Collaboration, risk and ‘just’ outcomes: challenges and opportunities in policing HBA

ABSTRACT

Honour-based abuse (HBA) is a complex phenomenon breaching the boundaries of the domestic abuse and violence against women frameworks within which it is dominantly situated. Although HBA is recognised as a multi-agency issue, police are ultimately responsible for responding to reports of known or suspected HBA. In 2015, a HMIC report highlighted that the police still have some way to go in terms of understanding HBA and adequately protecting victims. This paper presents findings from a project on the early identification of HBA, which generated evidence surrounding the key challenges encountered by police responding to HBA cases that may or may not have been identified as such, and recommendations on how practice can be improved. The focus is upon three key areas emerging as in need of attention: Firstly, negotiating the complexities posed by the frequently collective and collaborative nature of HBA; secondly, conducting effective risk assessments; and thirdly, ensuring a ‘just’ outcome for victims of HBA. The final part of the paper outlines opportunities for police policy and practice to be enhanced so that ultimately, victims of HBA can be more adequately safeguarded.

KEYWORDS:

• HBA
• risk assessment
• early identification
• early intervention

Introduction

Honour-based abuse (HBA) refers to a broad range of violent and/or abusive behaviours that are carried out to protect, defend, or restore socially-constructed notions of community, family or individual honour (Gill 2008, Her Majesty’s Inspectorate of Constabulary [HMIC] 2015). Since the late 1990s, policy and legislation have gradually emerged surrounding how HBA should be responded to within the UK, with a particular focus on the role of police and partner agencies in working together to protect victims (HMIC 2015). Although multi-agency working has been highlighted as essential in addressing this often-hidden form of abuse (HMIC 2015, p. 48), police remain the first port of call for many victims in a state of crisis, and for people wishing to report suspected or known incidents.

HBA is an extremely complex phenomenon presenting many challenges to the police, and UK policy and practice has been described as both world-leading and problematic (Begikhani et al. 2015, p. 18). In the 2015 HMIC report on police responses to HBA, several improvements were recommended, including: improving knowledge and awareness of HBA and how to protect victims; ensuring consistent and effective risk assessment; engaging with communities and promoting public confidence in the police; and enhancing partnership working. Stemming from these recommendations, the findings presented in this paper derive from a project conducted between 2018 and 2019, which sought to improve the early identification of, and response to, HBA by police in one English force. The data were generated through in-depth interviews with police, support workers and victims/survivors of HBA, and from analysing over 300 police case files.

We begin by discussing the ‘problem’ of honour-based abuse, including the nature of the phenomenon and the frameworks within which it tends to be understood, whilst contextualising the issue through a review of the development of UK policy and practice, and existing literature on responses to HBA. The findings from our research develop this literature by focusing on the challenges experienced by police in responding to identified cases of HBA, the missed opportunities revealed in our data, and the potential to improve policy and practice in order to enhance understandings of how to obtain ‘justice’. The discussion is organised around three key themes arising from the data: firstly, the difficulties experienced by police in responding to the collaborative nature of HBA; secondly, the importance of developing appropriate risk assessment training and practices that ensure consistency, accuracy, relevance and effective risk management; and finally, ensuring a successful outcome for victims. This entails addressing the high level of incidents that do not result in prosecution and adjusting assumptions about what may constitute a sense of ‘justice’ for victims.

Understanding and responding to HBA

HBA is notoriously difficult to define and lacks any statutory definition (Welchman and Hossain 2005, Gill 2009, Roberts et al. 2014). The definition used by the police and Crown Prosecution Service (CPS) in England and Wales is provided by the National Police Chiefs’ Council (NPCC 2015, p. 5):

Honour-based abuse is an incident or crime involving violence, threats of violence, coercion or abuse (including psychological, physical, sexual, financial or emotional abuse), which has or may have been committed to protect or defend the honour of an individual, family and or community for alleged or perceived breaches of the family and/or community’s code of behaviour.

Conversely, forced marriage (FM) was explicitly defined and criminalised in the 2014 Anti-social Behaviour, Crime and Policing Act:

A forced marriage is a marriage in which one or both spouses do not consent to the marriage but are coerced into it. Duress can include physical, psychological, financial, sexual and emotional pressure. In cases of some vulnerable adults who lack the capacity to consent, coercion is not required for a marriage to be forced. (Her Majesty's Government 2014, p. 3)

Policy and practice addressing HBA in the UK only began to develop in the late 1990s, previously going under the radar due to a lack of data, visibility and a reticence of the Government to intervene (Julios 2015). Following the activism of Black and Asian Women’s organisations (Julios 2015), a number of high-profile cases of forced marriage (Begikhani et al. 2010, Gill and Mitra-Kahn 2012) and ‘honour’ killings, particularly those of Heshu Jones in 2002 and Banaz Mahmod in 2006, the New Labour Government set up the Working Group on Forced Marriage in 1999, followed by the Community Liaison Unit (CLU) in 2000 (re-launched in 2003 as the Forced Marriage Unit); tasked with addressing the international dimension of forced marriage (Dustin and Phillips 2008). Mounting pressure on the Government, and a consultation on the criminalisation of forced marriage (Home Office 2012), led to it being criminalised under the Anti-social Behaviour, Crime and Policing Act 2014.

Against this backdrop, awareness of the broader problem of honour-based abuse has, over the past 20 years, gradually increased. An Independent Police Complaints Commission (IPCC 2008) inquiry into Mahmod’s murder highlighted that there had been numerous opportunities to intervene prior to her death, and the conclusions ultimately led to the first ever national policing strategy on HBA (see ACPO 2008), which was updated in 2015 (see NPCC 2015).

The 2008 and 2015 strategies provided a series of aims for police in responding to HBA, including protecting those at risk, taking effective action against perpetrators, embracing multi-agency working, increasing awareness of HBA, eradicating the causes, building trust among victims, protecting witnesses, mainstreaming HBA in policies, and providing high quality training (ACPO 2008). The 2015 strategy emphasised the role of collaborative working in ‘eradicating HBA, FM and FGM, increasing the confidence of victims, survivors and affected communities and in identifying, prosecuting and bringing offenders to justice’ (NPCC 2015, p. 6). However, the achievements of the strategies have been brought into question by Domestic Abuse Commissioner Nicole Jacobs and Victims Commissioner Dame Vera Baird, who highlighted the slow progress in responding to the HMIC report and failure to follow up on key recommendations; and conviction rates remain low (Khan et al. 2021, p. 7).

Although not specifically addressed, HBA is provided for within the Government’s Ending Violence Against Women and Girls Strategy 2016–2020 (Her Majesty's Government 2016), and also falls under domestic abuse policy and legislation. The Serious Crime Act 2015 incorporated HBA in its definition of Coercive Control, which refers to ‘controlling or coercive behaviour in an intimate or family relationship’, and the Domestic Abuse Act 2021 developed the previous cross-governmental definition of domestic abuse into a statutory definition of domestic abuse covering anyone related to each other. The implications of these Acts mean that any reported or suspected cases of HBA should be responded to under the framework developed for domestic abuse and coercive control, including risk assessment tools, pro-arrest/prosecution policies, safeguarding measures, and referrals to other agencies.

To an extent, subsuming HBA within domestic abuse legislation, policy and practice is understandable and beneficial. There are clear parallels between domestic abuse in (ex-)intimate partner relationships and HBA, including similarities in types of abusive behaviours, impacts upon victims, and the tendency for abuse to be perpetrated by men against women. Integrating the two types of abuse thus enables HBA to benefit from existing policies, practice and resources (Siddiqui 2007 cited in Eshareturi et al. 2014). Furthermore, conceptualising HBA as a form of domestic abuse circumvents cultural explanations of HBA, which tend to locate the phenomenon within the ‘otherness’ of ethnic minority communities and perpetuate the ‘us/them’ binary (Aujla and Gill 2014).

But there are problems with HBA being primarily understood and responded to as a form of domestic abuse. Despite the similarities, there are crucial differences, including the motivations associated with HBA, which tend to centre around notions of ‘honour’ rather than power and control per se; the often collaborative nature of HBA, which frequently involves extended family and the wider community (Sen 2005, Brandon and Hafez 2008, Wikan 2008, Chesler 2009, Gill 2009, Begikhani et al. 2010, Payton 2010, Belfrage et al. 2012, Keeping 2012, Cooney 2014, Salter 2014, HMIC 2015); and the consequent complexities in risk assessing and appreciating the challenges involved in escaping HBA.

Responses to HBA have been criticised on a range of levels, including the over-emphasis on forced marriage, the tendency for responses to be reactive (often in response to high profile honour-killing cases) rather than preventive, focusing on the more visible elements of HBA, and the gaps in legislation specifically addressing HBA outside of the forced marriage context (Safelives 2017). In 2015, the HMIC report on the police response to HBA concluded that ‘the police service has some way to go before the public can be fully confident that [HBA] is properly understood by the police and that potential and actual victims are adequately and effectively protected’ (p. 8).

Research on HBA

Criminological research on HBA has primarily focused upon conceptualising and quantifying HBA, and victims’ experiences of justice, although more recent research by Aplin (2019, 2021) has critically examined the policing of HBA. Figures on the incidence of HBA are highly limited, due to a number of measurement issues, including differential definitions, a lack of identification, mislabelling, under-reporting, variation in recording practices within and between agencies, and inaccurate recording; meaning that the true extent of HBA remains unknown (Samad and Eade 2002, Hester et al. 2007, Meetoo and Mirza 2007, Julios 2015, Chantler 2018). The Crime Survey for England and Wales (CSEW) contains generic questions about domestic abuse, but does not specifically cover HBA (Chantler 2018), ruling out the availability of generalisable self-report data. The majority of UK-based data on the incidence of HBA thus derives from police data and varies considerably according to the time frame and number of forces providing data (cf. IKWRO 2015, Khan et al. 2021).¹

Outside of official figures, national charity Karma Nirvana (2021) report that they supported 2584 victims of HBA in 2020/2021, and academic research in England using survey methods estimates between 5000 and 8000 cases of forced marriage were reported in 2008 (Kazimirski et al. 2009); similar work in Scotland has estimated between 129 and 207 cases per year (Chantler et al. 2017). Attempts to count the number of honour killings in the UK are fraught with methodological and definitional difficulties, alongside questions over reliability, but the figure of 12 honour killings per year occurs regularly in the literature (Begikhani et al. 2015). Khan et al.’s (2021) analysis also points to the high level of attrition in HBA cases, with just over 4% of cases resulting in charges being brought. According to the CPS, in 2017/2018, just 145 referrals were made for HBA-related offences, of which 55% were charged (CPS 2018). The most common reason given for unsuccessful prosecution was ‘victim issues’.

These figures highlight not only grave inconsistencies in the prevalence of HBA within and across data sources, but also reveal the high level of attrition in HBA cases. Evidential issues, withdrawal of victim support and unsuccessful prosecutions have been well-documented in the domestic abuse literature and, most recently, specifically relating to coercive control (Brennan and Myhill 2022); however, they have yet to be well-documented in the literature pertaining to HBA. Aplin’s (2019) analysis of case files identified a range of problems with police investigations of HBA (discussed further below) which potentially reduce the number of HBA perpetrators that are recorded as such and ultimately prosecuted.

Responding to HBA

As well as documenting the lack of reliable data on HBA and acknowledging the low number of reported incidents culminating in a charge or prosecution, the 2015 HMIC report highlighted that HBA requires a multiagency response ‘at all stages’ (HMIC 2015, p. 42). In addition to identifying and appropriately flagging cases of HBA, it is also essential for police to execute robust risk assessments and safeguarding measures, which will necessarily require close working with other agencies.

Hester et al. (2007) identified several key motives linked to risk of forced marriage (including preventing unsuitable relationships, protecting family honour or ‘izzat’, achieving financial gain, securing care for a child/adult with special needs), and additional risks have since been highlighted, such as the death of a parent, or if an individual is lesbian, gay, or transgender (HMIC 2014). Research has also highlighted the increased risk of HBA faced by people with learning disabilities (Clawson and Fyson 2017, Patterson et al. 2018), and the relationship between mental illness, self-harm, suicide and experiences of HBA (Chantler et al. 2003, Hussain et al. 2006, Bhui et al. 2007, Hester et al. 2015, Aplin 2018, Khan et al. 2018). According to Safelives’ (2017) data on people accessing support services, 57% of those at risk of HBA had visited their GP in the previous 12 months and 19% had attended A&E, though only 6% of referrals were made by health professionals. Finally, truanting from school and withdrawing from education have been identified as potential indicators for those at risk of HBA/FM (Brandon and Hafez 2008, Britton et al. 2002, Kazimirski et al. 2009, HMIC 2014), as have running away from home or ‘going missing’ (Her Majesty's Government 2014). Running away is believed to be a strategy adopted by many young South Asian women to avoid forced or arranged marriages (Britton et al. 2002, Franks and Rees 2006, Sharp 2010).

Despite this body of evidence surrounding specific risk factors for HBA, the key risk assessment tool for determining risk in cases of HBA in England and Wales is the generic Domestic Abuse, Stalking and Harassment and Honour-Based Violence risk identification, assessment and management tool (DASH), developed collaboratively by the Association of Chief Police Officers (ACPO) alongside academics, and CAADA (now Safelives) (Richards 2009, Payton 2014). This tool was constructed to assess risk in domestic abuse cases, which are overwhelmingly characterised by (ex-)intimate partner abuse. The standard police version consists of a 27-question checklist, which is used to assist officers assign a risk classification of ‘standard’, ‘medium’ or ‘high’ (determined by a combination of identified risk factors from the checklist and professional judgement), which has important implications for subsequent support and safeguarding measures. Evaluations of DASH have documented a significant degree of variation in the approaches to administering DASH across forces, in terms of who completes the form and grades the risk, the extent to which professional judgement plays a role, and how the assessments are processed and actioned (Robinson et al. 2016, Turner et al. 2019).

In addition to general criticisms levied at DASH surrounding its evidential basis, transparency, accuracy and reliability, predictive ability, and practical implementation problems (Thornton 2011, 2017, HMIC 2015, Almond et al. 2017, McManus et al. 2017, Turner et al. 2019); Robinson et al.’s (2016) evaluation of DASH across three forces in England and Wales revealed inconsistency in the application and quality of risk assessments, alongside a poor understanding of coercive control (specifically relating to assessing risk) and a tendency to focus upon physical violence. The research also highlighted the need for more training of officers in domestic abuse and risk assessment, which was recalled as ‘both general and limited in terms of the amount of time spent and content covered’ (p. 20).

The use of DASH as the standard risk assessment tool has also been questioned in terms of its applicability to HBA, particularly because the DASH tool was indisputably designed with intimate partner violence in mind (Medina Ariza et al. 2016). As Payton (2014) points out, risk assessment tools for domestic abuse are informed by a wealth of resources, however, the availability of information for HBA cases is comparatively sparse. Equally, some of the DASH questions, specifically those relating to conflict over separation and child contact, may be wholly irrelevant in cases of HBA, meaning that the assessment erroneously produces a ‘standard’ (low) risk score. In other words, it is likely that the DASH tool asks the wrong kind of questions for identifying or determining risk in HBA cases, and with the limited generic training on domestic abuse and DASH (highlighted by Robinson et al. 2016), officers are ill-equipped to undertake complex risk assessments in cases of HBA. Subsequently, risk assessments can fail to accurately capture the specific threats posed by HBA (Mulvihill et al. 2019).

To overcome this, four DASH questions prompt officers to consider issues of HBA/FM (Q15,² Q20,³ Q21⁴ and Q26⁵). If the answer to any of these questions is ‘yes’, officers are directed towards an annex of ten additional HBA-related questions. However, there is little evidence on how often these supplementary questions are asked, or how effective they are in ascertaining risk. Robinson et al.’s (2016) evaluation of DASH across 3 police forces observed that none of the three research sites operated the additional questions. Woodward’s (2015) theoretical overview of DASH in respect of HBA, concluded that the current tool is unfit for purpose and highlighted the need to evaluate the effectiveness of DASH against bespoke HBA tools. Aplin’s (2019) analysis of DASH in HBA cases revealed a plethora of problems, including high levels of missing data and selective inclusion/withholding of information; the omission of DASH assessments in cases involving children; and, through the use of trivialising language, a tendency to minimise the severity of victims’ experiences and ultimately, levels of risk.

In light of the above, it is perhaps unsurprising that the 2015 HMIC report identified the need for a consistent approach to risk assessment as one of the ‘clearest areas for improvement’ (p. 15) in responding to HBA. Although the College of Policing recently designed and piloted a revised DA risk assessment tool, three of the questions pertinent to HBA (Q20, Q21 and Q26) have been removed, due to a lack of evidence about their effectiveness and the lack of use of the additional ten HBA questions by forces (Wire and Myhill 2018).

Victims’ experiences of criminal justice processes

HBA and forced marriage, like all forms of domestic abuse, are significantly under-reported to the police. Reasons for victims’ reluctance include fear of the consequences, the need to maintain family ‘honour’, close family ties, poor language skills, mistrust of the police, previous poor experience of police involvement, concerns about confidentiality, and reluctance to get their family into trouble (Sharp 2010, Hester et al. 2015, Chantler et al. 2017, Gill et al. 2018). Additionally, some victims do not recognise the abuse or see it as a police matter, and others (especially those from overseas) may be unaware how to report the abuse to the police (Hester et al. 2015, Gill et al. 2018), or choose to report HBA to support agencies rather than the police (Kazimirski et al. 2009).

The high attrition rates in HBA cases have been concerning (Gangoli et al. 2019, Mulvihill et al. 2019), with only a quarter of women interviewed by Mulvihill et al.’s (2019) expressing overall satisfaction with their experience of reporting HBA to the police. Factors contributing to this dissatisfaction included lack of appreciation of the collective nature of HBA, removing the victims from their homes, rather than the perpetrator, without ensuring they had access to alternative accommodation and resources, and considering matching the gender and ethnicity of victims and officers (although this point is contested by Aplin (2021) who argues that an officer from the same community could potentially pose a risk to the victims).

Aplin’s (2019) work made a number of criticisms of the police responses, including claims that police ‘dismantle the crime process’ (Aplin 2019, p. 303) through distortion, alteration, trivialisation through the dilution of the severity of allegations, and data omission. The findings also point to the acceptance of perpetrator accounts and reluctance to prosecute them, alongside falsely portraying victims as reticent to engage. Aplin’s (2019) findings strongly indicate the need for systematic research into the current police response to reports of HBA, and victims’ experiences of justice.

Relatedly, there lies an important question over what may constitute a positive or ‘just’ outcome for victims of HBA. It is well-established that women have historically struggled to attain justice when it comes to gender-based abuse, but there are complex challenges faced by global majority women, for reasons that may include their immigration status, class, caste and/or religious identity (Gangoli et al. 2019). It is essential that experiences of justice should not be assumed to be homogeneous for women victims/survivors of HBA. Broader forms of ‘justice’ maybe more desirable, which, in line with Gangoli et al.’s (2019) findings, can include community recognition of the perpetrator’s wrongdoing, freedom from racism and unconscious bias, and achieving basic human rights such as personal freedom, protection and security. This illustrates that justice, or a positive criminal justice outcome, goes well beyond the confines of prosecution processes.

Clearly, in addition to concerns over police responses, risk assessments, prosecution processes, and safeguarding of individuals who report HBA, there are important questions to be addressed over what might constitute a positive and just outcome for individuals. With research surrounding police responses and victims’ experiences of justice in England and Wales currently in its infancy, this article aims to develop the emerging body of literature by examining criminal justice responses from a policing perspective. The findings presented below focus on the numerous challenges faced by police in responding to HBA and make a number of recommendations as to how police policy and practice might be improved, in order to secure more effective safeguarding and justice for victims.

Methods

The research findings presented below derive from a year-long project driven by the recommendations made in the 2015 HMIC report on the police response to HBA. The collaborative project brought together knowledge and expertise from academic researchers, strategic leads from one English police force (a research partner), and the national HBA support organisation, Karma Nirvana. This collaboration enabled the research to benefit from expertise from the police and Karma Nirvana (both of whom formed part of the steering group and were involved in co-designing the project), as well as assistance with access to data (case file data and interview participants) from both organisations. The key aims were to improve early identification of HBA, highlight both good practice and existing gaps in police provision, and to make recommendations as to how the police response could be enhanced to better safeguard victims and enhance confidence in reporting HBA.

The one-year project received full ethical approval from the University Research Ethics Committee and involved case analysis of three distinct sets of police case files (n = 312 in total), and in-depth interviews with key stakeholders (n = 11) including police, practitioners working with HBA victims/survivors in a support capacity, and victims/survivors of HBA (n = 5). The fieldwork sites were concentrated on two cities in the North of England.⁶ These interviews enabled an exploration of victims’ experiences of the justice process, as well as the perspectives of those encountering HBA in their day-to-day work. Each of the (anonymised) interview transcripts were imported into NVivo. The data was then reviewed with reference to the key research questions, and a series of themes were identified.

The three case file datasets were provided by the police and consisted of a combination of cases already flagged as HBA, and other types of cases where it was possible that HBA may have been missed. Dataset 1 comprised 100 cases that had been flagged as HBA between February–December 2017; Dataset 2 incorporated 120 cases that had been flagged as domestic abuse (but not as HBA/FM) between January–December 2017, which enabled us to identify potential missed opportunities to identify HBA; and Dataset 3 consisted of 92 missing persons cases reported to the police force in 2017. The inclusion of this dataset was in recognition that ‘going missing’ can often be a response to the threat or experience of HBA. The case file datasets were provided in an Excel format and contained information on the source of referral, classification of the occurrence, victim and suspect demographics, offences, whether risk assessments were completed, case outcomes, and comments from the DASH. Where required, the police provided further information on cases from the incident/crime reports and control logs. Information from the datasets was analysed using SPSS.

All interviews were audio recorded and informed consent from each of the participants was secured. Both the interview and case file data were anonymised to remove any identifying details, with pseudonyms being used instead of names.⁷ Interviews were analysed thematically using NVivo.

Research findings

The findings presented below derive from the qualitative analysis of Dataset 1, the 100 cases flagged as HBA, alongside interview data. Three key themes arising from the analysis pertaining to police responses, are discussed below. Firstly, the difficulties faced by police in responding to the collaborative element of HBA; secondly, the problems surrounding risk assessing HBA cases; and finally, the issues surrounding the attainment of a ‘just’ outcome.

Recognising and responding to the collaborative nature of HBA

A key challenge in responding effectively to HBA concerned the often-collaborative nature of HBA. Unlike other forms of domestic abuse, where there is usually only one perpetrator, HBA frequently involves more than one perpetrator, from the victim’s immediate family, extended family and/or the community. As Amber (Police Safeguarding Investigator) highlighted, ‘it's not just about the family. You've got the extended family. You've got the community. So, you know, it's like, “Is it safe for me to live out there?”’. This presents a challenge to the police in terms of identifying and recording all those who have played a role in the abuse, and presents challenges for how cases should be responded to.

In our dataset of 100 cases flagged as HBA, suspects were identified in 87 cases. Of these, only 26% were recorded as involving multiple suspects: fifteen cases involved two suspects, four cases involved three suspects, and two cases involved five suspects. However, there was evidence in the case files that the actual degree of multiple perpetration was much higher. In many of the cases where only one perpetrator was recorded, the case files indicated potential involvement of other perpetrators. For example, ‘Case 7’ concerned a victim who had been shown a message, allegedly from her brother, in which he threatened to assault her because she had left the family home due to suffering HBA perpetrated by her father, who was not recorded as a suspect. In ‘Case 19’, which involved a victim in an unapproved relationship, the notes documented that the victim’s mother and father were both involved, yet only one suspect was recorded. And in ‘Case 66’, in which the victim had a restraining order against her ex-husband, who was the only recorded suspect, it was noted in the case file that the victim’s family were also making threats to disown her, and that the suspect’s uncle had attended her family’s home over the divorce. It is possible that the uncle may have also played a role in threatening the victim but was not recorded as a suspect.

There was evidence of multiple perpetration in some cases where no suspect was identified. In ‘Case 4’, explicit reference was made to both the victim’s mother and sister harassing her. Similarly, the notes for ‘Case 95’ stated that the victim was escorted everywhere by her parents and sisters. This evidence indicates that many of the individuals involved directly or indirectly in the perpetration of HBA are not recorded as such. Male suspects were recorded in 68 out of the 87 cases in which a suspect was identified, whereas females were only recorded in 35 cases. Males were over twice as likely to be recorded as single suspects (n = 47 cases) than co-perpetrators (n = 21 cases); females, on the other hand, were equally likely to be recorded as single suspects (n = 17 cases) and co-perpetrators (n = 18). There is, therefore, some evidence to suggest that female perpetrators might be under-recorded. This issue was also discussed in the interviews:

[W]hat is quite distressing is to know that professionals don’t think ladies can be perpetrators … And like mums and aunties and grandmas can do this stuff as well, it doesn’t always have to be the male of the family. (Nora, Helpline Call Handler)

If the police are not always able to identify multiple perpetrators then many perpetrators, especially women, are not being recorded, investigated, included in risk assessments, or brought to justice. More importantly, the lack of recognition of multiple suspects and the collaborative nature of HBA may have significant implications for the quality of police responses and for victims’ safety. Professionals working with HBA victims and survivors mentioned that police would frequently try to interview or risk assess a victim in the presence of other family members, who might potentially be co-perpetrators or at least supporters/enablers of the abuse. This could negatively affect the level of disclosure from victims and hence the authenticity of the risk assessments, and also place them at risk of further abuse.

The collaborative nature of HBA also needs to be recognised and reflected in police investigative and safeguarding actions. Examples were given by professionals and victims/survivors of HBA in which they were placed at a higher risk by the police unwittingly using interpreters or taxi drivers who were part of the community:

So, if you know anything about taxis and taxi firms in [town], you would know that forget reporting it to the police – if you want to find somebody, go to the Asian taxi driver. They'll find the person before you manage to file the report into the police station. (Saira, HBA Victim)

Saira’s quote illustrates the potential breadth of the network available to those involved in HBA perpetration, making it nearly impossible to assess in advance. As highlighted by Nora, the police need to listen to victims’ concerns about these potential risks and take them into consideration when bringing in third parties:

It’s kind of listening to the victim. Appreciating their risk, taking their concerns as seriously as they possibly can. And kind of knowing who to talk to and who not to talk to, asking the right questions, so the police can ask a victim questions such as, ‘Do you have any, anyone in the police force?’ (Nora, Helpline Call Handler)

The challenges in responding to the collaborative nature of HBA are not recognised or provided for within the domestic abuse framework, which is geared around abuse from current or former partners. In these cases, the roles of the victims’ family and community are crucial to supporting them in escaping abuse, rather than potentially facilitating further abuse. This important distinction points to the need for tailored police responses to reports of HBA, including risk assessment processes. It also highlights the need for more focused and regular training on HBA. Police interviewees consistently highlighted the limited nature of training on HBA, which tended to form part of generic domestic abuse or safeguarding training but was not followed up with regular of HBA-specific training. The brevity of training on HBA was highlighted by Stacey, a Police Call Handler:

I personally think when they do the police training there should be even a couple of hours. Like I say, I mean, we do, we do an awareness session that’s an hour … I’ve spoken to police officers before and they have no idea what it is, literally no idea what honour-based abuse … 

Similarly, Emma, who worked in a police safeguarding unit, explained that although she had specifically attended a 3-day course facilitated by Karma Nirvana, most officers had little or no specific training beyond an e-learning module that would have covered HBA. Consequently, once she received the training, she was perceived as the expert in her unit and therefore dealt with all of the HBA cases: ‘anything that came through that looked remotely honour or forced marriage, came straight to me’.

The lack of bespoke training in HBA across all ranks and roles, and the particular nuances and risks posed to victims as a result of its often collaborative nature that distinguishes it from other forms of DA, is highly problematic, and feeds into the following section, which specifically addresses the issue of risk assessment.

Risk assessment in honour-based abuse cases

Ensuring that all potential perpetrators or enablers of abuse are identified and fed into risk assessments is key to effective safeguarding, and this represents just one of the difficulties for police conducting risk assessments with this complex phenomenon:

Obviously, where there’s indications of honour-based abuse and forced marriage, we would be looking really at asking additional questions, such as, ‘Where are you at the moment? Are you safe? Children to consider? Where is the suspect now, and their wider family, if they’re suspects are well? What impact is it gonna have on you if the police attend your home address?'

[…]

[HBA cases] are a little bit different, in that we should be recognising that potentially the victim could be living amongst suspects, [and the] impact if the community were to see a police car park up outside the home address, that could have, you know, that could escalate the risk of harm basically. (Debra, Police Safeguarding).

The standard tool that the majority of police forces use when risk assessing HBA cases remains the DASH tool. Despite covering all forms of domestic abuse under its now statutory definition, this tool was designed with abuse between (ex)intimate partners in mind. Although the tool does include questions which prompt officers to use an additional HBA questionnaire, in practice there is little evidence of the supplementary questionnaire being used by forces.

It's very basic, and for honour-based abuse and forced marriage it's like two, three questions. It's nothing. We can't determine anything from that. (Nora, Helpline Call Handler)

[I]t’s [DASH] still ineffective because there’s only two or three questions on there around honour-based abuse, so it’s not really gonna give a, a clear picture. (Aisha, Helpline Call Handler)

[B]ecause there’s a little tiny section on it [HBA], on the DASH and some of the questions are completely irrelevant because they’re all partner based. (Jessica, Police Officer)

Our findings revealed the inconsistent use of risk assessment. In almost one-third (31%) of the cases flagged as HBA, there was no evidence that DASH had been completed and there was no risk level documented in the case file. In six of these cases, the notes explicitly stated either that no DASH had been conducted, or that the DASH was incomplete. In some cases, there were reasons for the non-completion of DASH, including five victims aged under 16⁸ and six victims who refused to comply with the police or who denied any incident had taken place.

In at least nine cases, there was no apparent explanation for the lack of risk assessment. The risk assessment in ‘Case 2’ was recorded as ‘none’ despite a 20-year-old male suffering physical, verbal and emotional abuse by both his parents, due to his career choice. Similarly, there were four cases in which a ‘threat to kill’ was recorded in the offences category but there was no accompanying evidence of any risk assessment being carried out. Given that a DASH risk assessment should be completed for every incident involving domestic abuse, it is unclear why such a high proportion of the victims had not been risk assessed, particularly in cases where there appeared to be a significant risk of harm to the victim. These findings underline the need for more consistency in risk assessment processes.

For those cases in which a risk assessment had been completed, six cases were assessed as ‘standard risk’, 39 cases were classified as ‘medium risk’, four were categorised as ‘medium/high risk’, and 22 as ‘high risk’. However, it was not always clear why the level of risk had been ascertained. For example, in ‘Case 40’, which involved a 19-year-old woman whose father had recently discovered that she was in a relationship with a man in Pakistan, the case notes documented that the victim’s father shouted and swore at her and went to hit her with a shoe, before being stopped by his wife. He also threatened to shoot the victim and her boyfriend, which the victim perceived to be a credible threat. The suspect had previous convictions regarding his ex-wife, including assault, and the case notes confirmed that he was controlling of the victim, telling her what kind of employment she could undertake and dictating that she could only work part-time. Despite this, the case was classified as ‘medium risk’. Conversely, ‘Case 76’ concerned a 17-year-old woman whose mother had discovered that she was in a relationship with another woman and had threatened to send her to Pakistan. Although there were no previous convictions or evidence of other verbal or physical abuse, this case was classified as ‘high risk’.

Given the lack of relevance of some of the DASH questions, and the omission of questions that might more accurately predict risk in HBA, there is a strong argument, supported by the evidence within the 100 HBA case files analysed here, that DASH inaccurately assesses risk in HBA cases through simply asking the wrong questions or by not asking them at all. The findings support the need for further research and an evidence base to be developed around the potential of a HBA-specific risk assessment tool, and crucially, for improved training for frontline officers in how to sensitively conduct risk assessment in cases flagged as HBA, considering the often collaborative nature of this form of domestic abuse and how to ask the right kinds of question:

[I]t’s just basically, asking open-ended questions to build a picture about the life that they’ve experienced and the life that they think they will have going ahead. Because that’s the thing about forced marriage, honour-based abuse, you already know what your life is going to be like. (Aisha, Helpline Call Handler)

Asking direct questions and saying, ‘Well, if, if this happened what would be the consequences, what would happen, what do you think would happen? You know, have you had any recent threats or have you had any threats in the past? You know is there anybody else that you’re scared of, afraid of?' (Nisha, IDVA)

Ensuring a ‘successful’ outcome

The final theme arising from our analyses relates to the overall outcome of police intervention. This section considers the sometimes opposing desires for the core policing goal of securing a prosecution and what the victim may regard as a successful outcome, which may involve an alternative form of justice other than a prosecution. The concept of ‘justice’ may, therefore, have very different meanings for the police, who are trained to collect evidence and facilitate convictions of perpetrators, compared to HBA victims, who may aspire to attain safety, freedom, control over their lives, and even basic human rights.

In our HBA case file dataset (n = 100), just six perpetrators were ultimately charged. In 31 cases, the reason for no charge was recorded as due to the victim declining or withdrawing support; in 28 cases there were ‘evidential difficulties’; and in 27 cases, ‘no crime’ was recorded. These extremely high levels of what in police terms constitutes an ‘unsuccessful outcome’ is concerning, and there is undoubtedly the need for further research into why there are such high levels of victim withdrawals, and at what stage. These outcomes also suggest that work needs to be done to collect evidence from those victims who are in support of prosecution, to combat the high level of evidential problems. The professionals we interviewed highlighted difficulties in keeping victims engaged in the prosecution process, and in acquiring sufficient evidence to effectively pursue a prosecution. This was especially true for issues of coercive control, which relates to acts or patterns of acts, such as threats, humiliation and intimidation designed to harm, punish or frighten victims (Home Office 2012):

But sometimes we, when we see a statement, we might have something of the background, but not enough to reach the evidential threshold of coercive controlling behaviour. And especially where there isn’t violence. Because proving a substantial adverse effect on the day-to-day activities, is, you know, it’s, it’s sometimes tricky. (Alice, Crown Prosecution Service)

Alice, who worked for the Crown Prosecution Service (CPS), went on to highlight the importance of the initial contact with a victim – the so-called golden hour – as well as any subsequent support from the police:

So, it’s that golden hour of get the evidence there and then while they’re supportive and assume that they will retract. (Alice, Crown Prosecution Service)

The significant proportion of incidents which result in ‘no crime’ within our dataset indicates that there are important problems in criminal justice processes surrounding and responses to HBA, including victim engagement and support for prosecutions, evidencing HBA, and applying the legislative framework surrounding forced marriage and coercive control in particular. Although the criminal law has been developed to create specific offences relating to forced marriage, in the case files we analysed, there was no use at all of the forced marriage legislation. This was in spite of a high level of evidence surrounding forced marriage within the police reports in 45 out of the 100 cases. In ‘Case 88’, a 30-year-old Asian woman became pregnant two months into a relationship. Her partner asked her to marry him and when she refused, he threatened to tell her family about the pregnancy. They then married in Islamic law. There are close similarities to the first ever prosecution under forced marriage legislation, yet there was no mention of forced marriage and instead the incident was labelled as a ‘domestic incident’.

In support of the broader literature on victims’ perceptions and experiences of justice, our interviews corroborated that many victims do not necessarily want to attain a conviction; rather, safeguarding and freedom are perceived as the most important form of justice:

[I]t's more about safeguarding and then dealing with the criminal offence if they want to prosecute. ‘Cause it is, that's a difficulty that we have. And I think that's where a lot of the policing side is, it's, they don't have an understanding why. We've got a criminal offence here, but why are we not dealing with criminal … ? Why aren't we arresting family members straight away? But we've got to, like, assess it first, risk assess it, and to see … That person might want to go back home, and by locking up nine family members, it's not gonna be … We need to do the safeguarding intervention first, and then look at, and justifying making, arresting people later. (Amber, Police Safeguarding Investigator)

[A] lot of the time it’s about balance, offsetting what’s the purpose of this investigation and balancing it against the potential risk to the victim. And sometimes an investigation won’t progress just because we know that if it does, it’ll put the victim at greater risk. (Jessica, Police Officer)

Discussion: how can police enhance identification of and responses to HBA?

It is important to acknowledge that in both the case file and interview data, there were numerous examples of good practice, particularly by individual police officers who used their personal motivation, dedication and initiative to develop specialised knowledge of HBA and who worked closely with support organisations to help improve police responses to HBA. However, our analyses have highlighted areas where there were missed opportunities to intervene, risk assess and adequately safeguard potential victims of HBA. Furthermore, it has identified the need for HBA-specific policy and practice, especially in assessing and managing risk to victims.

One of the key themes identified was the challenge posed to police by the collaborative nature of HBA, requiring investigating officers to look beyond the immediately obvious perpetrators, often fathers, and recognise the sometimes-subtle involvement of others, who may instigate, collude, execute or enable the abuse from the side lines. Aplin’s (2017) research revealed the significant role played by female relatives, who are often overlooked by police and/or perceived to facilitate HBA only under duress; thus not investigated to the same degree as their male counter-parts. Our case file analysis revealed numerous examples of cases where only one (male, usually the father) suspect was recorded, despite clear evidence of co-perpetration with other (male and female) family members, including mothers, brothers and uncles, who were not investigated as suspects.

The failure to identify and collect evidence surrounding the roles of multiple perpetrators not only means that they are not brought to justice, but also leads to them not being considered in risk assessments and safeguarding plans. As our case file data illustrated, the lack of awareness and understanding of the collaborative nature of HBA might inadvertently enhance the risk to victims if they are interviewed alongside or left with co-perpetrators, or returned home in taxis driven by members of the community. The additional risks posed to HBA victims because of its collective nature has been noted by Aplin (2019), who similarly found evidence of victims being interviewed in front of co-perpetrators, and highlighted the particular risk of placing children with other relatives: ‘trying to safeguard HBA children is far more problematic because of the added complexity of multiple perpetration’ (p. 259).

We argue that the collaborative nature of HBA perpetration needs to play a key role in police training on detecting HBA, identifying all suspects, conducting risk assessments, and implementing safeguarding measures. As our interviewees from one police force revealed, existing police training on HBA is often entirely lacking, brief, or delivered as part of generic domestic abuse or safeguarding training, aside from a minority of officers who actively engage in in-depth training and consequently become perceived as the force expert. This means that the majority of call handlers, frontline officers and specialist officers are not equipped with the knowledge and understanding required to adequately respond to HBA and safeguard victims. These findings build upon Robinson et al.’s (2016) research that highlighted the lack of training of officers on domestic abuse more generally, and DASH risk assessment in particular.

Relatedly, the research findings identified numerous problems with risk assessment processes within the police force we worked with for this project. The challenges in recognising HBA means that many victims are not risk assessed at all, supporting Aplin’s (2019) findings. Even when HBA cases are recognised as such, there is inconsistency in the completion of the DASH tool and within this dataset from one police force, there was no evidence of police using the additional HBA-focused risk assessment questions available. This supports Robinson et al. (2016), who found no evidence of the additional 10 questions being used in any domestic abuse flagged cases they examined across three police forces. As argued earlier, a key problem with the DASH tool is the lack of relevance of some questions, which are geared towards abuse in (ex-)intimate partner relationships, alongside the omission of questions specifically assessing risk in the context of HBA. This may lead to inaccurate risk classifications and consequently, a lack of effective safeguarding. As emphasised by Robinson et al. (2016, p. iii), ‘an evidence-based approach to risk-led policing is needed’. A key recommendation emanating from this research is thus for development of a bespoke risk assessment tool for HBA that is theoretically and empirically driven, drawing upon evidenced-based knowledge surrounding the behaviours, contexts and trajectories of abuse that are specific to HBA and distinctive to other forms of DA. Crucially though, officers across all ranks and roles require training in assessing risk specifically related to HBA, in order to administer whichever risk assessment tools are available and supported by their force.

Finally, the research findings confirmed (in support of other recent literature including Aplin 2019, Gangoli et al. 2019, Mulvihill et al. 2019) that processes and conceptualisations of ‘justice’ need to be re-considered and expanded to meet the needs of HBA victims and hold perpetrators to account. Low prosecution and conviction rates for domestic abuse have been long-since established and most recently, for coercive control (Brennan and Myhill 2022), but little research has documented this in HBA. The incredibly low level of charges brought against perpetrators (7% of cases) and high level of attrition in our dataset point to numerous issues: firstly, that current legislative provisions are insufficiently applied, and that even when HBA, coercive control and forced marriage measures are available, in the majority of cases charges are not brought; secondly, that evidential requirements are inadequate, given that in many cases, charges are not brought due to a lack of appropriate evidence; thirdly, that much work needs to be done in order to enhance levels of victim engagement with the police and prosecution process, so that prosecution procedures are not ceased due to victim withdrawal; and finally, that for many victims, prosecution of the perpetrator is low down, or not even on, on their list of priorities. As Gangoli et al. (2019) report, global majority women face a multitude of barriers to justice, and broader forms of justice including basic human rights such as personal freedom and security may be more important than achieving prosecution.

Our findings indicate that not all victims wish to prosecute members of their family, or bring what they perceive to be further shame on themselves or their families. Instead, what victims seek when disclosing HBA to the police, is help at a time of crisis. This might involve immediate help to escape a dangerous situation, and/or long-term support to gain freedom. Safeguarding measures are therefore paramount to victims, over and above more traditional forms of criminal ‘justice’. Police objectives and targets thus require a degree of recalibration, in order to enable a successful outcome for a victim that does not result in a criminal prosecution.

Conclusion

Our research findings emanate from a project examining the challenges faced by police in one force in responding to HBA, and highlight numerous opportunities to improve the safety and justice of those who come to the attention of the police as potential victims of HBA. HBA is an extremely complex and nuanced form of violence, presenting many challenges to frontline officers, specially trained officers, and those responsible for police strategy. The issues raised and discussed in this paper only scratch the surface and omit, for example, difficulties arising due to the difficulties outlined at the beginning of the paper in terms of defining and conceptualising HBA, and the dominant cultural and gendered frameworks within which HBA is currently understood. However, our findings contribute to an emerging body of evidence around the problem of policing HBA, and make a number of recommendations for how police can be better supported to more adequately safeguard and achieve justice for victims.

The three areas included within this paper highlight the complexities posed by the collaborative nature of this form of abuse, the problems with existing risk assessment measures, and the problem with dominant forms of justice. The key recommendations first and foremost highlight the need for structural change. Despite good intentions, motivation and dedication of key HBA ‘pioneers’ within the force we worked with, there remains a lack of robust training across in HBA, and addressing this is crucial. Despite some excellent understandings by some specialists of the specific risks posed to HBA victims, particularly surrounding issues posed by the often collaborative perpetration of and potential community-wide support for the abuse, many officers responding to HBA are insufficiently trained and may inadvertently enhance a victim’s level of risk. Relatedly, without rigorous training on assessing risk and in the absence of a bespoke, evidence-based HBA risk assessment tool that recognises the specific risks, contexts and trajectories that are specific to HBA, effective risk assessment will remain difficult to achieve. Finally, until structural change enables police to engage with and pursue alternative forms of justice, the current focus on collecting hard evidence and seeking prosecution will remain inadequate for victims, who require sensitivity and safeguarding above all else. HBA has been on the policy and legislative agenda for several decades now, however, our findings from within one police force reiterate the concerns expressed in the HMIC report from 2015 and suggest that there is still some way to go until police practice policy and legislation sufficiently meet the complex needs of victims.

Ethical approval

The research project underpinning this article received ethical approval from the University Research Ethics Committee at the University of Manchester, approval number: 2018-4946-7328.

Acknowledgements

The authors would like to thank and acknowledge the Research Assistant in this project, who has asked to remain anonymous and not appear in the list of authors, despite their significant contribution to this article. This is due to a conflict of interest with their current employment.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Anonymised transcripts were made available to N8PRP for archiving in its Leeds University repository.

Additional information

Funding

This work was supported by the N8 Police Research Partnership [grant number BH141990-012].

Notes

1 IKWRO (2015) data from Freedom of Information (FOI) requests reports that between 2010 and 2014, 11,744 HBA incidents were recorded by 29 of the 52 UK police forces that responded, averaging 2348 per year. Khan et al. (2021) report that in 2018 and 2019 there were respectively 1069 and 1106 HBA incidents recorded by police across the 15 forces who responded to FOI requests. Official statistics report that in 2020/2021 there were 2725 honour-related offences recorded by all police forces across England and Wales (excluding Greater Manchester Police) (Home Office 2021).

2 15. Does (…) try to control everything you do and/or are they excessively jealous? (In terms of relationships, who you see, being ‘policed at home’, telling you what to wear for example. Consider honour based violence and stalking and specify the behaviour).

3 20. Is there any other person that has threatened you or that you are afraid of? (If yes, consider extended family if honour based violence. Please specify who. Ask 10 additional HBV questions*).

4 21. Do you know if (…) has hurt anyone else? (children/siblings/elderly relative/stranger, for example. Consider HBV).

5 26. Has (…) ever breached bail/an injunction and/or any agreement for when they can see you and/or the children? (List of options includes forced marriage).

6 The locations of the fieldwork sites were largely dictated by the geographical locations of the research partners and the constraints of the grant funding provided.

7 See Fox et al. (2019) for more information on the project.

8 The domestic abuse definition and corresponding requirement to complete a risk assessment only applies those aged 16 and over.