Fear or Failure

FEAR OR FAILURE? Why victims of domestic violence retract from the criminal justice process Lillian Artz* [email protected] In 2008/9 MOSAIC,1 with the assistance of the Gender, Health & Justice Research Unit (UCT), embarked on research that sought to identify the factors that contribute to domestic violence victims withdrawing from the legal process before they finalise protection orders (POs) applied for under the Domestic Violence Act (DVA).2 This study was based on the 2008 work of this author who, in partnership with MOSAIC, interviewed 365 domestic violence victims in the Western Cape about their engagement with and retraction from the criminal justice process.3 The second tier of this project – reported on here – emerged with more focused interview schedules and the addition of eight jurisdictions from which the sample was drawn. The findings from this study were extensive and pointed to a range of personal, systemic and structural reasons why Domestic Violence Act [DVA] applicants disengage from the criminal justice process. This article will limit its focus on three areas that are relevant to the decision by survivors to withdraw their applications for protection orders: the history and severity of violence, deadly threats, and key findings relating specifically to experiences of DVA applicants with the courts. The Domestic Violence Act (DVA) (No. 116 of 1998) was promulgated in 1999 with the aim of creating a civil remedy that was accessible, affordable and ensured that ‘victims of domestic violence received the maximum protection from domestic abuse that the law could provide’.4 Under the DVA, victims of domestic violence may apply for a Protection Order (PO), which aims at preventing future acts of violence.5 It involves a two step process. The victim of domestic violence applies for an interim protection order (IPO)6 which is granted if the court is satisfied that there is prima facie evidence that the ‘respondent’ (the alleged abuser) has committed an act of domestic violence and that the applicant (the victim) would suffer undue hardship if a protection order was not issued immediately. Where urgency can be shown, an IPO is issued.7 The second part of the process involves finalising the order. Once the IPO is served on the respondent, the applicant and the respondent are required to return to court on a certain date – referred to as the ‘return date’ – for a hearing during which the respondent is afforded the opportunity to present to the court reasons why the protection order should not be finalised. If the court is satisfied, on a balance of probabilities, that the respondent has committed or is committing an act of domestic violence, the protection order can be finalised (or varied in some way).8 It is in the period between the application for an IPO and the finalisation of the protection order that cases fall out of the system. * Associate Professor, Gender, Health & Justice Research Unit, University of Cape Town. This article is based on the findings from L Artz, D Jefthas & Z Majeit, Reluctance, Retaliation and Repudiation: The Attrition of Domestic Violence Cases in Eight Magisterial Districts, GHJRU, University of Cape Town: South Africa, 2011. DOMESTIC VIOLENCE ATTRITION RESEARCH SA Crime Quarterly no 37 • September 2011 This study of domestic violence attrition was sparked by MOSAIC, an organisation that provides 3 court support services to victims of domestic violence who apply for protection orders. MOSAIC estimated that almost half of their clients were not returning to court to finalise their protection orders, and embarked on research to establish the reasons why victims of domestic violence who apply for POs do not return to finalise them. Attrition research shows us the different points in the criminal justice system where cases fall out, and, hopefully, the reasons for cases falling out. Without this kind of research, we can only speculate about the decrease in the number of cases from the time they are reported to the police, to when they go to court. We can use attrition research to establish whether the reasons for cases dropping out relate to the complainant (personal reasons) or relate to the performance or decisions of the criminal justice system (systemic reasons). Studies that have reported on the rates of attrition in domestic violence cases consistently find that out of the ‘thousands of women’ who initiate the process of obtaining protection orders in the US, less than half return to court to obtain final orders.9 It is critical to establish whether the fallout or ‘attrition’ of cases is a result of the failure of criminal justice to implement domestic violence legislation, or of a seemingly unrelenting cycle of violence.10 This research has shown that it is a combination of both. Despite progressive legislation such as the Domestic Violence Act, the inconsistent and faulty application of the law by the police and other criminal justice agents results in limiting the effect of the legislation, and discourages protection order applicants from continuing the legal process. On the other hand, victims of domestic violence are often caught in intractable personal situations that make it almost impossible to take effective legal action. METHODOLOGY Between 2008 and 2009, 503 DVA applicants11 were interviewed about why they applied for protection orders, but did not return to court to have these orders finalised.12 Their application forms for protection orders were also reviewed. They were interviewed at the following research sites: 4 Table 1: Research sites Research sites/courts Number of interviews % of total interviews Bellville Bishop Lavis Cape Town Khayelitsha Paarl Philippi Wellington Wynberg 42 20 78 87 56 88 13 119 8,3% 4,0% 15,5% 17,3% 11,1% 17,5% 2,6% 23,7% TOTAL 503 100% The sampling was purposive, specifically targeting complainants who did not return to court to finalise their protection orders. The research took place over a six-month period. The only requirements for participation in the study were that DVA applicants (a) consented to being in the study at their first appearance (application) at the court; (b) were on the court roll on the return date; and (c) did not appear in court on the return date during the period of the research. DVA applicants were asked to participate at their first court appearance, which is at the date of application for a protection order. It was only after applicants had been assisted with applying for a protection order that the prospect of participating in the study was introduced and explained in detail.13 On any given day there may be between 20 and 50 applications for protection orders and hearings for domestic violence cases at each court, and up to half of the hearings may involve ‘non-returns’.14 Sometimes, although not exactly procedurally correct, these cases are struck off the court roll. The researchers involved in the project noted who was struck off the roll and checked these names against the list of applicants who consented to participate in the study when they came to court to apply for protection orders. These applicants were then phoned for an interview. Afrikaans and Xhosa interviews were conducted by MOSAIC’s auxiliary social workers, who were extensively trained on how to conduct the interviews, including how to manage situations where the alleged perpetrator was in the vicinity when the researchers phoned for the interview. Institute for Security Studies The non-return interview schedule was a semistructured interview schedule specifically designed to interview research respondents telephonically. The interview schedule was divided into eight sections: 1. 2. 3. 4. 5. 6. 7. 8. Introduction to the project Re-establishing consent Nature of application for protection Reactions by abusers when interim protection order was served Experiences with police and the reporting process Experiences with the courts Why applicants did not return to court on the return date to finalise protection orders, and Additional narratives specifying reasons for non-returns. The findings were analysed through a post-coding process as well as a process of thematic content analysis of the narratives from the interviews. The findings were analysed based on the gender of the applicants and the nature of the relationship between the applicants and their abusers, and specifically whether the PO was sought for protection from an intimate partner or a family member.15 The four key categories for analysis were: • • • • Female intimate partner [FIP] – female applicants who applied for a PO against their intimate partners Male intimate partner [MIP] – male applicants who applied for a PO against their intimate partners Female family member [FFAM] – female applicants who applied for a PO against a family member, and Male family member [MFAM] – male applicants who applied for a PO against a family member. RESEARCH FINDINGS Demographic profile of the research respondents Demographic information on the research respondents was primarily captured in, and extracted from, the intake forms. Where information on the intake form was absent or unclear, the relevant information was taken from the interview schedules and/or affidavits attached to the consent forms. As Table 2 sets out, the majority of applicants in the study were female (84,5%; n=425), with a significantly smaller number of male applicants (15,5%; n=78) completing the sample. Approximately threequarters (74,7%; n=376) of the applications included in the study were brought against intimate partners, with the remaining 127 applications (25,3%) being brought against family members. Women applying for POs against an intimate partner made up 64,4% (n=324) of the applications, with the next biggest percentage, 20,1%, comprising women wanting protection from family members (n=101). Applications by men against intimate partners totalled 10,3% (n=52) of all research respondents, with men seeking protection from family members in only around five per cent of the cases (n=26).17 The majority of applicants in the study, 65% (n=327), were between the ages of 20-39. Of all Table 2: Gender of applicants and nature of relationships Applicants – sex category Number % 324 64,4% 101 20,1% 425 84,5% 52 10,3% 26 5,2% Total male applicants 78 15,5% TOTAL 503 100% Female intimate partner applicants Female family member applicants Total female applicants This approach was adopted on the basis of overwhelming empirical evidence from both local and international research showing that women tend to be the victims of domestic violence, and the strong association between femicide and a history of domestic violence.16 SA Crime Quarterly no 37 • September 2011 Male intimate partner applicants Male family member applicants 5 the female applicants (n=425), 67,5% (n=287) were in this 20-39 age group, with a further 26,6% of women (n=113) between the ages of 40-59. Of all the male applicants, just over half (51%; n=40) were between the ages of 20-39 and 38% (n=30) between the ages of 40-59 years old. of violence within the relationship results in a ‘reporting incident’. This incident may not be reflective of the historical expressions of violence within the domestic relationship, but is experienced by the victim as something which has historically led to more serious violence. Nature of abuse Victims of domestic violence often suffer a range of abuses, compounding in severity or frequency over time and as the relationship begins to disintegrate. We found that, on average, DVA applicants have known their partners for nine years, experiencing abuse for at least half that time (4,3 years) before approaching the police or courts for assistance. One out of ten women (12%) experienced domestic violence for 10-20 years of their lives. From the intake forms, physical abuse, as opposed to sexual abuse, was cited as the most common form of physical victimisation reported by DVA applicants. Of the female applicants wanting protection from their intimate partners [FIPs], 83% (n=269) reported being physically abused. In about two-thirds of cases (67%, n=321) applicants cited physical abuse among the reasons they sought a protection order. … I have told him on numerous occasions that it is not working and he doesn’t want to let go. He has been threatening my … life and because I know what he’s capable of, I take his threats serious [sic]... [BL22-FIP] The study included an examination of the nature and extent of abuse reported by applicants. In doing so, the research interrogated both the patterns of domestic abuse over a period of time as well as the specific incident(s) that led to the victim’s decision to engage with the criminal justice process at that particular point in time. The analysis of these experiences provides some indication of the personal, social and systemic contexts that may contribute to the reasons why victims of domestic violence find themselves unable to continue with the criminal justice process. To some extent it allows inferences to be made about the extent to which certain types of abuse patterns contribute to the decision of women to withdraw from proceeding with formal interventions. The boomerang effect Domestic violence cases reported to the criminal justice system may be triggered by seemingly trivial events, but can quickly degenerate into violent episodes. Sometimes, the very act of seeking a protection order results in a ‘boomerang effect’, whereby seeking help only amplifies threatening behaviours and escalates violence by the respondent (perpetrator). Finn argues that ‘unlike most other types of crime, the offender is motivated18 to retaliate against a specific victim’.19 In other cases, the incident of violence that instigates legal action may be a seemingly minor incident (i.e. a non-physical but controlling behaviour), but the victim’s cumulative experience 6 Of all the applicants (n=503), sexual abuse or rape was reported in 16,5% of cases (n=83), with all of these cases reported by females applying for POs against an intimate partner. This means that of the female intimate partner applications almost one in five women had experienced sexual abuse or rape by their intimate partner. The findings showed that of those who reported being sexually abused, the majority, 82% (n=68), reported being forced to have sex or being raped. The majority of these women (90%) had been raped by their current partners, while the remaining 10% had been raped by an ex-partner. In addition to threats of further physical abuse, threats of rape or having the applicant raped were identified by women as reasons why they feared proceeding with the criminal justice process. Deadly decisions Victims of domestic violence are often threatened with death or more violence if they initiate or Institute for Security Studies proceed with criminal justice interventions. The interviews revealed that 44% (n=222) of all DVA applicants stated that their abusers had threatened to kill them. Threats to kill were present in 46% of cases (n=152) among women in the intimate partner category. There are also instances when the abuser threatens to send a third party to kill or harm the applicant. The following excerpt provides a glimpse into the contexts in which such threats are made and why returning to continue with the criminal justice process is, in some instances, more dangerous than protective: … My ex boyfriend [name] came to my house and demanded that I must go with him. As I refuse to do so he started banging on the door and then proceeded in punching me with his fist. I had to phone my mother to come … as we got near to the police office we notice that [he] was waiting opposite the police office with a sharp stone in his hand – he then walked next to us and started swearing at me [and] saying [you are sleeping with the policeman], you are a whore… He then threatened to get somebody to kill me if he don’t get bail and if he gets a bail he will come and kill me himself ... [W11-FIP] abuse, it is clear that there are high risks associated with continuing the process. In the work by Zoellner et al, prior threats to kill and threats to children were also positively related to whether victims obtained a final order or not.23 The range of abuses that directly or indirectly involve children are also prominent in the experiences of women in our study, varying from threats to ‘kidnap’ children to actual physical (and in some instances sexual) abuse of children. Threats to kill children were not uncommon in situations where the victim has threatened to, or has actually left, the abusive relationship. In 50% (n=251) of the interviews, applicants reported having at least one child affected by the abuse. From these 251 applicants a total of 469 children were listed as being affected by the domestic violence, with an overwhelming majority of these applicants stating that their children had been, and continued to be, psychologically damaged by the abusive environments they were exposed to. The desire to protect and care for children is a formative factor with respect to how women ‘manage’ domestic violence and the decisions they make, particularly whether continuing to finalise a protection order is worth the risk. Systemic issues Other studies, such as that by Belknap and Graham, have found a positive relationship between lethal threats to victims of domestic violence and their (dis)continuance with the criminal justice process.20 Belknap et al found that lethal threats by defendants featured in one fifth of cases they examined,21 while Anderson et al found that two thirds of domestic violence victims feared for their lives.22 In 15% of these cases victims reported that their abusive partners actually tried to kill them at some point in their relationship. Our interviews with DVA applicants revealed that fear of the perpetrator and coercion to withdraw – through explicit or implied threats – were significant factors affecting DVA applicants’ decisions to return to court. When combined with the fact that in 28% (n=134) of all cases in this study DVA applicants reported that their abuser used a weapon during an incident of SA Crime Quarterly no 37 • September 2011 Although domestic violence victims’ experiences with violence and their reasons for applying for protection orders defines the context in which the attrition of domestic violence cases takes place, it is what women say about not returning to court to finalise protection orders that is essential to this analysis. A critical finding from the interviews is that there are a number of systemic reasons why applicants do not finalise their POs. The difficulties faced in navigating the application process becomes particularly glaring when considering the cases where applicants returned to court to finalise the orders, but due to a systemic reason were unable to finalise their orders. The findings presented in Table 3 highlight these key reasons. Note that the respondents could have reported more than one reason for not returning. 7 Table 3: Systemic reasons for not finalising the protection order Reasons for not finalising IPO Number % Didn’t get a copy of the PO from the court 43 9% Didn’t get any paper saying when I must come back to court 38 8% The respondent didn’t sign the papers/fled 65 13% Didn’t realise I had to come back to court 27 5% Still with him, but I didn’t realise I had to come back to court 25 5% Applicant returned to court but no return of service at court 25 5% Struck off the court roll 24 5% Case postponed by the court on the return date 10 2% IPO not served 113 23% Lost confidence in the system 86 17% 38 8% 494 – Other Total responses In 23% of the cases, the interim protection order was never served. In an additional 13% of cases, the respondent did not sign for the service or fled from the police. Seventeen per cent simply lost confidence in the system. Nine per cent of applicants reported that they had not returned to finalise their orders because the court had not provided them with a copy of the PO. These applicants were therefore unlikely to know their return dates. In eight per cent of cases, DVA applicants explicitly stated that the court had failed to provide them with information about their return dates and in an additional five per cent of cases the applicants were not aware that they had to come back to court to finalise their orders. And yet another group (five per cent) reported that they had returned to court to finalise the order, but upon arrival learned that their cases had been struck from the court roll. These findings show that many DVA applicants do not return to court as a result of systemic problems, such as the courts not issuing 8 protection orders timeously or at all – even after multiple attempts by applicants to secure them from the courts – or cases being struck off the roll when applicants are late or at the wrong venue. Understandably, numerous attempts to get a copy of the protection order, endless phone calls to the courts and continued difficulties with serving protection orders are resulting in a critical and unnatural attrition of cases from the system. There are three studies from the United States that were published in the 1990s and two more contemporary studies that have directly investigated why women do not proceed with the process of finalising protection orders.24 These studies similarly found that poor criminal justice responses (multiple visits to court and the subsequent impact on work and childcare); difficulty with the service of protection orders; the lack of knowledge about the criminal justice system and other individual factors, such as the fear of perpetrators, were positively associated with the retraction of protection orders. Our study also found other ‘personal reasons’ for nonreturns that have not been highlighted in this article – for instance, permission to leave work, childcare and financial dependency – which indicate critical structural barriers for not returning to finalise protection orders. These challenges to women’s continued participation in a system that is meant to alleviate the additional burdens of violent domestic contexts, lie at the heart of the ‘attrition problem’. DISCUSSION It is widely accepted that intensity and brutality of domestic violence increases over time. Psychological and mental abuse escalates to physical violence. The use of weapons, and threats to harm or kill children and other family members intensify as victims attempt to seek outside help. Restrictions on victims’ movements, association with others and daily routines increase. Regardless of how ‘serious’ the reporting incident may or may not be, it is the victims of domestic violence who are the best assessors of their own risk to further violence. One of the applicants in this study illustrates this point: Institute for Security Studies … He has a serious history of abusing women who needed to be hospitalised. He keeps telling me he is going to kill all three of us if I don’t want to be with him. He is very aggressive and doesn’t care how he swears and performs in front of [my daughter]. He knows where I work and where [my daughter] goes to school. I fear for my safety and the safety of my daughter. He told me last night he would rather go to jail, then he knows why he’s alone, and that if I feel I want to go to the police and get an interdict, he doesn’t care. Nothing will stop him from being with me or [my daughter], and if he can’t be with us he will kill all of us …[BL22-FIP] It is reasonable to suggest that when victims fear for their safety they are extremely vulnerable; vulnerable to further violence as well as other factors that may have profound personal, social and economic consequences. Baker argues that women make a range of ‘reasoned choices’ in domestic violence relationships: staying with the abuser, lifting restraining orders and refusing to call the police are just some of these.25 The act of requesting formal legal intervention through the court system – despite the risks and barriers presented by the use of that system – indicates that South African women seeking protection orders are using the system to actively interrupt the cycle of violence. Victims use the system to arrange security – and a woman may have negotiated terms with her abuser under which she would not proceed if he left her alone. It is thus a conscious, protective strategy, where the decision to continue the process to its finalisation is tangential to the decision to utilise external measures in the first place. brutality of the abuse is likely to increase over time, and, most importantly, that seeking help can increase risk – whether actual or perceived. In reflecting on the types and severity of abuse suffered by domestic violence victims who apply for POs but who fail to finalise them, it is hard not to question why these victims would remain in their abusive situations, and why, with a legal remedy available to them under the DVA, they would not follow through and finalise the order. This research suggests that fear is a significant and common denominator in many cases of domestic violence. These ‘fears’ range from aggravating already violent situations and fear of reprisal from their abusers, to the fear of navigating what is perceived as a less than reliable criminal justice system. From the findings presented it is clear that the issue of attrition in domestic violence cases is a complex one, with a number of often interconnected factors impacting on an applicant’s decision-making with regard to finalising the PO. Only a few of these factors were presented here. These factors – both personal and systemic – may work together in a way that makes continuing the process simply impossible. They also, as Belknap and Sutherland argue, highlight the extraordinary risks that DVA applicants will take to actually proceed with the criminal justice process.26 To comment on this article visit http://www.issafrica.org/sacq.php NOTES 1. CONCLUSION The primary question in this research was: Why do victims of domestic violence who apply for a PO against their abusers, not return to court to finalise those orders? The findings from this research have, in some instances, mirrored findings revealed in previous studies on domestic violence. Most notably they have highlighted again that victims of domestic violence suffer a wide range of abuses, that the intensity and SA Crime Quarterly no 37 • September 2011 2. 3. 4. 5. MOSAIC is a non-profit organisation that supports youth and adults who experience domestic violence and abuse. MOSAIC works in partnership with government and other service providers in communities, delivering a range of violence prevention and support services. See L Artz, D Jefthas & Z Majeit, Reluctance, Retaliation and Repudiation: The Attrition of Domestic Violence Cases in Eight Magisterial Districts. GHJRU, University of Cape Town: South Africa, 2011. See L Artz, An Examination of the Attrition of Domestic Violence Cases in the Criminal Justice System in Post-Apartheid South Africa, Queen’s University Belfast, 2008, 1-425. Preamble, Domestic Violence Act (116 of 1998). Internationally, protection orders are variously described as restraining orders, protective orders, nocontact orders or peace orders. 9 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 10 The application process is contained in Section 4 of the Domestic Violence Act (116 of 1998). In instances where urgency cannot be shown an applicant is required to wait for the court hearing to see whether the order will be granted. Section 6(4) of the DVA (116 of 1998). L Zoellner, N Feeny, J Alverez, C Watlington, M O’Neill, R Zager & EB Foa, Factors associated with completion of the restraining order process in female victims of partner violence, Journal of Interpersonal Violence, 15(10), 2000, 1081-1099. By identifying the specific individual and systemic factors that contribute to attrition, we can begin to separate the broader structural challenges victims of violence face from the more immediately remediable systemic ones; the latter being a starting point from which to alleviate certain factors that discourage victims from proceeding with the criminal justice process. Of the 533 DVA applicants contacted for interviews, only 6% (n=30) did not want to proceed with interviews or were not contactable. Many applicants mentioned that they were appreciative that MOSAIC had contacted them to ask about the reasons for not returning to court. For some of these applicants, the very act of a follow up call was seen as a ‘support service’. For a full description of the methodology see Artz et al, Reluctance, Retaliation and Repudiation. Informed consent was established. The applicants were asked to sign a consent form that explained the purpose of the study. It also explained that should the applicant not return to court on the return date, the researcher would have permission to contact the applicant to discuss the reasons why s/he did not appear. To ensure that women were aware of their rights in this regard, it was also explained that the applicant had the right not to proceed with the application process. Applicants who do not return to court on the return date to finalise their protection orders. P Parenzee, L Artz & K Moult, Monitoring the Implementation of the Domestic Violence Act: First Report 2000-2001, South Africa: University of Cape Town, 2001. The definition for what constitutes an ‘intimate partner’ and what constitutes a ‘family member’ application is taken directly from the DVA [Section 1 (vii)]. An intimate partner is a male or female partner or ex-partner of the protection order applicant; and any of the following are thus placed in the family member category: father, mother, brother, sister, son, daughter, in-laws, as well as any other relationship as defined by this section. See for example JC Campbell, D Webster, J KoziolMcLain, C Block, D Campbell, MA Curry, F Gary, N Glass, J McFarlane, J Sachs, P Sharps, Y Ulrich, SA Wilt, J Manganello, X Xu, J Schollenberger, V Frye & K Laughon, Risk factors for femicide in abusive relationships: Results from a multisite case control study, American Journal of Public Health, 93(7), 2003, 1089-1097. Close scrutiny of these applications by men reveals that there are a notable number of counter-applications, meaning these male applicants are attempting to get a PO against someone – in a majority of cases female intimate partners – primarily because that person already has an order against them. 18. Own emphasis added. Finn, Statutory authority in the use and enforcement of civil protection orders against domestic abuse, Family Law Quarterly, 23(1), 1989, 43-73. 19. See RE Palarea, MA Zona, JC Lane & J LanghinrichsenRohling, The dangerous nature of intimate relationship stalking: Threats, violence, and associated risk factors, Behavioural Sciences and the Law, 17(3), 1999, 269-283. This study on intimate partner stalking found that the likelihood of reprisal is high. When assessing for levels of dangerousness, intimate partner stalkers were more likely to exhibit factors associated with dangerousness than non-intimate stalkers. More pertinent to the notion of the estranged ex-partner is the disturbing fact that “those cases in which the victim was threatened and then physically harmed … occurred four times more often in intimate cases than in non-intimate cases” and were “more likely to ‘make good’ on their threats by following them with some form of violent behaviour, and use more physical approach behaviours in contacting their victims”. 20. J Belknap & DLR Graham, Factors Related to Domestic Violence Court Dispositions in a Large Urban Area: The Role of Victim/Witness Reluctance and other Variables: Final Report. Research Report (No. 184232), US Department of Justice, 2000. 21. J Belknap, JL Hartman & DLR Graham, Introduction: Statement of the problem and review of the literature, in Belknap and Graham (eds), Factors Related to Domestic Violence Court Dispositions in a Large Urban Area. 22. MA Anderson, PM Gillig, M Sitaker, K McCloskey, K Malloy & N Grigsby, “Why doesn’t she just leave?”: A descriptive study of victim reported impediments to her safety, Journal of Family Violence, 18(3), 2003, 151155. 23. Zoellner et al, Factors associated with completion of the restraining order process. 24. S Keilitz, P Hannaford & H Efkeman, Civil Protection Orders: The Benefits and Limitations for Victims of Domestic Violence, Publication No. R201, Williamsburg, VA: National Centre for State Courts Research Project, 1997; A Harrell, B Smith & L Newmark, Court Processing and the Effects of Restraining Orders for Domestic Violence Victims, Washington, DC: Urban Institute, 1993; J McFarlane, A Malecha, J Gist, K Watson, E Batten, I Hall & S Smith, Protection orders and intimate partner violence: An 18month study of 150 Black, Hispanic, and White Women’, American Journal of Public Health, 94(4), 2004, 613-618; Zoellner et al, Factors associated with completion of the restraining order process. 25. PL Baker, And I went back: Battered women’s negotiation of choice, Journal of Contemporary Ethnography, 26(1), 1997, Academic OneFile, Thomson Gale. Retrieved 16 May 2007. 26. J Belknap & JL Sutherland, Content analysis of court transcripts, in Belknap and Graham (eds), Factors Related to Domestic Violence Court Dispositions in a Large Urban Area. Institute for Security Studies