in intimate partner abuse cases by all of the 39 forces that recorded this information.

Restorative justice was used almost exclusively as a diversion from, rather than parallel to, a criminal prosecution
While community resolutions are a disposal in themselves, restorative justice responses are not and can be used alongside other criminal justice disposals including prosecution.We were given criminal justice prosecution data from 34 forces relating to 1,227 intimate partner abuse cases.Of these, only one single case was recorded as involving any form of further criminal prosecution.In this case, where criminal damage to a dwelling had occurred, the offender was given a police caution in addition to the restorative justice outcome which was recorded as agreeing to pay reparation for the damage caused.
The almost total absence of parallel criminal justice proceedings suggests that Level 1 and Level 2 restorative justice approaches were being used almost exclusively as a form of diversion from prosecution (Level 3 is by definition post-prosecution).This would be understandable for community resolutions (given these are a disposal in themselves) but is not in line with the potential of restorative justice which can be a parallel rather than diversionary action.

Out of court resolutions were used for a broad range of serious criminal offences which, if prosecuted, have potential sentencing tariffs of up to life imprisonment
We were given individual level data for 1,098 of the crimed incidents which involved intimate partner abuse and an out of court resolution.As shown in Graph 1, the most frequent offence categories were violence against the person (56 per cent of cases), followed by arson and criminal damage (36 per cent of cases).Theft accounted for one in twenty of the cases (5 per cent) and together public order offences, vehicle offences, burglary and miscellaneous crimes against society (coded together as 'other' in the graph) accounted for 3 per cent of the cases.
Tables 1 and 2 show the offence frequencies within the two most common offence categories: violence against the person (Table 1) and arson and criminal damage (Table 2).The violence against the person category includes 101 cases of actual bodily harm and 131 cases of common assault and battery.The arson and criminal damage category mainly consists of generic criminal damage (338) but does include one case of arson endangering life.
It is generally inappropriate to talk of 'more' and 'less' serious offences when it comes to domestic abuse because the context around the offending is so important.An offence that seems 'low level' on paper can in fact be part of a pattern of coercive, controlling, threatening behaviour which can ultimately lead to homicide (Regan et al. 2007).Therefore, all domestic abuse offences should be assumed to be serious because of how they can link into a pattern of violent and abusive behaviour.It is still worth commenting, however, on the level of sentence that some of these offences would potentially attract if successfully prosecuted.Offences in Tables 1 and 2 that are particularly serious in terms of sentencing tariffs include malicious wounding (up to five years), threats to kill (up to ten years) and arson endangering life (with a maximum of life imprisonment).In essence, very serious crimes are being dealt with by means of out of court resolutions.
Graph 1 Frequencies of offence categories recorded for intimate partner abuse conferencing or youth mediation that took place between current or previous intimate partners.There were no examples of adult conferences (although our data set is incomplete in terms of descriptions).The remaining Level 2 descriptions (n = 31) consisted of outcomes such as apology, reparation and 'words of advice'-all of which mirrored Level 1 response descriptions, and which do not appear to comply with ACPO's Level 2 restorative justice requirements.In responding to the FOI request, one force did say that they had a further 60 Level 2 restorative justice cases which related to youth cautions or conditional cautions which had 'an element of RJ'.
Particularly concerning were descriptions of intimate partner abuse cases coded as involving out of court resolutions but where the action seemed negligible and unconnected to either 'restoration', 'community' or 'justice'.In some cases, the recorded action amounted to no positive intervention-despite being coded as a restorative justice or community resolution.Examples of this included: 'parties separated'; 'no formal complaint made'; 'now separated'; and 'victim wanted no further contact'.These responses would seem to have far more in common with a 'no further action' disposal.In two cases, the action taken was described as being the offender's continuing engagement with their social and support workers.Some examples of actions had distinct overlaps with criticisms of pre-1990s policing of domestic abuse, e.g.: 'words of advice given' and 'note signed to effect that offender will behave'.Others reinforced the idea that domestic abuse is simply a form of relationship difficulty, e.g.: being advised to 'seek marriage guidance'; 'relationship intervention work '; and 'mediation via PC'.These examples raise the question of whether some out of court resolutions are being applied inappropriately, to 'square away' domestic abuse crimes by recording them as 'detected' and 'cleared up' when arguably no discernible positive action has taken place.In doing so, forces could be artificially improving their crime detection performance rates.By increasing the volume of cases that are 'detected', there is a Youth Triage Programme, relationship intervention work, conferencing through Youth Offending Team, referred to intervention team via youth services.
corresponding decrease in the proportion of domestic abuse offences that result in 'no further action' and remain 'undetected'.
Alongside the sheer variety of responses that were coded as some form of out of court resolution there existed major inconsistencies-even within the same force area for the same offence type.For example, within one force, 13 different crimes of assault and battery were responded to with eight different forms of 'restorative justice' in intimate partner abuse cases: (1) 'going to CBT [Cognitive Behavioural therapy]'; (2) 'dealing with childcare case formally'; (3) 'seeking CAB [Citizens Advice Bureau] advice and marriage guidance'; (4) 'working with social services'; (5) 'letter of apology'; (6) 'verbal apology'; (7) 'mediating with PC'; and (8) 'wife removed to specialist unit'.

There was an over-representation of same-sex cases and of cases involving male victims and female offenders in terms of how frequently out of court resolutions were used
The recorded gender of the victim or offender was provided by 34 forces, relating to 1,139 cases of intimate partner domestic abuse.These are shown below in Table 4.We were further able to identify the gender composition of victim and offender dyads in 1,084 cases, which are presented in Table 5.
These findings follow the general overall pattern we usually see in intimate partner domestic abuse: a higher proportion of female than male victims, a higher proportion of male than female offenders, manifesting most frequently in the context of male-onfemale domestic abuse in a heterosexual intimate relationship.However, the gender proportions of victim and offender categories within the data are less pronounced when compared to national statistics (Walby et al. 2015).Furthermore, the ratios between victim and offender gender dyads could reflect an over-representation of same-sex partnerships.Overall, same-sex partners comprised 7 per cent of the gender dyad cases.While there are currently no nationwide statistics on police occurrences involving domestic abuse and same-sex partners (Seelau et al. 2003), we believe this figure to be high.Therefore, the results suggest that out of court resolutions are being used disproportionately in cases involving female offender and male victims, and also in cases involving same-sex intimate partners.We are not able to assess within this study why  this is the case and this should be the focus of further research.Nonetheless, there are a variety of possible explanations.In terms of the higher prevalence of female offenders and male victims, these include: the police not taking female violence as 'seriously' as male violence; the police being more punitive towards male offenders; or that the incident in question occurred as part of an ongoing series of domestic abuse incidents where the female was otherwise the victim (Gerber 1991;Feather 1996).In the case of same-sex partners, data may reflect same-sex domestic abuse being treated differently to, or not being taken as seriously as, male-on-female intimate partner violence (e.g.Seelau et al. 2003;Pattavina et al. 2007).

Discussion-Under the Radar
This research has, for the first time, uncovered the widespread police use of out of court resolutions in responding to domestic abuse across the United Kingdom-responses that have until now been happening 'under the radar'.It reveals that out of court resolutions were being used in 2014 by all police forces in England, Wales and Northern Ireland to respond to a considerable number of domestic abuse cases involving both family members and current or former intimate partners.There is no reason to think that this has decreased at the time of writing ( 2016).These responses were used for a variety of criminal offences, some of which would attract significant sentences on conviction including life imprisonment.Data additionally revealed a potential over-representation of out of court resolutions in cases that involved women offenders and same-sex partners.Moreover, the approaches were predominantly informal 'on-the-street' restorative outcomes (ACPO 'Level 1') which served almost exclusively as an alternative to prosecution.Our research contributes new knowledge in two fields of criminology-firstly that of policing domestic abuse and secondly that of police use of restorative justice and/or other out of court resolutions.
In terms of policing domestic abuse, we argue that street-level (Level 1) restorative justice and community resolutions represent a step back in time in terms of policing domestic abuse and are not safe or appropriate in the context of intimate partner domestic abuse.Terms such as 'words of advice given' or 'verbal apology made' chime with pre-and early 1990s approaches, seeing domestic abuse as a private and non-serious matter that police officers can deal with without the need to progress through the formal criminal justice system.Police discretion was a problem in early studies of domestic abuse and continues to be a concern that is documented in more recent research (Myhill and Johnson 2016).That officer-led mediation is named as an outcome for several cases is salient, given concerns that domestic abuse has in the past been problematically framed in terms of the 'problem couple' (or more contemporarily the 'troubled family'), thereby minimizing the seriousness of domestic abuse (Westmarland 2015).
The ways in which different forces record domestic abuse incidents, offences, victims and perpetrators, and their actions still varies between forces and has the (intended or unintended) consequence of making it difficult to compare responses.We fear that using community resolutions as a disposal for domestic abuse may be providing some officers with the means to inappropriately inflate domestic abuse detection (clear up) rates (this is not the case for restorative justice as it is not a disposal).This would fit with a pattern of problematic responses documented in other research whereby some police officers continue to minimize the seriousness of domestic abuse, reducing its visibility and the rate it is prosecuted by 'cuffing' cases (Myhill and Johnson 2016).More 'report to court' studies are needed to see how this research fits into wider criminal justice system outcomes.As mentioned earlier, these are few in number because of their complexity, and cross-force comparisons increase this complexity.Researchers planning cross-force domestic abuse studies should be aware of these continued difficulties and inconsistencies.In addition, researchers should be attentive to the changing profile of disposals and interventions including those not traditionally associated with policing domestic abuse-restorative justice and community resolutions in this study-but potentially others.
This research has also added to the body of knowledge on the police use of restorative justice and other out of court resolutions.Here, we concur with Marder (2016), who has argued that the discretion inherent in Level 1 responses has enabled police officers to stretch the concept of 'restorative justice' to such an extent that it does not adhere particularly closely to most principles of restorative justice.He found that many such responses included limited dialogue between stakeholders, echoing Miers' ( 2004) concerns over a decade earlier.This suggests that our findings in relation to restorative justice and domestic abuse responses are not unique to domestic abuse but common to the way the police are using Level 1 restorative justice responses more generally.However, while a low level 'dispute resolution' style approach with a nod to restorative justice may be appropriate in some cases, the nature of domestic abuse (a pattern of often 'low level' incidents characterized by coercion and control) makes it inappropriate and unsafe.The fact that these are 'on-the-street' disposals negates the possibility of a full risk assessment being conducted including discussion with multi-agency partners and independent support for the victim to make a full and informed decision about whether to engage in such a process (as per the ACPO guidelines 2011).This represents a step back in time in terms of highly individualized responses by officers who may not hold the necessary knowledge or skills to ensure a positive outcome.While apologies and reparations are listed as appropriate examples of community resolutions in responding 'proportionately' to 'lower-level' crime (ACPO 2012), in the context of domestic abuse, the informality of the restorative outcomes raises significant concerns.
It is possible that there may be a place for out of court resolutions for cases of familial domestic abuse or for planned, formal Level 2 restorative conferences with support for victims and following an appropriate programme of research and evaluation.Criminologists have argued that there should be space made to discuss the possibilities of restorative justice in cases of domestic abuse.Perhaps most strongly, Strang and Braithwaite (2002) have argued that the evidence is sufficient to 'impose an obligation on criminologists to be open to the possibility that restorative justice has something to offer in the domain of family violence that courts do not have to offer' (p.4).We argue that this research still needs to be undertaken in relation to police use of restorative justice but agree that the door should be left open to this possibility given the long-standing limitations of the criminal justice system in responding to domestic abuse (see, e.g., Hester 2006;Westmarland 2015).Truly restorative approaches-developing practices in ways that are attuned to the specificities of domestic abuse-as well as harnessing the coercive and expressive powers of state criminal justice systems, should be about expanding the opportunities for women to engage in justice processes (see also McGlynn et al. 2017).Innovative approaches could be developed that meet survivors' needs and justice interests, address and understand the offending, offer safety and protection, as well as enabling survivors to regain their voice and autonomy (Pennell and Burford 2002).
At the moment, the restorative practices being employed by the police in responding to domestic abuse do not appear to be in line with this potential.We question whether restorative justice practitioners would recognize the police practices labelled as such, and propose that the way many forces are using the term 'restorative justice' is so porous as to be unhelpful.Rather than being associated with interventions involving prior risk assessment, planning, support structures and managed outcomes; the term restorative justice is currently being used as an umbrella term for a multitude of street-level practices.
As well as academic contributions, the research has highlighted important issues for policy and practice.This research has uncovered a wide chasm between policy and practice.The finding that out of court resolutions are being used by every force in England, Wales and Northern Ireland and routinely for some forces is, we argue, in contravention of domestic abuse policing guidance as it is not just being used in 'exceptional' cases.However, and confusingly, it does not contravene policy on victims and restorative justice-with the Victims Code guaranteeing that restorative justice should be considered for any victim.This inconsistency between two areas of policy should be clarified given the broad use we have found in practice.

Conclusion
Our findings challenge current scholarly and policy discourses on out of court resolutions-particularly those involving restorative justice-and domestic abuse in the United Kingdom which tend to focus on the feasibility of introducing restorative justice: who should do what, when and how.We argue that the use of out of court resolutions has been happening 'under the radar'-since we have demonstrated that both restorative justice and community resolutions are already widespread.Research and debate, therefore, needs fundamentally to shift from considering theoretical feasibility to interrogating actual current practice.Given the limitations of the criminal justice systems in responding to domestic abuse in the United Kingdom and elsewhere, it is of Policing policy on the use of Level 1 restorative justice and of community resolutions should be strengthened to reinforce the guidance against their use in cases intimate partner abuse, and familial abuse where there exists coercive control.At the same time, it is vital that current practices are not simply pushed (further) underground.We need a greater openness and transparency in the policing of domestic abuse so that there is less pressure to increase (or maintain) the number of 'detected' cases; a demand which might be driving the use of out of court resolutions.The risks attached to poor police responses to domestic abuse are now well versed and a return to individual, inconsistent police interventions without appropriate risk assessments and multi-agency involvement represents a large backwards step.

Funding
No funding was obtained for this research.
importance internationally to consider what positive actions can be taken inside and outside the criminal justice system.Such a programme of research could include research questions such as: what do police see as the incentives to using out of court resolutions for domestic abuse; what are the views of victims on out of court resolutions; to what extent would victims engage with Level 2 restorative justice case conferences; what are the safeguards needed to safely engage in restorative approaches in domestic abuse; what are the differences needed in approach for intimate partner versus familial domestic abuse; what is the extent of Level 3 (post conviction) restorative justice in domestic abuse; and what other forms of violence and abuse are being dealt with by way of out of court resolutions.

Table 3
Nature of out of court resolution (n = 414 types relating to 346 cases)Seek guidance behaviour from support team, visit GP [General Practitioner] for medication, seek help for addiction, seek CBT therapy, seek CAB advice, seek marriage guidance, seek counselling, attend anger management, engage with social services, attend pathfinder diversion scheme, seek substance abuse support, seek education, training and employment support.

Table 4
Gender of victim and offender (n = 1,139)

Table 5
Gender of victim and offender in intimate partnership cases(n = 1,084)