It gives me great pleasure to post this article. Cat is graduating in a couple of weeks from CCCU with a first class honours degree in Policing. Her dissertation was exceptional and as a result she was encouraged by staff and academic colleagues to publish it in a shortened version – this is the result. Well done Cat we are very proud of you here at CCCU.
Published in Police Professional
Is DASH Fit To Assess Honour-Based Violence Risk?
The ‘Domestic Abuse, Stalking, Harassment and Honour-Based Violence’ (DASH) risk assessment checklist (ACPO, 2008) (see http://www.dashriskchecklist.co.uk; Richards, 2009, 2015) is a risk assessment tool for use by frontline police officers throughout the UK responding to conventional domestic violence (DV) situations, as well as those involving Honour-Based Violence (HBV), stalking and harassment.
HBV has been linked with DV as the two forms of familial violence share many characteristics (Payton, 2014; Keyhani, 2013; Gill, 2004, 2008b; Baker et al, 1999; Belur, 2008; Burman and Chantler, 2005; Podaná, 2010; FCO and Home Office, 2014b; ACPO, 2008; Meyer, 2011; Baker et al, 1999; Domestic Abuse: Caught on Camera – Panorama, 2014; Dogan, 2014). In some ways, this provides useful contextualisation which can help the police deal with HBV – a relatively unfamiliar crime. In other ways, this can be counter-productive as, firstly, the policing of DV is not without its own issues (Home Affairs Committee, 2008; Hoyle, 2008; Dichter and Gelles, 2012; Grauwiler, 2008; Meyer, 2011; Felson and Pare, 2008; HMIC, 2014; The Police Foundation, 2014), and, secondly, as well as similarities, HBV has significant differences from DV which require alternative policing techniques (ACPO, 2008; Papp, 2014; Araji, 2000; Gill, 2004, 2009; Belfrage et al, 2012; Roberts et al, 2014; Dustin and Phillips, 2008; Salter, 2013; Cooney, 2014; Brandon and Hafez, 2008; Baker et al, 1999; Dogan, 2014; Siddiqui, 2005; FCO and Home Office, 2014b; Pertile, 2010).
Although DASH was launched in 2009 (Richards, 2015; Robinson, 2010; The Police Foundation, 2014), there is no evidence that it has been empirically validated (Wheller and Wire, 2014; Pease et al, 2014). This is concerning as it can be argued that an HBV-competent risk assessment tool provides the foundation for effectively identifying and responding to HBV cases (The Police Foundation, 2014; Home Affairs Committee, 2008). If DASH, the only such candidate officially used by UK police, is not fit-for-purpose in this regard, this could potentially put HBV victims in grave danger and undermine the public’s confidence in the police (Kropp, 2004, 2008; Nicholls et al, 2013; Robinson, 2010; Bennett Cattaneo and Goodman, 2007; Kercher et al, 2010; Walklate and Mythen, 2011; Roberts et al, 2014; Home Affairs Committee, 2008; Grant and Rowe, 2011). Having identified this gap, this article critically explores DASH to conclude whether it is fit to assess HBV risk.
DASH consists of a main checklist of twenty-seven questions relating to risk factors for DV (Richards, 2009; Robinson, 2010). When responding to a DV or HBV incident/complaint, all twenty-seven questions must be asked of the victim. If question twenty, “Is there any other person that has threatened you or that you are afraid of?” (Richards, 2009, p. 4), is answered positively, the practitioner is guided towards asking a further ten questions related to HBV. The answers to both sets of questions, along with the assessing officer’s professional judgement, are combined to achieve an overall risk score of ‘Standard’, ‘Medium’, or ‘High’ (Richards, 2009, p. 4). This score informs officers as to the appropriate risk management response.
With this structure in mind, it is sensible to explore how sufficiently DASH accommodates HBV-specific risk factors within the ten HBV questions. Then, as the main ‘generic’ checklist is also applied, it is prudent to examine the appropriateness of combining HBV and DV in the one risk assessment tool.
The Accommodation of HBV-specific Risk Factors
The ten HBV questions featured in DASH are all, arguably, valid (Gill, 2004, 2005a, 2006, 2008a,b, 2009, 2010; Gill et al, 2012; Atherley, 2007; Roberts et al, 2014; Dean, 2008; FCO and Home Office, 2014b; Knudsen, 2003; The Job, 2006; Reddy, 2008; IKWRO, 2014; Dustin and Phillips, 2008; Belfrage et al, 2012; CIMEL and INTERIGHTS, 2001; Siddiqui, 2005; ACPO, 2008; CRASAC, 2014; Pertile, 2010; Keyhani, 2013; Salter, 2013; Amnesty International, undated, cited in Pertile, 2008, p. 34; Payton, 2014; Krantz and Garcia-Moreno, 2005; Yurdakul and Korteweg, 2013; Papp, 2014; Dogan, 2014; Wilson, 2007). There are several questions which are clearly aligned to potential HBV presence, for example, victims can be withdrawn from education and forced into marriage (Gill, 2005a, 2006; CRASAC, 2014; Awwad, 2004, and Goodwin, 1995, cited in Atherley, 2007, p. 27; Roberts et al, 2014), and HBV questions one, four and five address these aspects. However, some of the other questions are non-HBV-specific and need modifying or removing, for example “Isolation – is the victim very isolated?” (Richards, 2009, p. 6). This question needs to elaborate whether the victim feels isolated from people outside of the family in order to make it indicative of HBV. As victims are more likely to be closely supervised by family – a collective – this will not necessarily, depending on the victim’s subjective interpretation, invoke a feeling of isolation (IKWRO, 2014; The Job, 2006; Gill, 2005a, 2006, 2008b; Roberts et al, 2014; CRASAC, 2014; CIMEL and INTERIGHTS, 2001; ACPO, 2008; Atherley, 2007; Singh, 2015). Therefore, if the victim answers negatively to isolation, when, actually, she is isolated from her friends, school or work, this false negative may contaminate DASH’s predictive validity. Question three regarding house arrest addresses intensive family supervision, and, therefore, the question around isolation could be removed, particularly as it already appears in the main checklist, in order to avoid this potential misinterpretation.
Question seven, “A pre-marital relationship or extra marital affairs – is the victim believed to be in a relationship that is not approved of?” (Richards, 2009, p. 6) is, again, generic to HBV and DV (Hoyle, 2008; Dichter and Gelles, 2012; Connor-Smith et al, 2011). It could be moved to the main checklist where it currently does not feature. This would leave room for a modified version of this question that is more HBV-specific. Female sexual impropriety is the most fundamental cause of honour-loss, which can largely be suspected through association, no matter how slight or platonic, with a male unapproved of by the family (Gill, 2003, 2005a, 2008b; Keyhani, 2013; Atherley, 2007; Onal, 2008; Reddy, 2008; Hussein, 2010, cited in Gill et al, 2012, p. 81; Siddiqui, 2005; Brandon and Hafez, 2008; Knudsen, 2003; USA: Department of State, 2006, cited in Nasrullah et al, 2009, p. 1; Meetoo and Mirza, 2007; Faqir, 2001, and Hoyek et al, 2005, cited in Dogan, 2013, p. 492; Shah Kakakhel, 2004; Roberts et al, 2014; Pertile, 2010; Belfrage et al, 2012). As anything other than inter-tribal/clan/familial/religious marriage is usually met with disapproval (Samad and Eade, 2002, cited in Atherley, 2007, p. 31; Singh, 2015; Gill et al, 2012; Knudsen, 2003; Roberts et al, 2014), a modified version of question seven such as, “Have you been accused/suspected of associating with anyone from an ‘inferior’/‘unacceptable’ racial/ethnic/caste/class/religious background?” (Singh, 2015) would be a more suitable replacement.
Question two asks if there is evidence of victim self-harm. Self-harm, indeed, features as a risk factor for HBV within academic literature (Atherley, 2007; Roberts et al, 2014). However, it is also, potentially, a symptom of DV victims, as, although the literature does not specifically label self-harm as a DV risk factor, there is evidence that DV victims can suffer diminishing mental health (Campbell, 2002, cited in Messing and Thaller, 2013, p. 1538), of which self-harm could be an expression (Mind, 2013). What makes self-harm unique from an HBV perspective is that victims can be encouraged or instructed to harm themselves by their perpetrators to achieve redemption for costing the family their honour (Singh, 2015; Onal, 2008). This can include forced suicide, which is a convenient and popular honour killing disguise (BBC, 2003, cited in Atherley, 2007, p. 28; IKWRO, 2014; Siddiqui, 2005; ACPO, 2008). Victims may feel pressured to comply for reasons such as fear, guilt, and dependence (ACPO, 2008; Salter, 2013; Dean, 2008; CRASAC, 2014; Siddiqui, 2005; FCO and Home Office, 2014b; Belur, 2008; Burman and Chantler, 2005; Roberts et al, 2014; Payton, 2014; Murray, 2008). As diminished mental health is covered in the main checklist (Richards, 2009, p. 3), this question around self-harm can afford to be modified to make it more specific to self-harm committed under duress.
Question eight asks about separation from the victim’s intimate partner and question ten deals with threats to hurt or kill. Both of these questions, although valid, already feature in the main checklist and are generic to both HBV and DV (Hoyle, 2008; Dichter and Gelles, 2012; Walklate and Mythen, 2011; Kropp, 2004; Robinson, 2010; Connor-Smith et al, 2011). Therefore, these questions could be removed to avoid duplication and make room for other more HBV-specific questions related to typical characteristics and behaviours, such as “Has your husband’s family complained about you to your own family?”; “Are you or does anyone suspect you to be homosexual and what is your family’s views of homosexuality within the family?”; “Has any perceived sexual indiscretion given rise to this incident or has sex been used for punishment?”; “Is any abuse directed at you because you are female or because you have only given birth to females?”; “Are you aware of any dowry or bride-price concerns within your family that relate to you?”; and “Is your community aware of your problems with your family?” (Singh, 2015; Atherley, 2007; The Job, 2006; Payton, 2014; Gill, 2004; IKWRO, 2014; Siddiqui, 2005; ACPO, 2008; Korteweg, 2012, and Sev’er and Yurdakul, 2001, cited in Yurdakul and Korteweg, 2013, p. 205; Roberts et al, 2014; Amnesty International, undated, cited in Pertile, 2008, p. 34; Gill, 2010; Abu-Odeh, 1996, and Werbner, 2007, cited in Gill, 2009, p. 476; CIMEL and INTERIGHTS, 2001; Knudsen, 2003; Chesler, 2009, cited in Belfrage et al, 2012, p. 20; Papp, 2014; Tripathi and Yadav, 2004; Ginat, 1979, and Kressel, 1981, cited in Baker et al, 1999, p. 171). Also, what is missing which is very important is asking the victim “Who is the head of the family and who follows their beliefs?” (Singh, 2015; Onal, 2008). This should be accompanied by asking the victim to draw a family tree, indicating the power hierarchy. This will assist officers in estimating the scale of collectivity as well as providing a list of key suspects. (Singh, 2015; Payton, 2014; FCO and Home Office, 2014a).
The above reveals that not all of these questions relate to HBV-specific risk factors, characteristics or behaviours. Out of those that do, some relevance is diluted by over-generalised wording. This may cause the victim to misinterpret the questions, thus, provide distorted answers, which may adversely influence DASH’s predictive validity. Therefore, DASH’s alignment to HBV-specific risk factors would appear to be insufficient.
The Appropriateness of Combining HBV and DV in One Risk Assessment Tool
The advantages of the combined tool are three-fold. Firstly, HBV shares many of the same risk factors as DV, for example, victim fear; victim attempt(s) to separate from her intimate partner; sexual abuse; and reaching for outside help (Robinson, 2010; Hoyle, 2008; Dichter and Gelles, 2012; Walklate and Mythen, 2011; Kropp, 2004; Connor-Smith et al, 2011; Dugan et al, 2004, Bernard and Bernard, 1983, and Campbell, 1992, cited in Kercher et al, 2010, p. 7; Atherley, 2007; Roberts et al, 2014; Payton, 2014; Gill, 2004, 2005b; Pertile, 2010; The Job, 2006; Salter, 2013; Belfrage et al, 2012; CIMEL and INTERIGHTS, 2001; ACPO, 2008; IKWRO, 2014; Siddiqui, 2005). Having one risk assessment tool which covers both DV and HBV reduces duplication and simplifies the process (ACPO, 2008; Home Affairs Committee, 2008). This avoids the temptation, if there were two separate tools, for officers to carry only one because it is less time consuming/less to carry, as was the case in a recent study on Swedish police officers who favoured the use of a DV risk assessment tool (B-SAFER) over its HBV counterpart (PATRIARCH) (Belfrage et al, 2012) when responding to all familial disputes.
Secondly, because HBV is relatively unfamiliar to the majority of officers, it may not be obvious to officers when answering a domestic dispute that HBV is present (Payton, 2014). And thirdly, having a combined DV and HBV risk assessment tool helps officers respond to familial disputes more appropriately without being influenced by cultural/ethnic/religious stereotypes (Payton, 2014).
There is, however, a disadvantage with the combined tool. As HBV is a sub-section of DASH, consideration must be given to the questions included in the main checklist. These are mandatory and, along with the ten HBV questions, influence DASH’s predictive validity regarding HBV risk. They should, therefore, be sufficiently relevant and understood by the HBV victim to avoid misinterpretation and to maintain their confidence in the police’s ability to help.
The whole family, and sometimes wider associates, usually collude in HBV perpetration, and this is a major distinction between HBV and DV (Roberts et al, 2014; Dustin and Phillips, 2008; Salter, 2013; Gill, 2009; ACPO, 2008; Chesler, 2009, and Pimentel et al, 2005, cited in Cooney, 2014, p. 409). Payton (2013), therefore, argues for positioning the HBV trigger question, currently at number twenty within the main checklist, nearer the beginning to establish early on whether collective perpetration is present. In this way, subsequent questions can be adjusted by the assessor to reflect collective perpetration to keep them relevant to the HBV victim.
HBV situations can consist of long periods of coercive control, but can very quickly escalate into extreme/fatal violence depending on the extent to which the family perceives an honour transgression has become public knowledge (Atherley, 2007; Ginat, 1979, and Kressel, 1981, cited in Baker et al, 1999, p. 171; Glazer and Abu Ras, 1994, and VanEck, 2003, cited in Payton, 2014, p. 2865-6; IKWRO, 2014). Therefore, it is possible that there is no physical harm until the victim is killed (Onal, 2008; McVeigh, 2007). By having the first question as, “Has the current incident resulted in injury?” (Richards, 2009, p. 3), this may, early on, set a distorted perception that DASH equates absence of physical injury to low risk (Singh, 2015). An HBV victim who has not yet been physically assaulted may accept this perception, erroneously gauging their risk as lower than it is (Gill, 2004; Campbell, 2004; CRASAC, 2014), or they may be very aware of their accurate risk level, but become sceptical of DASH’s ability to reach the same conclusion (Hoyle, 2008; Brewer et al, 2007, cited in Dichter and Gelles, 2012, p. 45; Bennett Cattaneo and Goodman, 2007; Walklate and Mythen, 2011; Campbell, 2004; Connor-Smith et al, 2011; Kropp, 2004; Robinson, 2010; Bowen, 2011). This may lead to the victim’s loss of confidence in the police’s ability to help and they may disengage or become uncooperative (Singh, 2015; Payton, 2014; CRASAC, 2014; Belur, 2008). It might be prudent to amend this question to read, “Has the current incident resulted in injury, or threat of injury or threat to life?” (Singh, 2015) and may be better positioned further down the main checklist after “Are you very frightened?” and “What are you afraid of?” (Richards, 2009, p. 3). This will provide a more gradual lead-in to the question of injury whilst reducing the likelihood that it is misinterpreted.
Some of the wording of questions within the main checklist could be amended in line with HBV terminology so that they are more relatable to HBV victims. For example, “Does (…) do or say things of a sexual nature that makes you feel bad…?” (Richards, 2009, p. 4) could be changed to “Does (…) do or say things of a sexual nature that makes you feel bad or ashamed…?” (Singh, 2015; Gill, 2004; Payton, 2014). HBV victims can relate to the word ‘shame’ as they understand this to be the consequence of losing honour (Gill, 2003, 2008b; Brandon and Hafez, 2008).
There are other examples within the main checklist where questions are lacking in relevance to HBV, such as perpetrator jealousy of victim, perpetrator abuse of drugs and/or alcohol, and perpetrator abuse of pets (Roberts et al, 2014; Robinson, 2010; Singh, 2015; Klemperer, 2015). Removing them would be appropriate for HBV, but not necessarily for DV.
Combining HBV and DV in the one assessment tool promotes efficiency by reducing duplication and helps to detect HBV where it may not be obvious, whilst, conversely, preventing an inaccurate diagnosis of HBV based on stereotyping. However, as some questions within the main checklist are not positioned or worded appropriately in order to make them sufficiently relevant or understood by HBV victims, the overall risk prediction may be incorrect if influenced by answers from misinterpreted questions in this checklist. This is a significant concern, which is not mitigated by the advantages. Therefore, it might be appropriate to combine HBV and DV in the one risk assessment tool, but the current arrangement in DASH impedes its ability to identify and accurately assess HBV risk.
In summary, DASH is comprehensive in its content and the majority of questions are, arguably, valid. However, given the findings presented above, this article concludes that DASH is currently not fit to assess HBV risk.
With careful consideration towards modification and re-positioning of certain questions in the main checklist and HBV sub-section, DASH’s predictive validity could be improved in order to make it fit-for-purpose within an HBV context. However, a condition of this recommendation is that any amended version is swiftly followed with comprehensive empirical testing (HMIC, 2014; Wheller and Wire, 2014; Pease et al, 2014), which should include analysing the effectiveness of keeping the tool combined with DV versus using a bespoke HBV tool.
Additionally, this article recommends that further empirical testing should be conducted to establish a definitive range of HBV risk factors which will help inform this process whilst adding to the very few items that already exist in academic literature (Payton, 2014; Sartin et al, 2006; Gill, 2003, 2008b, 2010; Salter, 2013; Smith, 1989, cited in Belur, 2008, p. 429; Dogan, 2014).
Provision of an HBV-competent risk assessment tool is a very important first step in getting the policing response to HBV right. More widely, there is room for other initiatives that would help raise awareness and contribute to an overall improved policing response to HBV. This article recommends the following such initiatives:
Research to develop the most appropriate response and investigative techniques for HBV (CPS, 2010, cited in Salter, 2013, p. 108; Gill, 2008b; Dogan, 2014). An effective response technique is offered by DS Palbinder Singh, of the Greenwich Community Safety Unit, who this month won a Metropolitan Police Service Domestic Abuse Achievement Award for ‘Outstanding Individual Contribution to Victim Care’ during HBV investigations. Based on his extensive experience in HBV cases, he suggests that if an officer is unable to speak to a victim without family/community supervision or away from the family home, they should consider placing the victim under pseudo arrest, using the guise of ‘wasting police time’, in order to transfer them to an environment free of coercion where they can open up in confidence about their concerns. This also creates an opportunity to connect the victim with agencies who can arrange emergency housing and other means of support so that the victim need not return to the family home.
Training for police officers and frontline police staff, and those who influence police decisions, i.e. Crown Prosecution Service (CPS) lawyers, to understand these techniques, recognise HBV and learn how to deal sensitively with HBV victims (Gill, 2003, 2005b, 2008b; Gill et al, 2012; Keyhani, 2013; Tripathi and Yadav, 2004; Ammar et al, 2014; Home Affairs Committee, 2008; IKWRO, 2014). This includes ensuring guidance documents/‘toolkits’ offer bespoke response advice. DS Singh suggests that such advice could include considering resisting the arrest of suspects until sufficient evidence has been garnered and presented to the CPS. This would require the participation of lawyers who specialise in HBV cases, akin to those CPS lawyers who specialise in crimes such as rape, who will understand the need to charge suspects using a ‘threshold test’. The rationale for this is that in an HBV scenario the premature arrest of a suspect, in the absence of tangible evidence, can fast-track the victim’s risk of harm to severe as well as undermine the police case. This is because the arrested suspect, once bailed, is likely to recruit the whole community to force the victim to withdraw from police action, as well as subjecting them to further control, isolation and abuse. However, delaying the arrest does not work in every HBV scenario, for example, where the suspects are aware of police interest and/or there is an imminent risk to the victim/witnesses, warns DS Singh, so training should include how to discern this on a case by case basis.
The establishment of a specialist HBV policing unit, policing secondary investigations into HBV. DS Singh suggests the need for such a unit, which should comprise pro-active and reactive investigation methods and be staffed by dedicated and knowledgeable detectives/personnel. This would prevent the need to fracture HBV investigations – which can be multi-faceted – between current specialist units, which dilutes the effectiveness of the police response.
The above recommendations will help reduce the likelihood of investigations being discontinued – a problem illustrated by Greater Manchester Police in last month’s Channel 4 documentary, ‘Forced Marriage Cops’ (Forced Marriage Cops, 2015).
The discussion on effective UK policing of HBV is very relevant within today’s climate as Her Majesty’s Inspectorate of Constabulary is currently undertaking an inspection of this matter (HMIC, 2015). This follows its identification that HBV still appears to be overlooked by police and, to date, forces have never been independently scrutinised by any inspectorate on this matter.
It will be interesting to note the results of this inspection, when published, to establish whether DASH was scrutinised and whether the conclusion complements or contrasts with this article. Whatever the outcome, it is clear that DASH, and, indeed, the wider topic of policing HBV, is under-debated within academic literature, and not understood fully in policing circles, so there is ample scope for further research which would hopefully lead to a more effective policing response (Wheller and Wire, 2014; Pease et al, 2014; Salter, 2013; Singh, 2015; Gill, 2008b, 2010; Smith, 1989, cited in Belur, 2008, p. 429; Dogan, 2014).