5SA Crime QuArterly No. 56 • JuNe 2016 * Heidi Mogstad is an MPhil student in justice and transformation at the University of Cape Town (UCT). Dominique Dryding is an MA student and African Leadership Centre Fellow at King’s College, London. Olivia Fiorotto is an MA student in global health and public policy at Edinburgh University. When the data for this study were collected, all were students at UCT and Mogstad and Dryding were working for Ndifuna Ukwazi (NU); a Cape Town-based non-governmental organisation that uses research and strategic litigation to campaign for justice and equality in poor and working class communities. NU cooperates closely with community-based organisations in Khayelitsha and was one of the organisations that lodged a formal complaint against the South African Police Service (SAPS) in Khayelitsha and called for a commission of inquiry in 2011. Policing the private Social barriers to the effective policing of domestic violence The South African Domestic Violence Act of 1998 (DVA) is widely recognised as being a progressive law.1 By including comprehensive definitions and remedies, the act’s drafters sought to give victims and survivors of domestic violence the best protection and assistance a legal system could provide.2 The DVA imposes specific obligations on police to ensure that domestic violence is not neglected.3 However, despite this, domestic violence remains pervasive and under-reported in South Africa.4 A large body of empirical research shows that most victims of domestic abuse have not gained effective protection from the DVA or from the criminal justice agencies charged with its enforcement.5 In response to the gap between South Africa’s progressive legislation and the reality experienced by victims, researchers have sought to identify barriers to the effective implementation of the DVA. These evaluations have focused on structural and institutional barriers to implementation, such as the South African Police Service’s (SAPS) lack of resources, inadequate training and knowledge.6 Several studies have also documented the structural obstacles that marginalised women face in accessing justice.7 These factors are crucial, especially in poor and resource-scarce communities where the police–to–civilian ratio is low, the relationship between the community and the police is characterised by a significant lack of trust, and most victims are unemployed and poor. However, law enforcement interventions are not neutral or value free. Policing domestic violence requires authorities to interfere Heidi Mogstad, Dominique Dryding and Olivia Fiorotto* heidimog@gmail.com d.dryding@gmail.com olivia.fiorotto@gmail.com http://dx.doi.org/10.17159/2413-3108/2016/i56a414 The limited ability of police to assist victims of domestic violence is often viewed as an institutional failure; a consequence of a lack of resources or inadequate training. This article presents key findings from a qualitative study of perceptions of and attitudes towards domestic violence in the South African township of Khayelitsha that highlight the complexity of responding to this form of violence. The research found that prevailing social norms and beliefs in Khayelitsha prevent domestic violence victims from seeking help from the police and that, unless there is a change in social norms, it is unlikely that there will be an increase in the reporting of cases of domestic violence. InstItute for securIty studIes & unIversIty of cape town6 in private and previously unregulated spheres. To be effective, the policing of this crime depends on community norms that recognise and support police intervention as acceptable and appropriate. This article is based on a 10-month qualitative research project on domestic violence in Khayelitsha, a partially informal township on the outskirts of Cape Town.8 Khayelitsha was established by the apartheid regime in 1983 under the terms of the Native Urban Areas Act, to consolidate Cape Town’s legal African population in a racial enclave on the urban periphery.9 Despite a massive rollout of social grants and significant differences in wealth and living standards between Khayelitsha neighbourhoods, poverty and unemployment remain widespread.10 Khayelitsha is also burdened with high levels of crime and social violence.11 As noted by Seekings, crime is a constant consideration in the lives of people living in Khayelitsha.12 The township is particularly notorious for its high rates of gang violence, vigilantism and public and sexual assaults.13 Although local organisations and gender scholars recognise domestic violence as a prevalent social problem in Khayelitsha, violence in the household is usually overshadowed by the overwhelming focus on violence in the public sphere.14 The research was prompted by the Khayelitsha Commission of Inquiry (KCoI), also known as the O’Regan–Pikoli Commission. After substantial lobbying by local organisations, Western Cape Premier Helen Zille appointed the commission in August 2012 to investigate allegations of police inefficiencies and a breakdown in the relationship between the community and the police in Khayelitsha.15 The commission was tasked with investigating all policing activities in the area, including the policing of domestic violence. This attention to domestic violence was unprecedented, as other South African commissions of inquiry focused on uncovering violent crime or events have either silenced or sidelined domestic violence and violence against women.16 Through a combination of expert and victim testimony, the commission revealed the systematic failure of police to comply with the DVA and National Instruction 7/2009, which regulates enforcement of the act.17 However, the commission’s narrow mandate left several important issues unexplored. Firstly, by focusing almost exclusively on structural and institutional problems in policing, the commission did not consider or hear evidence about the role of cultural and gendered norms and beliefs in shaping the social limits of appropriate behaviour.18 Secondly, although the commission confirmed that there was a breakdown in the relationship between the community and the police, not enough was revealed about why so few victims in Khayelitsha chose to report cases of domestic abuse and to what extent the low rate of reporting could be explained by a lack of trust in the police.19 To help address these knowledge gaps, the authors undertook a qualitative research project in partnership with the Social Justice Coalition (SJC). The research was explorative in nature. Its purpose was to map local perceptions of, and attitudes towards, domestic violence and its policing. In this article, we reflect on some of our key findings, focusing specifically on how social norms and beliefs regulate experiences of and responses to domestic abuse. We begin by describing the research methods, followed by a presentation of the findings. Next, we discuss the role of social norms and beliefs in shaping the reluctance to involve police in cases of abuse. Finally, we conclude by summarising and discussing the implications of our findings. Methods Between September 2014 and June 2015 data were collected from five focus groups and seven in-depth individual interviews. The five focus groups were held in Khayelitsha in venues provided by the SJC. There was a total of 40 participants. Two of the focus groups consisted of men only, and three of women only. We divided our focus groups along gendered lines to identify the differences between how men and women speak and feel about domestic violence, and to ensure that participants would feel safe to speak openly. Since all participants were first-language Xhosa speakers, Xhosa-speaking translators helped facilitate the discussions. 7SA Crime QuArterly No. 56 • JuNe 2016 Focus groups were used because they produce data and insights not easily accessible in individual interviews.20 As noted by Albrecht: Given that focus groups are social events involving the interaction of participants and the interplay and modification of ideas, such a forum for opinion gathering may render data that are more ecologically valid than methods that assess individuals’ opinions in relatively asocial settings. A focus group responding to a new idea might generate opinions more like those of the public than would even a large number of isolated respondents.21 Focus groups are especially useful when studying group cultures and exploring degrees of consensus.22 To supplement our data and mitigate problems associated with focus groups,we also conducted three selective in-depth interviews with women willing to talk about their personal histories of abuse and relationships with the police.23 These interviews provided rich and detailed data and allowed us to further probe what factors and beliefs shaped victims’ responses to abuse. We conducted two interviews with local counsellors of domestic violence survivors, two interviews with local activists working with gender-based violence, and one interview with a female member of one of Khayelitsha’s community policing forums (CPFs).24 These seven interviews provided important contextual knowledge and offered the opportunity to discuss preliminary findings. Our interviewees were identified and recruited with assistance from SJC and the social justice organisation Ndifuna Ukwazi (NU), where Mogstad and Dryding worked during the research period. All individual interviews were conducted in English. The focus group discussions and interviews were transcribed into English and analysed, using thematic analysis.25 Our primary data were supplemented by informal background interviews with local activists, members of the CPFs, and a journalist working on the KCoI. The KCoI’s final report, transcripts of victim and expert testimonies, and meeting minutes from the community policing sub-forum dealing with domestic violence were closely read and analysed.26 The study was self-funded and not reviewed by an ethics committee. Limitations and clarifications The terms ‘local’ and ‘the community’ are ambiguous. They are not used here to imply that all people in Khayelitsha share the views discussed. We cannot assume the presence of common values and beliefs across this large, heterogeneous, diverse township consisting of people with different experiences, knowledge, living standards and educational levels.27 Participants in the focus groups were recruited by the SJC on a voluntary basis and identified using purposive sampling.28 Because we were interested in exploring norms and beliefs, personal experience of abuse and/or of engagement with the police in the case of abuse were not considered necessary criteria. However, in order to stimulate clear, focused and in-depth discussions, it was important that the participants in our focus groups had some shared experiences of the challenges of policing domestic violence in their particular area.29 We therefore recruited participants who were of similar age, from similar cultural backgrounds, and who shared similar living standards and income levels.30 The majority of participants in the focus groups were in their late 30s and unemployed. Most of the participants were married, and almost all lived with a partner and children. All focus group participants were isiXhosa speaking. Many participants had been born in the Eastern Cape and had been residing in Khayelitsha for various lengths of time.31 Participants lived in informal settlements and had limited or irregular access to sanitation, water and electricity.32 This is important to note, since the KCoI revealed particular problems and challenges with policing in the informal areas of Khayelitsha.33 The focus group participants were all members of the SJC, although their history of membership and participation in the organisation varied significantly.364 The focus groups did not include anyone who self-identified as lesbian, gay, bi-sexual, transgender or intersex. The findings cannot therefore shed light on this population’s particular concerns and challenges in dealing with police or domestic violence. Although our findings are not representative, they provide useful insights into how cultural norms and InstItute for securIty studIes & unIversIty of cape town8 beliefs can complicate well-intended legal inter- ventions, such as the policing of domestic violence. It is also important to acknowledge our positionalities as three young, coloured and white women who do not speak isiXhosa.35 Our personal traits and backgrounds influenced not only the questions we asked and what our participants chose to reveal but also how we interpreted responses and framed findings in this article.36 While the fact that our participants were of the same gender, from similar backgrounds and close in age may have helped facilitate trust, participants’ responses were also likely influenced by their relationships to other participants and by the sensitive nature of the topic. Although many of our participants were comfortable speaking English, our partial reliance on translators’ interpretations meant that we lost some nuance and richness in responses. However, using local translators also reduced barriers to participation and helped us understand culture-specific references.37 Findings In this section, we present empirical findings suggesting that prevailing social norms and beliefs in Khayelitsha prevent domestic violence victims from seeking help from the police. While our arguments are supported by the data we collected in the field, we also draw on the insights from a wide range of studies questioning the ability of legal reforms to influence entrenched cultural and gendered norms, attitudes and practices in South Africa.38 Barriers to reporting Nearly all of the research participants maintained that involving the police in cases of domestic abuse is inappropriate because domestic abuse is a private issue. Although they acknowledged that abuse is harmful and that something ought to be done about it, participants said that involving the police was unacceptable, or disloyal. Police interference was also seen to violate culturally correct procedures. When discussing appropriate ways of dealing with domestic violence, nearly all participants agreed that it would be best if the couple involved settled the issue without any external interference. The exception was one male participant who suggested that street committees could step in as mediators.39 Participants also said that the only culturally accepted alternative to settling the dispute between partners was to seek guidance from in-laws. As one male participant explained: You see, here in Khayelitsha … if I do something to my wife or she does something to me, it is very important to not go first to the police station. If I am abusing my wife, she may get out of my home and go to her home and tell her relatives, and after that, they will call my relatives … and then we will have a meeting of some sort and solve the problem without interference from the police.40 Our research indicated that married women faced especially strong pressure to restore peace in their families without police intervention. Whereas some men suggested that using in-laws as mediators was an example of ‘culture working’, female participants emphasised that ‘solving the problem’ was usually done without much consideration of women’s personal opinions and well-being. In addition, it was stressed that the in-laws were involved not to end the abuse but to broker the peace and ‘keep the family together’. As a female interviewee explained: In the white world, people go to therapy to find out what is really causing this problem … but in our lives, we have the option to sit down with the elders, and then they will give you advice on how to make your marriage work … Sometimes the family gives you good advice, but let’s say, if you are a makoti [daughter-in-law] they don’t like, they do not think about you. Female focus group participants, interviewees and counsellors unanimously stressed that in-laws generally took the husband’s side in a dispute, neglected women’s opinions and suffering and left women with ‘little control over the situation’. While some women experienced pressure from their in-laws or family members to stay with abusive partners, many female participants had also internalised cultural norms prioritising the welfare and maintenance of the family above their own well-being.41 Several women suggested that they tolerated abuse because they did not 9SA Crime QuArterly No. 56 • JuNe 2016 want their children to grow up without a father. In such situations, involving the police was seen as especially problematic as the women did not want to be responsible for sending their children’s father to prison. A number of female participants suggested that what they experienced was neither unique nor sufficient reason to break familial bonds and cultural expectations. As one interviewee succinctly put it, ‘My mother was able to endure it [an abusive relationship], so why shouldn’t I?’42 Behind these statements was an acceptance of spousal abuse as ‘normal’ and an experience that did not justify intervention by police or any other outside actors. In addition, both men and women said that police intervention was inappropriate because both parties were responsible for the abuse. Men felt particularly strongly about this and repeatedly stressed that women also abused their husbands and boyfriends. Several female participants also insisted on sharing the blame for domestic violence with their partner. The following reflection by one female participant is typical of many of the stories we heard. It illustrates how some participants internalised blame for the abuse they experienced and exemplifies how many female participants described domestic violence as normal.43 It happens every weekend … We shout at each other, he beats me, I try to hit him back … But when you wake up in the morning, you will feel very sorry for your partner and what you have done, and you will never go to the cops. A few men and women said that women were guilty of deliberately provoking men, for example by shouting or nagging. In one male focus group, it was suggested that some women want their partner to beat them so ‘they can feel that they are being loved and fought over’. While the idea that some women interpret abuse as an expression of love or care was repeated in all the female focus groups, participants were careful to emphasise that they spoke about other women, not themselves.44 Participants expressed reluctance to involve the police in domestic violence when the victim was a person whom they knew or cared about. When asked if they would call the police if they saw or heard a friend or neighbour being violently abused, most participants said they would be highly unlikely to do so. When asked why, several stressed that it was inappropriate to meddle in other people’s affairs. Some female participants stated they would help the victim in other ways, for example, by allowing the victim to sleep in their house or by encouraging the victim to leave the abusive partner or seek help from a non-governmental organisation (NGO) or social worker. One man said that he would call the police if he heard his neighbours fighting and the abuse was so loud that he could not sleep. Social costs of reporting Participants were also discouraged from seeking help from the police due to the social costs of doing so. While sanctions may take various forms, we focus specifically on the social costs attached to identifying as victims and seeking help from the police.45 Both men and women worried that police interference would affect their status and reputation in the community. When discussing male abuse, all male participants mentioned concerns with being ridiculed and humiliated. As one male participant stated, ‘If I were to report a case that I was raped by a girl or tell my friend … tomorrow the whole community would know what happened to me, and it would become a joke in the community.’46 The following extract from the transcript of a male- only focus group illustrates that cultural ideals of what it means to be a ‘real man’ appeared to have an especially strong influence on the behaviour of men, who said they often kept their personal experiences of abuse secret due to fear of being seen as weak or ‘controlled by their women’. Man 1: The problem is that we as men are ashamed to come out. Interviewer: Why is that? Man 2: The problem is the way we grew up … because men can’t cry. Something like that ... So you take it as a disgrace to go to the police station to report [abuse] … and people in the community will also laugh behind your back. Let’s say, your girlfriend kicked you or whatever, and you got bruises … You come and tell people, ‘She did this InstItute for securIty studIes & unIversIty of cape town10 to me’ … Then you will notice that they laugh at you … and keep asking, ‘What did she do to you?’ and laugh. Interviewer: Is this what makes it difficult to go to the police? Man 2: Very difficult. Very difficult. Man 3: You don’t want to feel inferior. That she has the power. Feel like a coward. Like the men, the women primarily feared gossip and judgement. One woman explained that she could not talk about her abuse with her closest family members or friends, as ‘you know it is going to spread and everyone is going to look at you differently’.47 Women viewed self-identifying as victims as shameful and embarrassing. Female participants’ reasons for shame differed. Some women said that it was shameful to go to the police as others might think that they had overreacted or acted in a way that justified the abuse. Others viewed involving the police as disloyal to both their partner and family. Some women stressed that identifying as victims would make them appear weak and powerless in their relationships. When the women who said this were asked what they would do if they were exposed to abuse, they suggested that it was ‘better to just leave [their abusive partner]’ than involve the police.48 One of the interviewees was in an abusive relationship for more than eight years but never considered reporting the abuse to the police. When asked why she explained that she was known as a strong, opinionated woman in the community and feared that identifying as a victim would reflect badly on her. She said it felt shameful to admit to being abused, even to herself, as it contrasted with her own self-image. Only once she had chased her husband out of the house did she tell her friends and family about her abuse. Contrary to what she had believed, this did not result in a loss of status in the community. However, she said she was happy that she had never brought the police into the picture, because it would have called into question her ability to deal with things on her own. Another interviewee confessed she had hidden her abuse from her friends and family for nearly 10 years because she believed they would judge her for staying with a man who was abusing her. You make means for people not to judge you for staying. Even your family. There were times when I would pack my things and take a bus home to Eastern Cape, no matter how much it would cost me. [But] when I was home I would act as if I was there just for fun, visiting, whereas I was there to express my feelings … and I will come back to Cape Town and he will be scared that ‘Ooh, she has told the mother everything’ but I didn’t say anything … I covered it up because I did not want my family to look down on him. Because I am his partner. And when you look down on him, you look down on me too.49 After having kept the abuse a secret for nearly a decade, the woman eventually told her family and in-laws about the abuse and filed for a divorce. At this point, the woman had a sustainable job and income and was not economically dependent on her partner. But after confronting the stigma of self-identifying as a victim, the woman faced considerable external pressures to keep her family together. Some of the pressure came from her in-laws, who were largely unresponsive to her interests and arranged family meetings to prevent the divorce from going ahead. Her own mother, who she had initially been afraid would judge her for staying with an abusive partner, also begged her to stay in the relationship for her sake and for the sake of the children. This woman’s story illustrates that abused women may face various forms of external and internal pressures to stay in abusive relationships. Even after taking the important and difficult step to self-identify as a victim and filing for a divorce, it took the woman an additional five years before she finally managed to leave her partner. Attitudes towards police For some participants, the unwillingness to involve police in cases of abuse appeared to be informed by their distrust of police in Khayelitsha. When asked what they believed would happen if they approached the police as victims of abuse, most participants suggested that the police were unlikely to provide any meaningful assistance because the police shared the same attitudes towards abuse and victimhood as they did. Participants were particularly sceptical of the police’s motivation to assist male victims, believing officers 11SA Crime QuArterly No. 56 • JuNe 2016 would not take them seriously and might laugh at them for ‘acting like a woman’.50 Man: If you are violated as a woman, you can get help, but if you are violated as a man, you can’t get any help. Interviewer: Why is that? Man: They [the police] take it as a joke. They laugh at you as a man. And if you fight back yourself, you are arrested. Visibly frustrated at being scapegoated as abusers while their own alleged suffering was left unacknowledged, the men emphasised that involving police would never work to their advantage as officers would always take the woman’s side. The problem is that if I am fighting with you, the police will not ask ‘What is happening?’ or ‘How did this thing start?’ … They will just take me to the station, although it was the woman’s fault. In one focus group a few men complained that some women reported consensual sex as rape, or abused their male partners knowing that they would never go to the police.51 While male participants indicated that female victims of abuse would receive better help and support from police, several female participants insisted that involving the police was pointless as they would not provide any real assistance.52 As the following brief extract from a female focus group illustrates, the women believed that police were not interested in helping them as they, too, considered abuse to be a private matter. The women also suspected that police were tired of dealing with women’s complex needs. Woman 1: It is a challenge in our police station. If it is domestic violence, the police say it is a family matter. (Other women nod and agree.) Woman 2: Because if the wife goes and reports it [the abuse] and opens a case to go to court … all of the sudden, after a month, she drops that case. You see … [In the eyes of the police] I just use a government article. Misuse it, you see. While most of our participants expressed a strong distrust in the police’s willingness to assist victims of abuse, it is important to note that not all criticisms were based on first-hand experience. While some participants shared personal experiences of encounters with the police, others’ disapproval was based on second-hand accounts or assumptions about how the police would respond. In contrast, the member of the police sub-forum and the two local counsellors who were interviewed argued that the police in Khayelitsha are, in their experience, better trained to respond to domestic violence today than a few years ago. The counsellors emphasised that today police are more sensitive and respectful of male and LGBTI victims.53 Regardless of whether this is true or not, participants’ negative view of the police is likely to reinforce their unwillingness to approach them in cases of abuse. Discussion of findings Norms play a crucial role in individual choice, by specifying what is acceptable and what is not in a society or a group.55 Norm-compliance is ensured in two ways. Firstly, people are encouraged to conform to a set of norms by expectations or threats of sanction. Sanctions can be both positive and negative but often include exclusion, ostracism or violence. Secondly, norm-compliance is ensured through the more subtle process of internalisation, in which members of society are socialised to think of certain ways of being and doing as normal and natural. If norms are successfully internalised, external sanctions are not needed to elicit conformity, as ‘norm-abiding behaviours are perceived as good and appropriate, and people will typically feel guilt or shame at the prospect of behaving in a deviant way’.55 Our data suggest that both these dynamics shaped participants’ reluctance to involve the police in cases of domestic abuse. Our research indicates that participants’ reluctance to involve the police was strongly influenced by the social shame, stigma and humiliation expected from self-identifying as a victim and seeking help from police; this was the case for both men and women. The consequence of this is that abusers enjoy de facto impunity while victims are left isolated, disempowered and ashamed.56 InstItute for securIty studIes & unIversIty of cape town12 However, there are nuances that should be explored. Although participants emphasised that abuse was bidirectional, this does not mean that women were considered as abusive as men, or that participants believed that men suffered as much as women. Female participants who spoke about abuse inflicted on their male partners usually framed it as an act of resistance. It is also important to recognise the performative nature of interviews and focus groups. Participants do not simply communicate information but define and position themselves in front of their audience and bring certain truths into being. With this in mind, female participants’ eagerness to share stories of their acts of abuse against their partners might be interpreted as attempts to distance themselves from an image of women as passive, powerless victims. Similarly, male participants’ frequent insistence that ‘men are also abused’ is a clear contestation of the one-dimensional image of men as violent aggressors.57 There were limits to participants’ reluctance to involve the police in cases of abuse. Both men and women said they would consider approaching the police if they believed the abuse had an extremely negative influence on their children, for example if it resulted in a failure to provide food or pay children’s school fees, or was also directed at children. However, in these situations police interference was identified as a last resort, and both men and women said they would rather leave with their children or go to a social worker. Some women said they would consider reporting their husbands to the police if they did not have children, or if their children were older. Several women explained that they might choose to stay in an abusive relationship to protect their children from anticipated economic hardships. However, our research indicates that economic factors often interact with social norms, placing added pressure on women to accept and endure abuse, and increasing the costs associated with seeking help. When pushed, some participants admitted that they would want to get the police involved if the abuse became very violent and they feared they (or a loved one) might be seriously hurt or killed. This finding aligns with other studies showing that cultural norms might condone and privatise domestic abuse, but only within certain boundaries of severity.58 However, even in these scenarios participants had highly ambivalent feelings about seeking help from police, partly due to the anticipated personal and familial costs of police intervention. Finally, the research indicates that the reasons participants felt shame were strongly influenced by their own gender and their views about gender. The research indicates that a dominant model of masculinity in Khayelitsha is associated with power and control over both self and others. Consequently, identifying as a victim was seen as unmanly, shameful and humiliating. Female participants’ reasons for shame differed, and were influenced by the model of femininity they endorsed. Women who endorsed a traditional form of femininity based on cultural ideals of submissiveness and endurance feared that they would be perceived as overreacting or deserving of abuse, as they had misbehaved or failed to act like a proper woman.59 As indicated, involving police was believed to incur specific social costs, as women feared they would be stigmatised by family members or others who would see this measure as an act of disloyalty or a violation of culturally accepted procedure.60 In contrast, women who endorsed a more progressive form of femininity feared that identifying as victims of abuse would make them appear weak and powerless in front of their family and friends.61 Here, involving the police was defined as deeply embarrassing and was expected to have a negative impact on women’s status and reputation in the community. Importantly, however, these models of femininity are ideal types. As the personal stories of our interviewees indicate, women’s decisions to stay in abusive relationships and not seek help from the police can be influenced by various forms of external and internal pressure, operating simultaneously or at different times. This indicates that thinking in terms of a rigid traditional/ progressive binary is not always useful. Discussion The research findings reveal the external and internal pressures on women to keep families together in the face of abuse, almost at all costs. The internalisation of norms led women to downplay and tolerate 13SA Crime QuArterly No. 56 • JuNe 2016 abuse to the extent that it compromised their physical and psychological health and well-being. These factors clearly complicate the job of policing domestic violence, as the private nature of this crime makes police intervention largely dependent on victims’ identification of abuse as a crime worthy of intervention. The research also indicates that women’s response to abuse is particularly influenced by social norms defining what is best for their children. Following Carol Gilligan, this thinking might be understood as a distinctly female moral reasoning guided by a moral orientation towards relationship maintenance and care for others.62 As Meyer stresses, such decisions should not be interpreted as irrational acts. In contrast, ‘costs and benefits are simply assessed on a broader, less selfish scale, taking into account the costs and benefits for individuals close to the rational decision-maker’.63 The fact that abused women may often prioritise their children’s well-being above their own suffering demonstrates the importance of existing laws and policies designed to address the safety and well- being of both mothers and their children. The findings highlight the significance for victims of social shame and stigma attached to public revelation and help-seeking in cases of abuse. Rather than dismissing victims’ decisions not to involve the police in cases of abuse as a sign of passivity, non-cooperation or acceptance of the status quo, our findings suggest that non-reporting is a calculated, legitimate strategy to protect oneself from a variety of social costs, including social stigma, gossip, humiliation and shame. This builds on other studies conducted in South Africa.64 Our findings offer a deeper and more nuanced understanding of the social costs of reporting domestic violence by showing that victims’ reasons for shame and embarrassment are dependent upon the victims’ gender and the model of masculinity and femininity that they endorse. Non- reporting can be a calculated strategy to avoid social and other costs, and as such, victims may be served better by interventions that do not rely on a criminal justice response. However, steps should also be taken to reduce the social costs associated with seeking help from the police. Since sensitive and empathic policing is necessary to overcome the stigma associated with reporting domestic violence, the KCoI’s recommendation, that improved internal and external oversight over implementation of the DVA is necessary, should be supported.65 However, reducing the costs of help seeking depends on transforming social and gendered norms and attitudes towards police intervention and victimhood held by police and society at large. By attending to the performative nature of focus groups discussions, this article has drawn attention to men and women’s reluctance to identify with harmful yet prevailing stereotypes of ‘women-as- passive-victims’ and ‘men-as-aggressive-abusers’. This reluctance indicates that many women may see advocacy that emphasises female victimhood as disempowering. Women-centred advocacy may also alienate men who are frustrated at being scapegoated as abusers, while at the same time being fearful of the consequences of identifying as victims of abuse. To reduce the social costs associated with self-identifying as victims of abuse, further steps must be taken to confront simplistic and disempowering discourses and create opportunities for men to explore alternative masculinities. Conclusion Problems in policing are commonly framed as institutional failures. When thinking about policing in this manner, it is easy to conclude that the solution lies within the institution itself, or requires more resources and support from the government. Combrinck and Wakefield, for example, argue that ‘the South African Police Service holds the key to a successful implementation of the [DV] Act’ and recommend that persistent shortcomings be addressed with better and additional training.66 The Khayelitsha Commission of Inquiry also focused on structural and institutional challenges in policing and crafted recommendations with these concerns in mind. This article has examined the challenges and limitations of policing domestic violence from a different angle. Reflecting on key findings from a small qualitative study of local perceptions and attitudes InstItute for securIty studIes & unIversIty of cape town14 towards domestic violence and the policing of this crime in Khayelitsha, we have drawn attention to the powerful disciplinary influence of social norms and beliefs in regulating responses to abuse. While acknowledging that victims’ experiences of and responses to abuse are shaped by a variety of factors, our findings suggest that victims’ responses to domestic violence are constrained by dominant social norms and beliefs, framing police involvement in cases of abuse as being inappropriate and shameful. The findings suggest that the social norm defining household violence as a private issue regulates the behaviour not only of victims but also of potential witnesses and third parties. Whereas higher compliance with the DVA will necessarily require considerable resources, this suggests that the effective policing of domestic violence is predicated on shifts in norms and beliefs, defining police interference in cases of abuse as problematic, if not unthinkable. Acknowledgements This article has its roots in a collaborative project by the SJC and a UCT Justice and Transformation postgraduate class taught by Dr Hugo van der Merwe. We would like to thank Chumile, Khanyiswa, Zimkita, Ntombi and the rest of SJC and NU for helping us to arrange and conduct our focus groups and interviews, and for providing useful and constructive feedback and information throughout the process. We would also like to thank Dr van der Merwe for his instruction and feedback, as well as Katherina Mann, who participated in the first stage of the data collection process. For helpful comments on earlier drafts we would also like to thank Dr Chandre Gould, Dr Andrew Faull, Prof. André du Toit and two anonymous reviewers. Lastly, and most importantly, we thank the men and women who gave their time to share their thoughts on and experiences of domestic violence and policing with us. To comment on this article visit http://www.issafrica.org/sacq.php Notes 1 See L Vetten, Domestic violence in South Africa, Institute for Security Studies (ISS), Policy Brief 71, 2014. 2 Ibid.; Anticipating problems associated with statutory definitions, the law also includes a ‘catch-all’ category covering abuse that is otherwise not covered. 3 L Vetten, Addressing domestic violence in South Africa: reflections on strategy and practice, Centre for the Study of Violence and Reconciliation, 2005, 5. 4 Vetten, Domestic violence in South Africa. 5 See, for example, C Albertyn et al., Women’s freedom and security of the person, in E Bonthuys and C Albertyn (eds), Gender, law and justice, Cape Town: Juta, 2007, 323. 6 See, for instance, P Parenzee et al., Monitoring the implementation of the Domestic Violence Act, Cape Town: Institute of Criminology, University of Cape Town, 2001; L Vetten, Show me the money: a review of budgets allocated towards the implementation of South Africa’s Domestic Violence Act, Politikon: South African Journal of Political Studies, 2, 2005, 19–26; S Mathews and N Abrahams, Combining stories and numbers: an analysis of the impact of the Domestic Violence Act (No. 116 of 1998) on women, Gender Advocacy Programme and Medical Research Council, 2001; J Smit and F Nel, An evaluation of the implementation of the Domestic Violence Act: what is happening in practice?, Acta Criminologica, 15, 2002, 3; L Artz, Magistrates and the Domestic Violence Act: issues of interpretation, Institute of Criminology, University of Cape Town, 2003; M Manamela et al., Policing domestic violence effectively at Rietgat Police Station: an assessment, Acta Criminologica: CRIMSA Conference: Special Edition, 2, 2010, 99–113; D Smythe, Missed opportunities: confiscation of weapons in domestic violence cases, SA Crime Quarterly, 10, 2004, 19–26; H Combrinck and L Wakefield, Going the extra mile: police training on domestic violence, SA Crime Quarterly, 31, 2010, 27–34; L Artz and D Smythe, South African legislation supporting victims’ rights, in D Lillian and R Snyman (eds), South African handbook on victimology, Pretoria: Van Schaiks, 2005. 7 See specifically D Smythe and L Artz, Money matters: structural problems with implementing the DVA, Agenda, 66, 2005, 24–33; L Artz and D Smythe, Bridges and barriers: a five year retrospective on the Domestic Violence Act, Acta Juridica, 2005; P Parenzee and D Smythe, Domestic violence and development: looking at the farming context, Cape Town: Open Society Foundation South Africa and Institute of Criminology, University of Cape Town, 2002; D Smythe and P Parenzee, Acting against domestic violence, in W Dixon and E Van der Spuy (eds), Justice gained? Crime and crime control in South Africa’s transition, Cape Town: Juta, 2004; L Artz, Violence against women in rural Southern Cape: exploring access to justice through a feminist jurisprudence framework, Institute of Criminology, University of Cape Town, 1999; Matthews and Abrahams, Combining stories and numbers. 8 The Domestic Violence Act (DVA) recognises domestic violence as a complex crime that can involve many different victims and perpetrators. Our research focused on violence between intimate partners, broadly interpreting violence in line with the act to include physical, sexual, emotional and economic forms of abuse. 9 For a detailed overview of Khayelitsha see J Seekings, Economy, society and municipal services in Khayelitsha, report for the Commission of Inquiry into Allegations of Police Inefficiency in Khayelitsha and a Breakdown in Relations between the Community and the Police in Khayelitsha, Centre for Social Science Research, University of Cape Town, 2013. 15SA Crime QuArterly No. 56 • JuNe 2016 10 Ibid. See also G Super, Violence and democracy in Khayelitsha: governing crime through the community, Stability: International Journal of Security & Development, 4:1, 2015, 1. 11 See, for example, EA Pieterse, Counter-currents: experiments in sustainability in the Cape Town region, Johannesburg: Jacana, 2010, 261. 12 Seekings, Economy, society and municipal services in Khayelitsha, 23, 19. See also Mthente Research, Community safety barometer 2011/2012 – precinct reports for Lingelethu West, Harare and Khayelitsha (Site B), commissioned by Western Cape Government, 2012, http://www.khayelitshacommission.org.za/media-press/ press-statements/100-general.html?start=48 (accessed 1 September 2015). 13 For an overview see C O’Regan and V Pikoli, Towards a safer Khayelitsha. Report of the Commission of Inquiry into Allegations of Police Inefficiency and a Breakdown in Relations between SAPS and the Community of Khayelitsha, ch. 4, 40–45, http://www.khayelitshacommission.org.za/ images/towards_khaye_docs/Khayelitsha_Commission_ Report_WEB_FULL_TEXT_C.pdf (accessed 1 September 2015). 14 Informal conversation with representative from MOSAIC, personal communication, April 2015. See also MOSAIC, Fact sheet 1: Domestic violence in Khayelitsha, http://www. mosaic.org.za/kfactsheet.pdf (accessed 15 June 2015). 15 See Khayelitsha Commission, http://www. khayelitshacommission.org.za (accessed 15 August 2015). 16 The widely heralded South African Truth and Reconciliation Commission has, for instance, been criticised by gender scholars for privileging the recovery of truths pertaining to political and public violence and thereby marginalising the suffering women experienced during apartheid, both inside and outside their homes. See Fiona Ross, Bearing witness: women and the Truth and Reconciliation Commission in South Africa, London: Pluto Press, 2003; S Meintjes, Gendered truth? Legacies of the South African Truth and Reconciliation Commission, African Journal on Conflict Resolution, 9:2, 2009. More recently, both the Marikana Commission of Inquiry itself and the media and academia’s coverage of it have been criticised for focusing narrowly on police brutality towards (male) miners while silencing women’s narratives and needs and the wider socioeconomic context. See A Benya, Absent from the frontline but not absent from the struggle: women in the Marikana Massacre, Femina Politica, 22:1, 2013, 144–147. 17 For a victim testimony, see Ms Ntsilane’s testimony in O’Regan and Pikoli, Towards a safer Khayelitsha, Appendix B, August 2014. For an overview of the expert testimonies from L Vetten and L Artz, see O’Regan and Pikoli, Towards a safer Khayelitsha, 140–145 (phase 1), 336–340 (phase 2). 18 Gail Super argues that the commission’s narrow focus on policy inefficiency also diverted attention away from the socioeconomic conditions that generate crime. See Super, Violence and democracy in Khayelitsha. 19 The commission concluded that the relationship between the community and its police is characterised by a significant level of distrust. See O’Regan and Pikoli, Towards a safer Khayelitsha, xxxv. Lack of trust in the police is a general problem in South Africa. The 2007 Human Sciences Research Council (HSRC) South African Social Attitudes Survey found that 60% of citizens do not trust the police. Human Science Research Council, Between trust and scepticism: public confidence in institutions, http://www.hsrc.ac.za/ HSRC_Review_ Article-85.phtml (accessed 1 September 2015); A Faull, Corruption in the South African Police Service, ISS, Paper 226, November 2011. More recently a national survey conducted by Futurefact revealed that ‘three quarters of South Africans believe that a lot of police are criminals themselves’ and 33% said that they were ‘scared of the police’. See Futurefact, Futurefact finds: Three quarters of South Africans believe that a lot of police are criminals, 3 February 2015, http://www.futurefact. co.za/futurefact-finds/futurefact-finds-three-quarters-south- africans-believe-lot-police-are-criminals (accessed 10 February 2016). 20 For a methodological reflection on focus groups based on research in South Africa see I Macun & D Posel, Focus groups: a South African experience and a methodological reflection, African Sociological Review, 2:1, 1998, 114–135. 21 TL Albrecht et al., Understanding communication processes in focus groups, in D Morgan (ed.), Successful focus groups, London and New York: Sage, 1993, 54, quoted in ibid., 119. 22 DL Morgan and RA Krueger, When to use focus groups and why, in DL Morgan (ed.), Successful focus groups: advancing the state of the art, Newbury Park: Sage, 1993, 3–19. 23 A weakness of focus groups is that participants can collude or collaborate to silence a particular issue. For a thorough discussion of the advantages and disadvantages of focus groups, see S Wilkinson, Focus groups in feminist research: power, interaction and the co-construction of meaning, Women’s Studies International Forum, 21:1, 1998, 114. 24 Community police forums (CPFs) involve partnerships between representatives of the community and local police. The purpose of the forums is also to allow for input from the community and promote better communication and relationships between the community and the police. In Khayelitsha there are CPFs at the police stations in Site B, Harare and Lingelethu West. See Safe Khayelitsha, Community police forums and sub-forums, http:// safekhayelitsha.org.za/local-safety-structures/community- police-forums-and-sub-forums/ (accessed 1 February 2016). For a critical discussion of the CPFs and other measures to promote community-centred strategies in Khayelitsha, see L Freeman and C McDonald, Mapping Khayelitsha: the complexity of everyday policing in a high crime area, SA Crime Quarterly, 53, 2015. 25 Thematic analysis enables the identification and analysis of themes that speak to something important and relevant to the study. See V Braun and V Clarke, Using thematic analysis in psychology, Research in Psychology, 3, 2006, 77–101. 26 Following the commission’s recommendations, eight sub- forums were established to deal with specific crimes and thereby bridge certain ‘policing gaps’. One of these sub- forums deals with gender-based violence and has, according to one of its members, defined policing of domestic violence as a key issue. Personal communication, Khayelitsha, April 2015. For more information on the police sub-forums see Freeman and McDonald, Mapping Khayelitsha, 27–28. InstItute for securIty studIes & unIversIty of cape town16 27 For an overview, see Seekings, Economy, society and municipal services in Khayelitsha. 28 For more information see Social Justice Coalition (SJC), About us, http://www.sjc.org.za/about-us (accessed 25 May 2015). Purposive sampling is a non-probability sampling method that allows for the selection of participants who fit into the particular research. See R Schutt, Investigating the social world: the process and practice of research, New York: Sage Publication, 2015, ch. 5. 29 Macun and Posel, Focus groups: a South African experience, 4–5. 30 Ibid. As noted by Macun and Posel, homogeneity is also important to ensure that the members of the group ‘feel as comfortable and uninhibited with each other as possible’. 31 As Seekings notes, Khayelitsha remains a largely ‘immigrant community’. According to the 2011 census, most adults living there (69%) were born in the Eastern Cape. See Seekings, Economy, society and municipal services in Khayelitsha, 12. 32 According to the 2011 census, 46% of Khayelitsha’s population live in informal settlements, and the majority of residents live in shacks. Although access to public services has expanded steadily, a significant minority of residents still rely on communal, generally unsatisfactory facilities, including inadequate sanitation arrangements. See Seekings, Economy, society and municipal services in Khayelitsha. 33 During the commission’s hearings repeated complaints were raised about the failure of the Khayelitsha SAPS to patrol in informal neighbourhoods. See O’Regan and Pikoli, Towards a safer Khayelitsha, ch. 6, 128. For a discussion of the particular challenges of policing in informal settlements see the testimony of Major General D Molo, ibid., 252. 34 According to an SJC representative, the organisation has approximately 2 000 members in 12 branches in Khayelitsha. On average, 10–30 members attend branch meetings while the remaining members attend only larger events. The majority of SJC members are unemployed and live in informal settlements. Personal communication, Khayelitsha, 21 April 2015. 35 In terms of ‘racial classification’, one of us is white and European, one coloured South African and one white South African. 36 See, for example, J Flavin and A Desautels, Feminism and crime, in C Renzetti, L Goodstein and S Miller (eds), Rethinking gender, crime and justice, New York: Oxford University Press, 2006, 11–28. 37 For a discussion of the use of a translator in qualitative research see, for example, A Young, Qualitative research and translation dilemmas, Qualitative Research, 4:2, 2004, 161–178. 38 For example, see Albertyn et al., Women’s freedom and security; M Hunter, The master’s tools revisited: can law contribute to ending violence against women?, IDS Bulletin, 37:57, 2006; R Jewkes, J Levin and L Penn-Kenkana, Risk factors for domestic violence: findings from a South African cross-sectional study, Soc Sci Med., 55:9, 2002, 1603– 1617; J Goldscheid, The parallel processes of law and social change: gender violence and work in the United States and South Africa, Feminist perspective on international justice: from international and criminal to alternative forms of justice, Intersentia, 2013, 313; Penelope Andrews, Violence against women in South Africa: the role of culture and the limitations of the law, Temple Political and Civil Rights Law Review, 8, 1998, 425–515. 39 Notably, the man who suggested this was a member of a street committee (a community structure that works to mediate conflicts and promote safety in communities). According to Freeman and McDonald, street committees in Khayelitsha operate largely outside the remit of the SAPS and are seen as ‘belonging’ to the South African National Civics Organisation (SANCO). See Freeman and McDonald, Mapping Khayelitsha, 27–37. In the focus group his suggestion received no support from other participants. Due to the relatively small sample, we cannot exclude the possibility that other community members would be more appreciative of and willing to engage in local and informal dispute resolution mechanisms, as was found by Artz and Smythe, Bridges and barriers. 40 Another man in the same focus group explained that his closest kin were not in Cape Town, so the right thing to do would be to find somebody from his extended family or clan to step in and ‘help solve the problem’. Focus group interview, male group, Khayelitsha, 24 April 2015. 41 See also R Jewkes, Intimate partner violence: causes and prevention, Lancet, 359:9315, 2002, 1423–1429. 42 Personal communication, Khayelitsha, 17 April 2015. Notably, the argument that ‘my mother accepted it so I should too’ was also mentioned by a woman in our focus groups. According to one counsellor we interviewed, this statement was a typical sign of what she described as an ‘old mentality’. Personal communication, Khayelitsha, 21 April 2015. 43 See also S Bollen et al., Violence against women in metropolitan South Africa: a study on impact and service delivery, ISS, Monograph 41, 1999. 44 One of the counsellors we interviewed confirmed that victims in counselling frequently explained their partner’s abuse as acts of love or care, but hypothesised that it was often an excuse women made for themselves to stay in the relationship. Personal communication, Khayelitsha, 15 April 2015. 45 In the South African context, studies have demonstrated that women’s reasons for non-reporting are influenced by the economic and violent sanctions that involving the police are anticipated to bring. See L Artz, Fear or failure? Why victims of domestic violence retract from the criminal justice process, SA Crime Quarterly, 37, 2011, 3–10; Jewkes et al., Risk factors for domestic violence; L Vetten et al., Implementing the Domestic Violence Act in Acornhoek, Mpumalanga, Tshwaranang, Research Brief 2, 2009; L Vetten et al., Micro-study of the DVA: implementation of the Domestic Violence Act at nine South African courts, Report prepared for the Department of Justice and Constitutional Development, 2009. Further rationalising victims’ decisions to not involve the police in cases of domestic abuse, studies have also demonstrated that police interventions into these matters are temporary and inconclusive at best, and might often exaggerate the violence. See, for example, A Altbeker, Policing domestic violence: the enthusiasm gap, SA Crime Quarterly, 12, 2005, 13–18; J Steinberg, Thin blue line, Jeppestown: Jonathan Ball Publishers, 2008. 46 Focus groups interview, male group, Khayelitsha, 25 September 2014. 17SA Crime QuArterly No. 56 • JuNe 2016 47 Focus group interview, female group, Khayelitsha, 25 September 2014. 48 Focus group interview, female group, Khayelitsha, 21 April 2015. 49 Personal communication, Khayelitsha, 15 April 2015. 50 Interestingly, it was suggested that lesbian women would experience similar forms of ridicule if they looked or dressed like men. 51 Focus group interview, male group, Khayelitsha, 24 April 2015. 52 Research conducted by Lillian Artz suggests that involving the police is not necessarily ‘pointless’ as female victims of abuse can use the system to negotiate security and non- violence. See Artz, Fear or failure. 53 Personal communication, Khayelitsha, 15 and 21 April 2015. 54 C Bicchieri, The grammar of society: the nature and dynamics of social norms, Cambridge: Cambridge University Press, 2006. 55 C Bicchieri and R Muldoon, Social norms, in Edward N Zalta (ed.), The Stanford encyclopedia of philosophy, Stanford: Stanford University Press, 2014. 56 See KA Kelly, Domestic violence and the politics of privacy, New York: Cornell University Press, 2003, 2. 57 Our male participants’ eagerness to challenge the image of men as violent and aggressive may have been influenced by the fact that we were female researchers. 58 Jewkes, Intimate partner violence. 59 Women who are complicit in the unequal structuring of gender relations and at least tacitly accept their subordination, endorse or enact a traditional or emphasised femininity. See R Connell, Gender and power: society, the person and sexual politics, Palo Alto: University of California Press, 1987. Importantly, Connell underlines that these models of femininity should not be read as fixed categories as the relationships in which gender is constituted are dynamic. 60 See also Jewkes et al., Risk factor for domestic violence. 61 See also R Jewkes and R Morrell, Sexuality and the limits of agency among South African teenage women: theorising femininities and their connections to HIV risk practices, Soc Sci Med., 74:11, 2012, 1729–1737; TE Barlow et al., The modern girl around the world: a research agenda and preliminary findings, Gender & History, 17, 2005, 245–294. 62 C Gilligan, In a different voice: psychological theory and women’s development, Cambridge, MA: Harvard University Press, 1982, quoted in S Meyer, Why women stay: a theoretical examination of rational choice and moral reasoning in the context of intimate partner violence, Australian & New Zealand Journal of Criminology, 2, 2012, 181. 63 Meyer, Why women stay, 181. 64 For similar findings, see S Bollen et al., Violence against women in metropolitan South Africa: a study on impact and service delivery, ISS, 1999; Human Rights Watch, Violence against women in South Africa: state response to domestic violence and rape, New York: Human Rights Watch, 1995. 65 For the commission’s recommendations pertaining to domestic violence see O’Regan and Pikoli, Towards a safer Khayelitsha, 458–459. 66 H Combrinck and L Wakefield, Going the extra mile: police training on domestic violence, SA Crime Quarterly, 31, 2010, 27–34.