Abstract
Over the last decade, coercive control has entered the legal lexicon, having developed from a social construct (Wiener, 2020) into transformative offences in all of the jurisdictions of the United Kingdom. 1 In this article, we refer to coercive control as conduct, whether physical or non-physical, employed by an abuser to undermine the victim’s autonomy, freedom and integrity and entrapping them within the relationship (Stark, 2007). Psychological abuse, physical or sexual violence, or threats of physical or sexual violence are typical tactics of coercive control. Section 76 Serious Crime Act 2015 (SCA) in England and Wales and s. 1 Domestic Abuse (Scotland) Act 2018 (DASA) adopt different offence models, with each recognising that domestic abuse can entail a pattern of ongoing abusive behaviour. Both offences do capture physical and non-physical behaviour, however, differ in that the Scottish offence was designed to do so, while s. 76 was intended to address the non-physical only (Home Office, 2014). In fact, the English offence does capture some low-level physical abuse which is tangential to it (McGorrery and McMahon, 2019). As such, these offences are unlike traditional criminal offences that are generally incident based and focused on physical harms and injury. By capturing more accurately the reality of how domestic abuse is often perpetrated and experienced, the offences are potentially transformative. They allow for a more holistic understanding of the dynamics of abuse and its various manifestations including how non-violent forms of behaviour can support or supplant the use of physical force by the perpetrator. These offences have therefore been welcomed by many practitioners and agencies working in the field of domestic abuse (Scott, 2020). However, they cannot, and will not, be truly transformative unless they are successfully implemented by criminal justice agencies.
While the respective offences within the two jurisdictions have previously been examined through a comparative law lens (Bettinson, 2020), in this article, we draw on original empirical research to examine how the new offences are being implemented in practice. The research examines whether, from the perspective of the criminal justice professionals tasked with implementing the law and third sector agencies working with them, the offences have been successful, and if not, what they felt were the most significant barriers to success. By exploring the factors that have been limiting or assisting in operationalising coercive control offences in the two jurisdictions, lessons can be learnt from one jurisdiction to inform the other and potential issues with coercive control prosecutions more generally can be highlighted.
Methods
Our study aimed to explore the perspectives of criminal justice professionals and voluntary agencies on the implementation of coercive control offences, in Scotland and in England and Wales. For context, the statistics show that the uptake of prosecutions for coercive control in both jurisdictions has been slow. In England and Wales, Official National Statistics illustrate that police recording and subsequent prosecutions of coercive and controlling behaviour have been increasing year on year. The current figures are 41,626 offences of coercive control recorded by police in England and Wales in the year ending March 2022, compared with 33,954 in 2021 and 24,856 in 2020 (Office for National Statistics (ONS), 2022). In the year ending March 2020, this resulted in 1,208 prosecutions (Home Office, 2022) and 1,403 in the year ending March 2021 (ONS, 2021). To the extent that this reflects increasing utilisation of the offence, it suggests a degree of successful operationalisation. However, the number of police-recorded offences is not particularly high relative to the wider profile of domestic abuse; and two recent studies have found only marginal police usage of the offence (Barlow et al., 2020; Myhill et al., 2023a). Indeed, in its review of the operation of section 76, the Home Office noted that the 24,856 coercive control offences recorded by police in England and Wales in 2019/20 required to be situated in the context of a total of 758,941 domestic abuse–related offences that year. Meanwhile, in Scotland, it is somewhat more difficult to evaluate the success or otherwise of the implementation of s. 1 DASA given the dovetailing in time between its enactment and the COVID-19 pandemic (Scottish Government, 2023b: 2). What is known, however, is that a larger proportion of total domestic abuse offences are charged in Scotland – 92% in 2020–2021 (Scottish Government, 2023a) – compared with England and Wales – almost 72.7% in the year ending 2022 (ONS, 2022). In 2021/22, there were 33,425 charges reported by police to the Crown Office and Procurator Fiscal Service with a domestic abuse identifier, reflecting a 9% increase on the previous year. The vast majority of these offences remained, however, for Breach of the Peace, Common Assault or Crimes against Public Justice (Scottish Government, 2021). Indeed, statistics for both 2021–22 and 2022–23 in Scotland indicate that DASA charges accounted for under 6% of all domestic abuse charges reported (COPFS, 2023) – though this constitutes a greater proportion than in England and Wales, it is thus still relatively rare.
This study was designed to explore whether, despite their differences in legislative design, evidential procedure and organisational structures, there were aspects that could be learnt across jurisdictions in terms of the framing of offences, investigative practices, training and trial processes. We were interested in particular in criminal justice professionals’ opinions about what challenges and opportunities they had experienced since the offences were enacted and what they considered needed to be done, if anything, to ensure their effective operationalisation. To explore these issues, during 2021–22, we undertook a series of semi-structured interviews lasting approximately 60 minutes and conducted on MS Teams. An interview schedule was designed and modified slightly to accommodate each constituency, with the same broad themes addressed across discussions. The interviews were recorded, transcribed and coded for subsequent qualitative analysis using Nvivo. Interviews were conducted in accordance with ethical approval granted by the University of Warwick Humanities and Social Sciences Research Ethics Committee. In total, across the study we spoke with 20 participants in Scotland – comprising of 5 police officers, 6 Crown Office prosecutors, 6 Sheriffs, and 3 domestic abuse sector specialists from voluntary organizations – and 25 participants in England and Wales (predominantly the former) – comprising of 12 police officers, 7 Magistrates, and 6 domestic abuse specialists within voluntary organisations.
Police officers were recruited from a range of locations in Scotland and were either specialist domestic abuse investigation officers within local policing divisions or worked in the national Domestic Abuse Co-ordination Unit. In respect of England and Wales, officers were recruited from two police forces – one based in the Midlands and the other in the South of England. These officers were either based in a local specialist domestic abuse unit or in departments which gave them exposure to domestic abuse investigations, and so were distinct from frontline officers who are first responders to calls of domestic abuse incidents. We received permission to recruit police participants from senior colleagues in each force area and secured approval through the individual forces’ processes. A designated point of contact provided the researchers with a list of colleagues of varying seniority for the research team to approach.
Prosecutors (Procurator Fiscals) were based in various Scottish locations, having a wide range of experience across a large number of domestic abuse cases. While access was also requested to interview Crown Prosecution Service (CPS) colleagues in England and Wales, it was not possible to secure such access in the timescales of the research. To recruit the Scottish prosecutors, we first secured permission from the Crown Office and then issued an open call, supplemented by targeted approaches to specific prosecutors identified to us by others. Applying a snow-balling technique based on the researchers’ existing networks, highly experienced third-sector specialists were also recruited in both jurisdictions who worked in a variety of capacities for leading domestic abuse organisations providing advocacy and/or support services to victims.
In England and Wales, our judicial participants were magistrates, who are lay judges and a section of the criminal justice community in relation to whom there is little available research on their approach to domestic abuse outside of specialist courts, with the most recent study carried out over two decades ago (Gilchrist and Blissett, 2002). Given the timeframe for this project, we decided to focus on magistrates, rather than the professional judges who sit in the Crown Court. The s. 76 offence is triable either way and consequently is heard at either the magistrates court when charged as a summary offence or the Crown Court when charged as an indictable offence. As retention of cases at the magistrates’ court is encouraged by mode of trial guidelines, it is likely that most s. 76 cases are heard by the lower court (CPS, 2023). Having secured approval to conduct judicial research with magistrates by the Judicial Office, we approached – by email – Regional Leadership Magistrates who cascaded a recruitment request to members and identified individuals for us to contact. The result was a mixture of magistrates who sat in specialist domestic abuse courts and those who did not. Sheriffs in Scotland are professional judges, unlike the lay magistracy in England and Wales, and they hear the majority of criminal and civil cases in the sheriff court. Our sheriff participants were recruited once permission was received by the Scottish Courts and Tribunals Service who also assisted us, through head clerks in each sheriffdom, with approaching judicial participants. The sheriffs were involved in hearing cases across the country and had extensive experience presiding over criminal trials involving domestic abuse and coercive control offences.
The gender of our participants were generally mixed except for those working for domestic abuse voluntary organisations all of whom were female. Of the 7 magistrates, 5 were male and 2 were female; and in respect of Sheriffs all 6 were male (in 2021 only 29.31% of the total number of Sheriffs were female). 2 Four of our 6 prosecutor interviewees were female, while 11 of our 18 police participants were male. Across policing and prosecutor participants, only 1 had been involved in their role for less than 5 years, with 4 of the prosecutors and 12 of the police officers having worked in the service for more than 10 years.
The offences
It is important to acknowledge from the outset that, though intended to address the same problem, the coercive control offences in the two jurisdictions have very different designs. In England and Wales, the offence is contained in s. 76 SCA 2015. It is committed where a person (1) repeatedly or continuously engages in behaviour towards another to whom they are personally connected that is controlling or coercive, (2) that behaviour had a serious effect on the other person such that it caused them serious alarm or distress having a substantial adverse effect on their daily activities or causing them to fear on at least two occasions that violence would be used against them, and (3) the perpetrator knew, or ought to have known, that the behaviour would have this serious effect on the person to whom it was directed. The term personally connected is now defined in s. 2 Domestic Abuse Act 2021 and includes current and former intimate partners and family members. In Scotland, the relevant offence was introduced by s. 1 Domestic Abuse Scotland Act 2018 (DASA). While it avoids the specific terminology of coercive control, unlike s. 76, it provides a detailed description of the forms of behaviour that it seeks to target, which map broadly to Stark’s (2007) coercive control paradigm. While s. 76 sought to target behaviour that was not criminalised under existing laws (Wiener, 2023), the DASA provisions adopted a more flexible approach, seeking to address the wrongdoing at the heart of domestic abuse, even where this involved behaviour already criminalised under existing offences (Burman and Brooks-Hay, 2018). Section 1 of DASA applies only to current or former partners rather than wider family relationships (Cairns, 2017; Cairns and Callander, 2022) and is committed when a person (1) engages in a course of behaviour which is abusive to a partner or ex-partner, (2) they intend or are reckless as to whether that behaviour will cause physical or psychological harm, and (3) a reasonable person would consider the behaviour likely to cause such harm to the recipient. ‘Abusive behaviour’ is defined as that which is violent, threatening, or intimidating, or which has, or would reasonably be understood to have, as one of its purposes making the recipient dependent on or subordinate to the perpetrator, isolating the recipient from sources of support, controlling, or regulating their daily activities, depriving or restricting their freedom of action, or otherwise frightening, degrading, humiliating, or punishing them. Unlike s. 76, there is no requirement in DASA for the prosecution to prove that the abusive behaviour caused harm to, or impacted, the victim.
Identifying coercive control
Implementing coercive control offences, regardless of the specificity of their definitional framing, requires professionals first to identify the conduct as a crime and record it as such. Shifting from an incident focused model to a longer view of patterns of behaviour presents challenges for identification. Wiener has explored the difficulty of using a new tool like s. 76 to make visible that which was previously invisible, such as abusive behaviour characterised by non-physical methods (Wiener, 2017). She has documented how the failure to properly define behaviour may result in it being ‘cloaked’ and recorded under different crime types, increasing the chances of coercively controlling behaviour being missed (Wiener, 2023). Other research has similarly suggested that identifying a pattern of abuse for the purposes of s. 76 can be challenging for police officers, who may continue to focus on isolated incidents rather than the wider context in which those incidents have occurred (Barlow et al., 2020; HMICS, 2023; Myhill et al., 2023a). Meanwhile, Robinson et al. (2018) have reported that officers placed greater emphasis on physical violence with injury when presented with vignettes depicting different forms of coercive and controlling behaviours. Despite this, our participants, excluding the third-sector interviewees, expressed the view that there had been significant changes in policing approaches and attitudes to domestic abuse cases over the last decade, accelerated by the introduction of these new offences.
For example, as one officer put it, ‘as a young PC attending domestic abuse cases back in 98, it [domestic abuse] wasn’t glossed over but it wasn’t considered as serious as it is now . . .. [T]hankfully, things have changed’ (Police 1, male, Midlands). Similarly, a prosecutor noted that in Scotland when they started as a procurator fiscal in 2005, ‘there was no point almost, of police investigating all of the nasty little things that perpetrators would do to make people’s lives more awful and to control them because that was not a crime’ (Prosecutor 2, female).
The introduction of coercive control offences was considered an important driver of change. In the Scottish context, for example, Prosecutor 5 (female) explained that since the introduction of the s. 1 DASA offence, the Crown Office were receiving more cases from the police than they had anticipated. From Prosecutor 6’s perspective (male), the DASA legislation had meant that ‘people are more aware of things than perhaps they used to be and it’s kind of very much on people’s radar’. It is interesting that the prosecutor should use this terminology, given it has been suggested that the use of out-of-court disposals in domestic abuse cases demonstrated they are ‘off the radar’ (Westmarland et al., 2018). However, the view that criminalising coercive control had increased awareness of, and their capacity to respond to, non-physical domestic abuse was also echoed by all the judicial participants in Scotland. For example, Sheriff 1 (male) noted that ‘prior to the enactment of things like the DASA, that sort of controlling, psychological, manipulation, although it was always there, wasn’t really something that the criminal law could tackle’.
These quotes suggest that the criminalisation of coercive control has enjoyed a degree of success, enabling some professionals to better identify domestic abuse and take it seriously. However, shifts in attitudes and understandings will not translate into improved outcomes unless there is consistent and confident identification throughout the whole justice process. In this context, it should be noted that low confidence was expressed particularly in the ability of magistrates in England and Wales to identify and tackle coercive control. Magistrate 1 (male) said, ‘I think there would be quite a large swathe of people, magistrates, who would struggle to articulate what [coercive control] was from any kind of knowledge base, rather than just in their head’. This view was echoed by a domestic abuse organisation representative (Third Sector 6, female) who also had prior experience serving as a magistrate – ‘[I]t’s walking on eggshells and not recognising that as a magistrate is really concerning because they are the ones who are the decision makers’. The magistrates that we interviewed often regarded themselves as more knowledgeable about domestic abuse than many of their peers, because it was an area in which they took a particular interest. However, in reflecting on the attitudes and practices of their peers, concerns were commonly raised about the identification of coercive control, despite the suggestion that domestic abuse is treated ‘as more serious’ than any other crime in the Magistrates’ court (Magistrate 3, male).
In line with previous research that highlights a priority afforded to physical violence in coercive control cases (Barlow et al., 2020; Robinson et al., 2018), we found that professionals (for example 9 of the 12 English officers) often expressed the view that it is easier to build and prosecute cases involving physical assault and injury: ‘it is difficult to charge when it’s not physical abuse’ (Police 3, Male, Midlands). At the same time, the need to recognise non-physical coercive control was also noted: As Magistrate 2 (male) put it, the fact that coercion is not necessarily physical coercion, abuse is not necessarily physical abuse, and it’s in the legislation, so we’re seeing it. The definitions are no longer just, you’ve hit someone, or you’ve slapped somebody, it’s all of the other areas as well. So, the legislation has changed, so we have to change with it.
There is, of course, more to the concept of coercive control than merely understanding that abusive behaviour can include non-physical manifestations and involve a systematic pattern of behaviours designed to reduce the victim’s space for action (Sharp-Jeffs et al., 2018). Stark (2007) considers coercive control to be a gendered concept that operates through the power dynamics of expected gendered roles in an intimate relationship, falling at the extreme end of those relations that exist in ‘normal’ family life. In Stark’s view, coercive control as a gendered phenomenon has become more prevalent as women gained steps towards equality in wider society. The normalised expectations of gendered roles mean that coercive control can be difficult to discern from accepted behaviours in relationships (Bishop and Bettinson, 2018: 8). Consequently, this has led Tolmie (2018) to question whether criminal justice professionals would identify coercive control in such contexts: ‘[i]f abusive behaviour exploits existing gender norms when does ‘normal’ end and ‘abuse’ begin?’ (p. 56).
One factor that might separate the acceptable from the unacceptable in this context is the threat that accompanies the demand of the perpetrator. That threat may be expressed in a seemingly innocuous way such as a gesture, however, where there has been a previous experience of violence or serious damage to the victim’s autonomy, that gesture serves to remind the victim of the possibility that it will happen again (Dutton and Goodman, 2005). A threat may be an obvious one to an outsider where it implies death or serious injury or be more subtle and designed to affect the specific victim based on knowledge gained by the abuser during the relationship (Wiener, 2017).
Issues relating to gendered roles and the identification of coercive control were also raised by our participants. For two English officers in our study, the existence of a spectrum of normal gendered behaviours explained their view that coercive control could be too readily raised by complainants. For example, Police 7 (male, South) commented that people may say that they are being controlled every day, but really they are complaining that they don’t have the freedom to do what they want because they are in a relationship . . . at the end of the day . . . you may have responsibilities and you can’t just dump things and run off and go out with your mates or whatever all the time.
In the same vein, Magistrate 2 (male) said – If you’re husband and wife having a go at each other, is it just an argument? Is it just you’re in a bad mood with each other because you’re tired, you’re stressed, you’ve got children, you’ve got aggravations, whatever? Normal arguments, that’s not domestic abuse, that’s just relationships.
This officer and magistrate are correct that relationships involve compromises to work successfully and that is distinct from exerting coercion or control upon a partner. What is missing from their understanding here, however, is the need for a threat to the person’s autonomy to be present for it to be a coercively controlling relationship and indeed a need to consider the wider context of the relationship, such as the consequences should the victim resist compliance with the abuser’s demands. Identification of coercive control is hindered where criminal justice professionals draw ‘upon their own gendered values and assumptions relating to the division of labour [or sexual relations] in intimate partnerships’ (Myhill et al., 2023a). This was also prevalent in Myhill et al.’s (2023a) study in England and Wales where an officer trivialised a woman’s disclosure of sexual coercion by normalising male sexual entitlement. In Scotland, moreover, HMICS found that victims had reported inappropriate remarks from some responding officers, highlighting a lack of understanding about the dynamics of coercive control and gendered assumptions. Examples of such remarks included asking whether a victim was only reporting to discredit the reputation of an ex-partner or an officer observing that he could not put a man out of his house (HMICS, 2023).
Our interviewees also discussed examples where the alleged perpetrator might have been interpreted at first glance to be ‘doing something nice’ or ‘looking after’ the victim, in order to underscore the difficulties that can arise with identifying coercive controlling behaviour, or evidencing its existence in ways that jurors will be sufficiently convinced by. As Sheriff 1 (male) puts it, You know, he bought flowers every Friday, isn’t he lovely, but then when you set it against umpteen other weird and wonderful things that are occurring over that period, it suddenly doesn’t sound so lovely anymore. But if you said to somebody, oh he buys her flowers all the time, it’s hard to see how that’s a bad thing; but it can be.
Crowther-Dowey et al. (2016) have previously raised this concern noting that the perpetrator’s behaviour can be confused and misinterpreted as an indication of love and affection. In this regard, while prosecutorial guidance does clarify that context is important, it continues to be a concern that such context may not be easily understood by professionals or by jurors who have had relatively little prior experience of domestic abuse (Bishop and Bettinson, 2018).
Also problematic here is the fact that these gendered norms ‘are culturally specific and influenced by the shared history and circumstances of particular communities’ (Tolmie et al., 2024: 62). Scholars have thus argued for a further reframing of coercive control by locating it within a model of social entrapment (Tolmie et al., 2018, 2024). This would look beyond the behaviours adopted by individuals and consider the safety responses available to victims, as well as how the complex intersection of inequities associated with race, class, sex, gender, and other forms of oppression inform both their experience of the abuse and their realistic expectations of agency responses. Wilson et al. (2019) explain, for example, that experiences of institutional racism mean that a victims’ experiences of state violence and neglect may be more harmful than, and compounded by, violence from abusive partners.
In our interviews, representatives from domestic abuse organisations were typically sceptical that justice professionals were consistently identifying non-physical tactics used in domestic abuse, and some highlighted a concern that they were particularly poor at doing so in cases involving minoritized women. In this context, culture was seen as obscuring the abuse in the eyes of many professionals, leaving the women increasingly vulnerable. As one interviewee put it, women [who have no recourse to public funds] are quite often expected to be at home, be a stay home mum, because that’s just something that is being told to them, that that’s the expectation from them as a mum. That creates lots of avenues for the perpetrators to abuse women financially but kind of present that . . .[as] not abusive behaviour to start with, but also, normal practice within relationships, within families. I think the perpetrators are very skilled in justifying their abusive behaviour. (Third Sector 5, female)
This interviewee went on to underscore that financial control can manifest in food rationing and the woman ‘not being allowed to buy sanitary products’, for example. Physical violence can be disguised by restricting women’s access to medical assistance to private ‘culturally specific gynaecologists, dentists, GPs’, and poor responses from state agencies when victims come to their attention (for example, through police call-outs) can increase the danger they are exposed to while also undermining their expectations of ‘rescue’. This is crucial to understanding the victim’s circumstances and her subsequent responses to the abuse. A further barrier to identifying domestic abuse for minoritized women that was highlighted was the use of ‘interpreters from their community, who they are likely to know’ (Third Sector 2, female). This may well undermine victims’ ability to disclose freely and fully. Moreover, even where there is not this concern, there are other challenges associated with the mediating role of interpreters in flows of communication. One interviewee recounted, for example, a case in which ‘we know . . . where there was somebody else in the courtroom who spoke that language and the interpreter wasn’t actually even interpreting what she said’ (Third Sector 2, female).
Participants also commented on the prospect that victims may often not recognise that they were being subjected to coercive and controlling behaviour. However, only those from domestic abuse organisations highlighted the extent to which this may be a particularly acute problem, or have particular manifestations, within specific cultural communities. A lack of appreciation of this can impact on how domestic abuse complaints, and complainants, are responded to (Graca, 2018); and as such, our findings suggest that even where the offences may be transformative, they are not equally so for everyone.
Training
The need for specialist training on coercive control for criminal justice professionals has been highlighted by a number of commentators (Bishop and Bettinson, 2018; Robinson et al., 2018). The importance of this, particularly in order to facilitate successful identification and prosecution of cases where there is a lack of physical violence, was discussed by all of our participants. It was clear, however, that different approaches had been taken towards the provision of training during the roll out of the new legislative regimes in the respective jurisdictions. In England and Wales, individual police forces have autonomy over their training allocation and are not mandated to provide training specifically on coercive control, although the Domestic Abuse Matters Programme, led by the College of Policing, is available to all forces (Brennan et al., 2021). Engagement with e-learning modules by officers has been encouraged by the College, but Brennan et al. (2021) have noted that ‘it is not possible to know how many viewed it. As with similar mandated e-learning in police forces it is likely that the take-up was limited’ (p. 1157). This story of limited provision of training was certainly in keeping with what we were told by police officers in England as part of this research. They observed that they did not feel they had received adequate training on coercive control, even where – in some cases – they had undertaken online training when the s. 76 offence was introduced. As one officer put it, we did have some online packages, just to brief us about what the offences, what sort of things to look for. So, like the pattern of assaults and what have you. . .that was pretty much blanket training for all officers. Apart from that, I’m struggling to think of any times I’ve had specific training about coercive control, to be honest. (Police 3, male, Midlands)
This reflection on the limitations of online training, where it is undertaken, corresponds to findings from Brennan et al. (2021) that the form training takes influences the learning achieved, with a greater increase in arrests for coercive control where there was mandatory, force-wide in person training on the Domestic Abuse Matters programme, compared with e-learning. At the same time, it is also important to caveat that with the further finding from Brennan et al. (2021) that increases in arrests following such training were temporary, as rates began to reduce after 8 months highlighting a need for a rolling programme of training. Likewise, though Wire and Myhill (2016) found that College of Policing provision of a 1-day classroom-based training for first responders in England and Wales had a positive effect for some indicators of knowledge and understanding of coercive control, there was no apparent impact on officers’ general attitudes to domestic abuse. All of this underscores the importance of ongoing and robust training initiatives.
In contrast, Scottish police participants spoke markedly more positively about the mandatory Domestic Abuse Matters training they had received so far, which they often connected quite directly to a confidence in their ability to investigate DASA offences. Coinciding with the roll out of the DASA legislation, 13,510 frontline officers and call handlers were provided with 1-day core training (HMICS, 2023; Scottish Government, 2017), while comparable training was mandated for all prosecutors before they could handle a domestic abuse case (COPFS, 2018). Integrated within that training, moreover, was involvement and input from third-sector specialists and those with lived experience, which participants underscored had ensured a more powerful and impactful result. As prosecutor 1 (female) put it, ‘[we heard] presentations from victim support organisations and other people who’ve had lived experience. And . . . part of our training now involves actually spending a half day with a victim support service as they help victims’. HMICS (2023) considers the incorporation of the lived experience of victims to be an essential element to all domestic abuse training. However, HMICS’s (2023) Thematic Review of Domestic Abuse has also recommended further training for Scottish police officers, to eradicate outdated attitudes and ensure they consider the background of an incident to determine whether it forms part of a course of behaviour.
While, in Scotland, the Judicial Institute provided the judiciary with compulsory training to support the roll out of DASA legislation – consisting of online training and a face-to-face course, with input from domestic abuse organizations, for example, Scottish Women’s Aid (Scottish Housing News, 2019) – as with their policing and prosecutorial counterparts, the degree and consistency of training provided to magistrates in England and Wales is less transparent.
This in turn has generated a greater risk of fracture and tension across different stages of the criminal justice process in England and Wales than might arise in Scotland where a ‘whole system’ approach has at least been recognised as the gold standard, reflected in an organisational priority to coercive control offences. In England and Wales, there was certainly evidence from our participants of a sense of frustration and demotivation among police officers in the fact that they felt prosecutors were too hesitant to take forward cases. In the next section, we will explore in more detail this question of professionals’ perspectives in relation to evidencing-gathering for coercive control offences within the two jurisdictions.
Lengthy investigations and evidence-gathering for coercive control offences
Coercive control complaints recorded by police are often still not resulting in a charge. Indeed, in England and Wales, the proportion that have done so has now fallen to less than 6% and research indicates that coercive control cases are more likely than other domestic abuse cases to be closed due to evidential difficulties (Brennan and Myhill, 2022). Despite some important differences between them, the design of offences under both the SCA and DASA is such that evidence is required of a pattern of behaviour represented as ‘repeated or continuous’ (s. 76) or ‘a course of behaviour’ (s.1) that can include either physical or non-physical abuse. As noted above, this differs from traditional criminal offences that require evidence of a single incident, that is generally physical in nature. Harassment and stalking differ to this requiring evidence of a course of conduct, although judicial interpretation in England and Wales has limited its potential of capturing a pattern of behaviour by requiring two or more incidents that amount to harassment, having sufficient connection to each other (Bettinson and Bishop, 2015). For the coercive control offences, officers in both jurisdictions have been required to gather various types of evidence, building cases in a way that may be new to them, and which often generate larger, longer, and more complicated investigations.
It was clear, however, that this was met with greater reservations by the English officers in our sample than their Scottish counterparts. Police 11 (male, South) said, for example, its’ an awful lot of work to prove to the criminal standard . . . you have a very large case file with an awful lot of materials, for then what ends up being treated in the courts as not the most serious of offences.
This factor could reduce officers’ motivation for pursuing cases, which is needed to complement skills developed in training (Brennan et al., 2021). Other officers in England similarly expressed concerns regarding the limited time and resources available to them to investigate coercive control cases. Police 10 (female, South) explained that investigations are ‘very in depth, you have to go through looking back months, years build a lot of evidence in various things, and there’s just no way officers have the time to be able to put those kind of cases together’. Although participants noted that specialist units might have greater capacity, they observed that this was not always realistic in practice since ‘it’s that funding, the time, we haven’t got any officers’ (Police 12, female, South). Where officers are first responders, without sufficient domestic abuse training, there will be missed opportunities to investigate coercive control offences (Myhill et al., 2023b). Lack of policing resources was also identified as a substantial obstacle by other professionals, with Magistrate 6 (male) commenting, for example, that ‘when it comes to the bobby on the beat it’s a case of resources, it’s a case of time’. Likewise, third-sector domestic abuse organisations acknowledged that the temporal design of the s. 76 offence represented a change in approach which had ‘implications for investigations and timeframes, at a time when the criminal justice system is really stretched’ (Third Sector 1, female).
This position contrasts strongly to the tone of responses from Scottish officers in our study, who – while also noting that inquiries ‘had become much bigger than they would have [been] a few years ago’ (PSCOT 2, female) – were more sanguine about their capacity to conduct them. As one officer put it, ‘to have the time to go and do that, it’s something that we here have, we’re lucky, we’ve got a luxury that we can go in and give ourselves times to do that’ (PSCOT 5, male). Prosecutors similarly confirmed that officers were building larger cases as a consequence of DASA but were broadly positive regarding their skills and capacity to do so. As Prosecutor 2 (female) noted, in their view, police are good at ‘proactive investigations’. Importantly, our Scottish police participants all worked as specialist domestic abuse investigators and research indicates that their experience of time and resources does also contrast with that of non-specialist responding officers. In that context, HMICS (2023) has continued to underscore that responding officers need to recognise ‘the possibility of a DASA in the first place’ otherwise ‘. . . there is a risk of missing investigative opportunities’; (p. 88) and that adequately resourcing specialist domestic abuse units is therefore a high priority.
Corroborative evidence
One possible explanation for this apparent difference in attitude towards building large cases for coercive control investigations (alongside the priority afforded to specialist resourcing) may lie in the existence of a corroboration rule in Scotland, that has been abandoned in England and Wales. In Scotland, all crimes require corroboration meaning there must be at least two independent forms of evidence to support the essential or crucial facts of the case ‘including that a crime was committed and that it was the accused who committed’ it (Cairns, 2020: 397). In domestic abuse and sexual offence cases in particular, this has been a potential obstacle to securing criminal convictions, given that much of the prohibited conduct occurs in private. The fate of the corroboration rule in Scotland remains contested: the Scottish Government made recommendations to abolish the rule in 2011 (Scottish Government, 2011), which were later abandoned (Scottish Government, 2015), but the topic has been opened for consultation.
Cairns (2020) has explained that the design of a ‘course of conduct’ model within DASA could be particularly problematic under the
Scottish police officers talked about gathering a variety of different forms of evidence, beyond the victim’s testimony, in a positive way. This evidence was considered to be effective according to Prosecutor 2 (female) who believed that the ability to show that others had experienced the same behaviour by the abuser was not only corroborative but helped to redress victim-blaming attitudes in court. As one officer out it, by approaching previous partners of the suspect, evidence could be gathered so ‘we’ll not struggle to prove on pretty much the majority of the stuff because they act the same way with most victims’ (PSCOT 3, male). In making these approaches, however, the need for empathy towards previous victims was also identified, with PSCOT 4 (male) showing an awareness that relaying past events could trigger victims’ trauma, or that–where the perpetrator was still in their community–the witness could justifiably fear reprisals for talking to the police. Despite these concerns, however, officers emphasised that it was necessary to pursue cases, since otherwise victims would be left at risk and ‘the state had a duty to look after victims of abuse’ (PSCOT 1, female).
Prosecutors did acknowledge that the corroboration rule could be ‘quite tricky’ (Prosecutor 3, male) in relation to coercive control, particularly as more than one incident had to be supported with two pieces of evidence and the incidents typically took place behind closed doors (Prosecutor 6, male). Prosecutor 1 (female) observed, for example, that a physical assault would be easier to corroborate because there would be evidence of an injury, or the s. 38 Criminal Justice and Licencing (Scotland) Act 2010 offence where someone overheard or saw an incident. However, the prosecutors generally endorsed the police view that corroboration requirements were not a significant problem in practice and were broadly positive about the police’s ability to gather sufficient evidence. Prosecutor 4 (female) noted, for example, that ‘the police are good at looking for [evidence] and exhausting all lines of inquiry’. Separately the evidence may be circumstantial, but together ‘it is quite compelling’. Other prosecutors underscored that this focus on ‘looking at the bigger picture’ (Prosecutor 1, female) was not something that had been generated particularly as a result of DASA, since in their view this was police practice previously also. At the same time, it was noted that the passing of DASA had created a greater enthusiasm among officers to investigate the s. 1 offence, sometimes to the extent that cases were sent ‘to the prosecutors that were not actually s. 1 DASA cases’ (Prosecutor 5, female). This was often because the abuse in question spanned a period before and after the implementation of DASA, and although prosecutors were able to ‘split the charges’ in such cases, it was suggested that it could make it more difficult to lead evidence at court when having to dance ‘around dates and times’ and some felt it risked the impression they were ‘overzealous prosecutors’ (Prosecutor 6, male).
This contention that Scottish police and prosecutors were being ‘overzealous’ in bringing coercive control cases to court also had some traction with Sheriff 3 (male), who suggested that cases he adjudicated were often weak, given the high standard of proof required to find guilt. However, the same participant did also concede that the DASA offence had made ‘it easier for the person bringing the case to establish the relevance of a particular line of evidence, which might otherwise, maybe twenty years ago, have been seen as irrelevant’.
Evidence-gathering and evidence-led prosecutions
When asked to consider what made for a robust case, police officers in England and Wales – more so than their Scottish counterparts – tended to focus predominantly on evidence gathered from the victims themselves, with less consideration afforded to other forms of evidence, such as third-party material. Police 1 (male, Midlands) observed, for example, that while ‘you can’t blame’ victims who retract cooperation, ‘it does tend to take your evidence away with it’ and Police 4 (female, South) noted that ‘I wish there was a way in which we could kind of capture that evidence better with disengaging victims’. While Police 11 (female, South) described the process as involving ‘long and intrusive statements’ from victims which was an ‘all encompassing experience for the victim to be subject to’, it was clear that the train of thought for English officers was often that coercive control cases require evidence from the victim in order to be realistically pursued. In part, this may be encouraged by the design of s. 76 SCA which requires proof that the prohibited behaviour had a serious effect on the victim (s. 76(4)), entailing either that they feared violence would be used against them on at least two occasions or that their everyday activities were adversely affected substantially. In contrast, the DASA offence has not made the impact upon the victim a constituent part requiring proof. Police 1 (male, Midlands) highlighted that demonstrating that the cause of the serious effect under s.76 was the perpetrator’s coercion and control could be difficult because ‘other things happen to people in that time’. Addressing this may place a particular premium on the account of the victim as to the ways in which their behaviour was impacted or the fear under which they operated.
Participants from third sector domestic abuse organisations also noted that, in cases of non-physical abuse in particular, victims may be unable ‘to lead the police towards the evidence because some of the abuse, it’s everyday life’ (Third Sector 5, female). As a consequence, they were unconvinced as to the ability of police to navigate these cases appropriately. Linking to the difficulties of identification discussed above, one IDVA observed, for example, that if somebody turns up and says, my husband’s broke my arm, . . . the officers go, great, because there’s medical records, I understand this, it’s physical abuse, it’s absolutely not acceptable. . .But when somebody comes to them and says, over thirty years this is what has happened . . . [they] don’t know what to do. (Third Sector 2, female).
This difficulty for victims in being positioned as the primary source of investigative material was not appreciated by Magistrate 5 (male), however, who believed
It is well documented, of course, that victims find it difficult to engage, and continue to engage, with criminal justice processes (Cretney and Davis, 1996; Porter, 2019). The reasons are myriad, but common features are that they may be fearful of reprisals from the perpetrator, or the perpetrator may manipulate them into retraction. In recognition of this, policing and prosecution policies have – with varying degrees of enthusiasm – encouraged the use of evidence-led cases (HMCPSI, 2020). Although such prosecutions should not be undertaken lightly, given the safety implications for victims, a third sector participant in our study was of the view that such opportunities were under-used (Third Sector 4, female).
Officers in one of the English police forces in our study discussed the use of evidence-led prosecutions. Police 8 (female, South) suggested they were often undertaken, but this could be contrasted to Police 5’s (male, South) view that they were not popular as they had not proven to be particularly successful. This ambivalence was also echoed among magistrate interviewees, several of whom acknowledged the significance of pursuing evidence-led cases and observed that it was ‘good’ to see them at court (Magistrate 4, female). However, ultimately, they also considered that such cases were difficult to prosecute ‘without a witness statement’ (Magistrate 4, female).
Overall, then, the impression given by our English participants was of an ongoing reliance on the victim-witness to provide evidence in order for a case to proceed under s. 76 SCA. This may undermine the goal of the offence to the extent that its implementation can be hampered by the very effects of coercive control itself. In contrast, the Scottish landscape was depicted by participants as being somewhat different in this respect. Although the corroboration requirement continues to have its detractors, it was suggested by our Scottish participants that its pre-existence as part of the criminal justice infrastructure in Scotland meant that police were less phased by the need to undertake larger and more complex investigations in coercive control cases than their English counterparts for whom the idea of triangulating evidential sources by necessity to make out an offence was more novel. Rather than discussing what evidence the victim could provide them, Scottish police and prosecutors perceived the system as a ‘victim-centred not victim-led’ one (Prosecutor 1, female). In addition, the different design of the offences across the two jurisdictions meant that police officers in England and Wales were more conscious of the need to prove that the victim was harmed by the prohibited behaviour, encouraging a focus on the victim’s evidence. At the same time, as noted above, the confidence expressed by Scottish police was also clearly linked to the more expansive training they had received on the DASA legislation, as well as perception of greater resourcing and capacity in specialist units than in England and Wales.
Victim-related issues
Participants in both jurisdictions raised similar victim-related issues in our data. For the police in England, their responses generally focussed on evidence, or the lack of, provided by the victim and issues that they felt negatively impacted on the victim’s credibility such as mental health or substance misuse. They also intimated that they felt the CPS were too heavily focussed on factors seen by them to undermine the credibility of the victim in the eyes of the jury. Bishop and Bettinson (2018: 15) have highlighted how the traumatic effects of domestic abuse will often make the process of giving evidence difficult for victims. Physiologically, trauma effects memory processes making it difficult to recall events in a coherent, detailed, and consistent manner and this may imply a lack of credibility to prosecutors and others who are not appropriately trained. Despite this being well-established and recognised in CPS policy, Roberts and Saunders (2010) found that decision-making often continued to be influenced by problematic perceptions of witness credibility, impacting upon case progression. Although training on the impact of trauma on memory is now routinised by the CPS in respect of RASSO cases, police in our study remained concerned about understanding in domestic abuse cases. As Police 2 (male, Midlands) put it, ‘I think we [police and prosecutors] come at it from different angles . . . As soon as they see a challenging victim, it’s almost as if, well we’re on a no hoper here. They’ll say your victim isn’t reliable’. Likewise, Police 8 (female, South) remarked in relation to CPS decision-making: ‘If it’s just a witness saying this happened, it’s like they don’t believe them or they think they’re not quite good enough witnesses’.
Since we were unable to secure CPS participation in this study, we obviously have not been able to probe this further with prosecutors to obtain a clearer sense of their current approach, and to evaluate the extent to which policies around trauma, myths and stereotypes might have impacted. In the Scottish context, however, input from prosecutors did similarly highlight the significance of victim credibility. Prosecutor 5 (female) described, for example, how victims’ presence at court could ‘make or break the case’. And in terms of the trial experience itself, there is recent evidence in Scotland that underscores the extent to which victims do indeed continue to find that process difficult and potentially re-traumatising (Scottish Government, 2023b). In particular, the experience of cross-examination within the adversarial trial has been highlighted as being substantially distressing, with victims reporting that it ‘undermine[s] their ability to provide their best evidence’ (Scottish Government, 2023b: 26).
Sheriffs we spoke to in this study also provided insights that highlighted the pressure placed on victims when giving oral testimony, which could frustrate their ability to tell their story in a meaningful way. Sheriff 6 (male), for example, thought that the trial may come at a stage in the victim’s recovery where they feel ‘defeated’ and not prepared to share their experience. While this Sheriff indicated that he would ‘give a prosecutor as much leeway as I possibly can’ to enable them to elicit the complainer’s account, a particular difficulty has been identified in relation to prosecutors’ and sheriffs’ willingness to interject during cross-examination or to address intimidating behaviour by the accused at court which can substantially undermine complainers’ capacity to give testimony (Scottish Government, 2023b).
Considering the alternative of an evidence-led trial without the victim providing oral testimony, Sheriff 4 (male) was sceptical of their likely success, ‘the fiscal (prosecutor), with all the best efforts. . . in the world, is struggling without the victim coming to court’. The same Sheriff felt that the presence of the victim-complainer in court was very persuasive. Even where victims were willing to share, trials were lengthy, as cases were complex with large amounts of evidence, leaving Sheriff 6 (male) feeling that the criminal justice process did not serve victims well, particularly as they became uncertain in their accounts where so much detail was required. Thus, while the s. 1 DASA offence was considered easier to evidence by police officers in Scotland than their English counterparts in respect of the s.76 SCA equivalent, and there was greater evidence of collaborative working across police and prosecutors north of the border, supported by a mandatory roll out of initial training, in both settings successful enforcement may yet be stifled at the latter stages of the criminal justice process.
Conclusion
Although a small-scale study, our data give a fresh insight into the implementation of coercive control regimes across Scotland, and England and Wales, from the perspective of those most closely involved in its operation. Our findings suggest that the introduction of coercive control offences has increased some criminal justice professionals’ awareness of the different forms domestic abuse can take and enhanced the ability to identify coercive and controlling behaviours. However, this ability is not uniform across the professions, with particular concerns expressed about the lay magistracy in England and Wales, who have not benefitted from the roll out of consistent training. Gendered assumptions were present in some of our participants’ responses and specific complexities for minoritized women were not recognised, indicating that the offences have not been equally transformative for all victims of abuse. Continued improvements require consideration of different training approaches to address normalised, outdated attitudes, as suggested by HMICS (2023: 7). The difference in training experiences around the respective offences was marked in the two jurisdictions and appeared to impact on the level of practitioner confidence as well as the willingness of police officers to build cases, gathering large amounts of a wide variety of evidence essential to these prosecutions. In Scotland, police and prosecutors described positive experiences of training, which was mandatory and timed with the roll out of the domestic abuse offence. The influence of training on Scottish sheriffs was less evident. Police officers and Magistrates in England and Wales had received lower levels (or no) training and consequently expressed less willingness and confidence to build cases (police), or knowledge about how the offence should operate (magistrates). For police participants in England, moreover, there was a sense of frustration that, in their view, coercive control cases rarely had a satisfactory outcome, given the effort made in investigating them, and this affected their motivation to pursue such cases. Their contributions indicated scope for improving the joint working between forces and CPS on domestic abuse and the operation of the coercive control offence specifically, developing clarity as to the evidential thresholds that must be met to proceed with a prosecution. Finally, our research highlights a difference in attitude towards the position of the victim in the investigation and prosecution of these offences in the two jurisdictions, reflected, in part by their different designs. In England and Wales, greater emphasis was placed on the victim’s perceived credibility and ability to provide evidence themselves, while in Scotland, at least at the police level, there was greater confidence in the ability to build cases without reliance upon the witnesses’ statement when required. However, this development may have little effect on the outcome of a case that goes to trial, where prosecutors and sheriffs often continued to consider the victims’ presence at court central.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors received financial support for the research of this article from the University of Warwick, through a Faculty of Social Science Research Development Award.
