FEAR IN THE SUBURBS

The policing sector I studied lay immediately north-west of Durban's city centre. The centre itself is a bustling hub of formal and informal trade that Berea residents tended to avoid in favour of the growing range of malls constructed to serve Durban's elite, many of whom were moving north along the coastline into exclusive, secured spaces.

The neighbourhood explored below remained one of the more elite areas around the original central business district. Those I spoke with in Berea were white, middle- to upper-class residents who lived in homes ranging from flats in affordable blocks for young professionals to sprawling old houses replete with pools and smaller outhouses where live-in domestic workers used to live (and sometimes still did). The sector's more affluent residents could have afforded homes in elite gated estates such as Mount Edgecombe (Durrington Reference Durrington2009), but they were drawn by the sector's schools; its lively streets with restaurants, cafés and galleries; and its proximity to the beach. In their minds, Mount Edgecombe was elite, but isolated. In contrast, they felt more engaged in the world around them and yet still firmly within the elite suburbs of the city. Thus, when Simon, a resident in his forties, spoke of being ‘part of the world’, the ‘world’ he spoke of was, he clarified, ‘the suburbs’.

The areas surrounding this policing sector have seen substantial racial desegregation since 1994 and have become accessible to lower socio-economic groups. This sector, however, had seen less change, although there had still been notable shifts in the race and class of the area's residents. Immediately after the transition, several people recalled a ‘fire sale’ of houses, as residents moved north along the coast and out of the city, or out of the country.Footnote ² This gave many younger residents I interviewed the opportunity to acquire substantial properties. It had also provided Indian and black residents of all ages with a chance to buy property in this previously white suburb. Explicit discussions of racial shifts were, however, rare. In fact, any direct discussion of race among white residents in Berea was limited, particularly in interviews.

It was not just white South Africans who avoided talk of race and racism during interviews (Erasmus Reference Erasmus2005). However, given the ‘racial structure’ of apartheid South Africa (Bonilla-Silva Reference Bonilla-Silva2006), white residents appeared to be particularly unsettled by direct discussions of race, which implicated them in questions of privilege or prejudice (Durrheim et al. Reference Durrheim, Mtose and Brown2011). Their avoidance and suppression implicitly recognized that, while race is neither an immutable nor a monolithic category of identification, it remained intimately entangled with everyday life in undeniable, uncomfortable, and ambiguous ways (Steyn Reference Steyn2001; Blaser Reference Blaser2007). Consequently, residents generally spoke in opaque, pejorative terms of demographic change ‘creeping’ towards the borders of their neighbourhood.Footnote ³ It was in everyday conversations, as opposed to interviews, that more explicit opinions and anxieties emerged, hovering at the intersection of class and race.

Despite ongoing references to ‘creeping change’, the housing ‘fire sale’ had dissipated in recent years – as had, it seemed, these residents' fear of crime. Many spoke of a time, after President Thabo Mbeki had ‘denied’ the problem of crime in 2004, when they felt that their fear of it had peaked. A significant number of people's friends and family had emigrated during Mbeki's presidency, and the talk at braais (barbeques), at the school gate, and over the dinner table was reportedly dominated by discussions of crime. More recently, however, residents claimed that this talk had declined, as people had acclimatized to life in post-apartheid South Africa. ‘Every conversation you used to have in this country was about crime,’ explained James, a male resident in his forties, ‘[but] this is one of the first conversations I have had about crime – and it is such a relief – in a long time.’

However, the fear of crime and a preoccupation with avoiding crime had not disappeared. On the one hand, they had become normalized in practices of everyday life, and, on the other, they had been actively repressed. Authors have highlighted how crime is a politically correct way to talk about racialized fears and apartheid nostalgia (see, for example, Judd Reference Judd and Dunn1994). However, the degree to which ‘crime talk’ (Sasson Reference Sasson1995) could act as a shield for such sentiments has arguably waned since Mbeki and others dismissed those afraid of crime as ‘reactionaries’ and ‘racist’. Crime has remained less taboo than outright apartheid nostalgia (Hansen Reference Hansen2012), but there was a sense among these white residents that to ‘obsess’Footnote ⁴ about crime was to resist fully integrating oneself into post-apartheid South Africa. Many denied the hold that such fears had over them, and yet their words and deeds showed how such fears had become part of their daily lives. As Sophie, a resident in her twenties, explained:

What people's fears focused on, primarily, was a criminal archetype that was young, poor, black, male and violent.Footnote ⁵ Two white residents I interviewed had been victims of robbery,Footnote ⁶ having been held at gunpoint by black males, but not injured. The remainder had been subject to petty theft or house break-ins in their absence and did not know, with any certainty, who had committed the crime. Everyone, however, believed that they were potential victims of violent crime and shared similar criminal archetypes, recounting incidents among friends or in the local press as evidence that their fears were founded. The ‘objectivity’ of such fears is difficult to measure and, in many contexts, not the most important analytical issue at stake (Douglas and Wildavsky Reference Douglas and Wildavsky1983). Nonetheless, we can tentatively gauge the relative level of crime that Berea residents faced by exploring the differences in recorded crime between there and KwaMashu, a comparatively deprived former black township.Footnote ⁷ In 2011, there were six recorded murders and seventy-three recorded acts of common robbery, compared with seventy-eight recorded murders and 112 acts of common robbery in KwaMashu.Footnote ⁸ In other words, the threat of violent crime was notable but, nationally speaking, comparatively low: statistics only partly explain why residents such as Ben, a middle-aged interviewee, stated: ‘I think we are just brought up with this looking over your shoulder type of attitude.’

White residents in Berea were, to borrow from Vincent Crapanzano's (Reference Crapanzano1985: 45) analysis, ‘waiting’. Focusing on Wyndal in the northern Cape in the 1980s, Crapanzano studied white residents who were in privileged positions of power and ‘paradoxically, in their privilege, victims of it’ (Crapanzano Reference Crapanzano1985: xiii). These individuals, he claimed, could not conceive of an alternative ordering of society to that of apartheid but knew that the current order of society could not last. Thus, Crapanzano argued, they were found ‘waiting’ – a state ‘marked by contingency – the perhaps – and all the anxiety that comes with the experience of contingency’ (ibid.).

Clearly, too much has changed since Crapanzano's study to apply his findings, in all their complexity, to this case. However, in his article on gated estates in Durban, Matthew Durrington (Reference Durrington2009) asked whether these affluent residents in post-apartheid South Africa were also ‘waiting’, and, if so, waiting for what. Ultimately, he left the question open. In Berea, the concept of ‘waiting’ is useful for highlighting how elite, white South Africans, while presenting themselves as ‘part of the world’, realized (at some level) the incongruence of their livelihoods with current rates of unemployment and economic deprivation in the historical context of apartheid. For residents, violent crime posed the most immediate threat to this life but fears of broader social, economic and political change also exacerbated the ‘ontological insecurity’ (Giddens Reference Giddens1991) at the heart of ‘waiting’ (Ballard Reference Ballard, Distiller and Steyn2004; Lemanski Reference Lemanski2006).

One of the most obvious manifestations of ‘waiting’ was that residents became hyper-vigilant, scanning the streets for would-be attackers. In keeping with the discussions on race above, many implied rather than described the characteristics of ‘suspicious’ characters in their area. Alistair, for example, was a resident in his forties, who claimed that he could spot suspicious characters because he knew his neighbours: ‘You will know who is not meant to be in the area before he does something … All of a sudden there is someone hanging around and you will know, “Hang on, that is not a maid or a worker,” you will say, “Guys, let's investigate.”’Footnote ⁹ Alistair's initial reference point was his neighbours (who were predominantly Indian or white and relatively affluent) but his concerns rested on those who looked like ‘maids and workers’ – that is, those who were predominantly black and relatively poor – but were not employed on the street. These figures, he concluded, were ‘not meant to be in the area’: having no obvious connections to residents in the neighbourhood, they were assumed to have malevolent intentions. They were ‘matter out of place’ and, therefore, dangerous (Douglas Reference Douglas2003).

Within white residents' imaginaries of crime, such figures were ever-present in the suburbs, looking for the chance to commit an opportunistic crime or monitoring people's homes to ascertain their daily routines and plan crimes around them. As Rebecca, a resident in her sixties, argued: ‘They watch your moves, they know exactly when you go out, which house has nobody in attendance during the day … They sit. And, if there is that facility where they can just sit and watch innocently, that is where they will hit.’ The criminal gaze Rebecca and other white Berea residents described was all-encompassing. Like prisoners in Jeremy Bentham's panopticon, who were rendered visible to guards that they themselves could not see, residents felt that they ‘must assume ubiquitous surveillance [and] that they may be under inspection at any time, day or night’ (Elmer Reference Elmer, Ball, Haggerty and Lyon2012: 23). However, whereas Bentham's watch guards were always invisible, suspicious characters were deemed to be both a visible and invisible presence.

Understanding residents' fears, and how these were entangled in their broader sense of being and belonging in post-apartheid South Africa, provides us with a foundation from which to explore the everyday policing strategies that emerged when residents became victims of crime.

A CONVERSATION WITH GEOFF AND RACHELFootnote ¹⁰

Geoff and Rachel are middle-aged white South Africans who live in a detached house on a relatively quiet road in Berea. As is the case for most residents in the area, Geoff and Rachel's house is protected with a gate, fence and armed security response, and they also have household insurance for their possessions.

In December 2010, family members visited Geoff and Rachel to celebrate Christmas. One morning, while the family was sleeping, two men broke into the house and began stealing their possessions. When the thieves disturbed Rachel, she woke Geoff. Realizing that the occupants were awake, one man jumped over the back fence while the other hid under the dining-room table. As Geoff recounted: ‘I accosted him with a chair … Basically we arrested him … He did put up a fight but at that point he was under the table and I laid into him with a chair … There was blood all around the place.’

Having placed the thief under citizens' arrest, Geoff phoned his security company. Hearing screams from the house, his neighbours shouted to the family, offering their help. Geoff instructed them to stay at home and call their security company. Acceding to his request, the neighbours phoned both their private security company and the police. It was their private security firm that responded first, followed by the police, while Geoff and Rachel's private security company went to the wrong address.

Entering the property, their neighbours' private security guards took over from Geoff – who was, by that point, pinning the intruder under the table with a chair – and handcuffed the suspect. Before the police arrived to relieve the private security guards, Rachel recalled that the suspect ‘proceeded to tell us that 10 Rand of the money in his wallet was his, and that that was his cell phone so we duly gave him back his 10 Rand and his cell phone. One has to be extremely careful to remain exactly on the right side of the law.’

While Geoff and Rachel both recalled that the police responded relatively promptly, they were frustrated with their apparent inability to write a statement. ‘[I] can accept that … the police are not going to be as effective as I would like to protect me, so I pay the security company,’ Geoff argued, ‘but when I do need the police to be involved, I need them to be competent to take the statement … And the person cannot take dictation.’

After recounting his story to me, Geoff reflected: ‘Although … I do not see myself as a vigilante, I did attack and arrest this guy who was in my house.’ In that way, he claimed he was similar to other neighbours who had been victims of crime. Nonetheless, Geoff argued that people rarely used violence against criminals in Berea because the constitution ‘protect[ed] victims and perpetrators … you darenʼt take the law into your own hands because you are going to end up in Westville prison’.Footnote ¹¹

In this instance, the police already had their suspect for the attempted robbery. Nonetheless, Geoff and Rachel reported that there were difficulties securing a prosecution. The docket that included the suspect's past crimes was lost (mistakenly or otherwise) and so, while he should have been sentenced to five years, two years of his sentence were suspended. Geoff and Rachel believed that a conviction had been secured not because of the police but in spite of them. Thus, several months later, when a guest staying at their house had his mobile phone stolen from a jacket in their back room, the couple encouraged him to report the case in order to obtain a case number for his insurance claim. As Geoff explained: ‘it was just really reported for the purpose of insurance. I think there is really no expectation [when] something is stolen that the police will actually recover it.’

Following the robbery, Geoff and Rachel changed their security provider. They also attended a meeting of their local Community Police Forum (CPF), having heard of its existence through a flier. These forums are interfaces between the police and local residents (the Berea CPF is discussed in more detail below). While Geoff and Rachel found some of the information gleaned at the meeting interesting, it was not useful enough to make them regular attendees. They also resisted the CPF's attempts to ‘educate’ citizens on crime avoidance:

‘I WENT THROUGH THE MOTIONS’: PROCEDURAL ENGAGEMENT WITH THE POLICE

Geoff had low expectations of the police that were only confirmed by his interaction with them. As noted above, he believed his intruder's conviction was secured despite the actions the police took. Research by the Law Commission confirms that his case was highly unusual in securing any conviction at all. In their longitudinal study following cases committed in 1998, only 3.57 per cent of robberies with aggravated circumstances committed in Durban ended in a conviction; 88.37 per cent never made it to court (SALC 2000: 29).Footnote ¹²

Geoff's criticisms focused on the bureaucratic behaviour of the police: he had sought to engage the South African Police Service (SAPS) in their administrative capacity, not their protective or investigative capacity. He believed that the police were useful for opening cases, writing statements, issuing case numbers, and passing dockets to the courts. In one sense, this is a remnant function: the bureaucratic remainder of a wider, more active role for the police. However, record keeping is also a key ‘language of stateness’ in and of itself (Hansen and Stepputat Reference Hansen and Stepputat2001: 21), as the country's obsession with crime statistics demonstrates (see Comaroff and Comaroff Reference Comaroff and Comaroff2006). While the state provided services from refuse removal to road maintenance that made Geoff's daily life possible, he felt that his most tangible, personal engagement with the state was driven by his need to access the definitive, official record of events, which the state still held. ‘Most often one's interaction with the state is administrative,’ he reflected. ‘We get our driver's licence, we get our passport. That is when it feels like you are engaging with the state.’Footnote ¹³ Ultimately, for Geoff, the police had failed to behave in a state-like fashion because they were administratively incompetent. He might have avoided the police entirely at this point had he not needed their case number for his insurance claims, to validate his claims to victimization (Cooper-Knock and Owen Reference Cooper-Knock and Owen2015).

Many residents I spoke with echoed Geoff's experience. In the words of Lawrence, a middle-aged Berea resident, they ‘went through the motions’ of reporting to the police purely to access their household insurance; they had ‘no expectation’ of seeing a successful investigation or conviction.Footnote ¹⁴ This drives us to look more closely at the diverse ways in which the expanding private industry around crime affects connections between citizens and the state. Certainly, the rise of private security per se and the growth of gated estates with their own internal, privatized governance have loosened the bonds between privately ensconced citizens and the state (Ballard Reference Ballard, Distiller and Steyn2004; Durrington Reference Durrington2006; Reference Durrington2009; Hook and Vrdoljak Reference Hook and Vrdoljak2002). However, those who had private insurance were locked into contact with the state even as they retreated from it: they were forced to ‘go through the motions’ of being a citizen.

In this policing sector, however, contact between residents and the police was fraught with ambiguities: were the police sometimes unable to perform their duties due to a lack of human and physical resources? Or were they unwilling to do so due to laziness, corruption or racialized resentment? It was rare for residents to definitively know what drove police success or failure in any one instance. The rationalizations they made, however, indicate how they ‘read’ the police as an institution, and the state more generally. Geoff and Rachel believed that police incompetence was tied to poor education and high levels of corruption. Such critiques often made implicit and explicit reference to the class differences between these affluent citizens and local police officers. Others emphasized the importance of race in shaping poor performance: black police, they argued, would seek to protect black suspects, ‘go slow’ when serving non-black complainants, and demand bribes from, or prosecute, non-black citizens where possible.Footnote ¹⁵ For several Berea residents, this apparent racialization of state services was understood as retribution for apartheid (see also Burger Allen Reference Burger Allen2002). Lorna, for example, was a resident in her fifties who had been brought up in the area and considered herself to be politically liberal. Like Lawrence and Geoff, Lorna reported crimes to the police for insurance purposes only. She argued:

Thus, residents' need to gain a case number for their insurance may have cemented the police's bureaucratic authority and secured their ongoing contact with the state. However, as Allport's (Reference Allport1954) ‘contact hypothesis’ highlights, contact itself can either exacerbate or allay prejudice and tension: what matters is the context in which this contact takes place. For residents in Berea, contact with the police often reinforced their sense of frustration with them, or alienation from them (Loader Reference Loader2006).

EMPLOYEES IN BEREA: BETWEEN PREDATOR AND PROTECTOR

Like Geoff and Rachel, most residents I spoke with in Berea lived in ‘fortified cells’ (Davis Reference Davis and Sorkin1992: 156), surrounded by defensive architecture. Many also employed an armed security response service or a permanently stationed guard. Residential private security in South Africa has its roots in the late 1970s and early 1980s, when commercial provision was encouraged by an increasingly beleaguered apartheid state (Grant Reference Grant1989; Baker Reference Baker2002).Footnote ¹⁸ However, it was after apartheid that the industry mushroomed (Schonteich Reference Schonteich1999). In recent years, this rapid expansion has continued: the number of security businesses increased by 18.35 per cent between 2010 and 2011, and registered security officers increased by 5.13 per cent in the same period (PSIRA 2011). Berea residents argued that a fear of crime drove their substantial investment in private security. However, as argued above, crime is a ‘polyvalent’ category (Buur Reference Buur2003) and contained within this stated fear were broader feelings of exposure, danger and dislocation that had accompanied South Africa's new dispensation (Ballard Reference Ballard, Distiller and Steyn2004; Lemanski Reference Lemanski2006).Footnote ¹⁹

Although the different constellations of private security employed in this sector had varying effects on those living in and around them (Durrington Reference Durrington2009), they all functioned on the basis of exclusion. Where these private security measures spilled into public spaces, those who fitted South Africa's dominant criminal archetype (i.e. young, poor, black and male) found their constitutional right to freedom of movement downgraded to a negotiated privilege. Guards described how they prevented anyone without a ‘purpose’ or ‘reason’ from entering or remaining in a patrolled area.Footnote ²⁰ In practice, this amounted to the exclusion of all those who lacked pre-existing ties of sociality or service to local residents: ‘purpose’ and ‘reason’ were ultimately not self-defined terms. Where people were suspected of a crime, or resisted removal, their right to ‘security of the body’ also became negotiable.Footnote ²¹ A comprehensive literature has documented this erosion of substantive citizenship (for example, Berg Reference Berg2010; Samara Reference Samara2003). Here, I pursue a different analytical track by analysing employer–employee tensions and their consequences for everyday policing. Starting with private security guards and then moving to domestic workers, I argue that, framed by race and class, employers saw their employees as potential objects and agents of policing.

Private security guards occupied a paradoxical space in the imaginations of Berea residents. On the one hand, private security was everything that the police were not: a spectacle of elite, technological professionalism.Footnote ²² As David, a middle-aged entrepreneur, stated: they ‘looked the part’.Footnote ²³ On the other hand, the groundwork of private security was usually performed by unskilled, low-paid labour. The demographic profile of private security guards differentiated them from their employers and overlapped (in whole or in part) with the criminal archetype they were employed to dispel. For many white residents in Berea, this difference produced a fear that hovered at the intersection of race and class, manufacturing distrust and complicating their reliance upon those they had employed (Burger Allen Reference Burger Allen2002; Dirsuweit Reference Dirsuweit2007; Durrington Reference Durrington2009). While residents paid for private security operators, some feared that the allegiances of the guards lay elsewhere. Highlighting the role of race in these fears, a manager from one of Durban's main private security firms stated:

Thus, private security was a terrain on which racial and class-based prejudices and fears were iterated rather than escaped. The act of commercialization itself also exacerbated residents’ fears. Private security was seen as a ‘big business’ whose market relied upon the continuation of the very crime and insecurity it promised to fight, producing perverse incentives.Footnote ²⁵

Consequently, suspicion and muttered accusations permeated exchanges with security personnel. Because any evidence linking private security guards to crimes was often circumstantial at best, these concerns were not taken to the police. Rather, residents either chose to continue this suspicious reliance, making it one of many daily relationships they maintained that were quietly fraught with ‘race trouble’ (Durrheim et al. Reference Durrheim, Mtose and Brown2011) and class-based anxiety. Alternatively, they terminated their contract. Lorna, for example, chose the former route. During her fiftieth birthday celebrations, she had mistakenly triggered her burglar alarm, activating her armed response. The response guard, she claimed, had surveyed a roomful of open presents. When her house was robbed the following day, she held him responsible. The problem with security guards, she claimed, was systemic: ‘security companies would definitely be my question mark. Definitely. Definitely. I think that a lot of what happens is through the security companies.’Footnote ²⁶ Subsequently, Lorna maintained her contract with the same company but lessened her reliance on these employees. As she explained, ‘I never put [the alarm] on. I never put it on. My dogs are better than anything.’Footnote ²⁷

Distrust was not the only reason, however, why private security could become little more than a facade. John, a resident in his sixties, had an armed response contract, ‘burglar guards’Footnote ²⁸ on his windows and doors, and an electric fence, but never activated the alarm or the fence because he found them inconvenient.Footnote ²⁹ Instead, as I explore in more depth below, he preferred to try to negotiate with anyone who broke into his property. What Lorna's and John's cases both demonstrate is the importance of looking beyond a private security aesthetic to uncover the lived complexity of everyday policing choices. Residents in Berea, it seems, were capable of ‘going through the motions’ with both private and state providers of policing.

Private security company staff were not the only employees who operated in and around the residences of Berea: most households also hired domestic workers. The latter were not primarily engaged to provide policing but they were drawn into everyday policing in two senses. Firstly, to an even greater degree than private security guards, domestic workers knew the intimate spaces of people's homes. In the midst of their daily work routines and their personal lives, they were assumed to know of potential opportunities for crime and potential criminal threats. Consequently, residents believed that they could act as informers for their employers or for local criminals. As a result, residents were keen to ‘recruit’ domestic workers to assist in everyday policing. This logic underlay the establishment of ‘maid watch’ by the local neighbourhood watch, who organized ad hoc meetings between themselves, local domestic workers and the police to ‘educate’ and ‘train’ these employees. This was on the understanding that, ‘when you go off to work, [domestic workers] are there at your house. So they notice what happens in the day. They sit on the road. To get their buy-in and support [is] to know what happens.’Footnote ³⁰ Conversely, the same logic also triggered suspicions that, when a house break-in occurred, domestic workers had provided ‘inside information’ to the criminals in question although evidence supporting these accusations was circumstantial at best.Footnote ³¹

The second sense in which domestic workers were drawn into everyday policing was as direct suspects of theft. Sometimes evidence of this was, again, circumstantial at best. However, a few residents had more solid proof that their employees had stolen from them on a small scale, taking things they ‘knew we would not miss’, from boxes of cereal to fur coats.Footnote ³² Residents were often conflicted about policing this level of crime given the context of domestic work in South Africa. As Cock's (Reference Cock1980) study demonstrates, relationships between white residents and their domestic workers were among the most exploitative working relationships in apartheid South Africa. Since 1994, this form of employment has continued to be economically unrewarding and fraught with racial tensions that are rarely vocalized but often present (Durrheim et al. Reference Durrheim, Mtose and Brown2011: 182; Ally Reference Ally2011).

This wider context complicated the framing and tackling of petty theft. Across Berea, residents debated the causes of the country's crime, which were variously understood in psychological, racial and economic terms, being fed by the country's vast inequality. In such conversations, economic inequalities could be expressed in general, abstract terms, which lessened any sense of responsibility these affluent residents might feel for the current status quo. However, residents were completely responsible for setting the pay and conditions for their domestic workers, and petty theft within the home could be read as a reaction to the poor wages they offered. As Phumzile, a domestic worker, claimed: ‘It is good to be satisfied with your salary, if you are not satisfied [you] … will think I will rather pay myself … I will come back later and take that [item from the house].’Footnote ³³ Employers across the sectors I studied in Durban, of different races, acknowledged this connection between meagre wages, poor working conditions and petty theft, arguing that ‘people who have no intention to steal would steal if you are rude and disrespectful … and paid [if they are] a pittance’.Footnote ³⁴

This framing made residents less likely to tackle petty theft directly. Sally, for example, was a resident in her thirties who had recently realized that her maid, Buhle, was taking food from her cupboards. Defending herself as an employer, she insisted that Buhle was paid reasonably and sent home with ‘ridiculous amounts [of food] each week’. Nonetheless, she accepted that Buhle's wages were still insignificant in comparison to her income and expenditure, and accepted theft as a morally justifiable form of informal redistribution:

In summary, in this section we have seen the ways in which the question of difference between employers and employees shaped everyday policing in Berea. This might make employees more or less likely to be relied upon for policing, and more or less likely to be policed, as the example of petty theft among domestic workers demonstrates. The broader conclusion emerging here is that a distrust of state policing does not equate to a trust in policing alternatives.

‘THERE IS NO COMMUNITY SHARING’: THE INFORMAL COLLECTIVIZATION OF PRIVATE SECURITY

When Geoff and Rachel were in trouble, their neighbours called their own private security company to come to the scene. This could be termed an act of ‘informal collectivization’.Footnote ³⁶ Although many residents I spoke with would agree with Peter, the local neighbourhood watch chairperson, that ‘there is no community sharing’ of security measures in Berea, there had been several acts of informal collectivization both in this sector and elsewhere.Footnote ³⁷ In his study on policing in Africa, Bruce Baker (Reference Baker2002: 31) argues that the terms ‘“public” and “private” get confused’ when public officials act outside the collective public interest. Likewise, in Berea, the individuated logic of private security was blurred when it was made accessible to those outside the strictly contractual relationships involved.

There were various ways in which private security could be collectivized that, in turn, required varying degrees of consent and proactivity from those paying for and delivering private services. At a minimum, informal collectivization emerged as a classic free-rider problem. For example, following a brutal murder on their street, Sylvia had convinced a large proportion of residents to contribute to a twenty-four-hour unarmed guard service. While these guards physically responded to individual incidents, their key power, in Foucault's words, was ‘based not on physical coercion but on a more subtle and pervasive form of coercion that draws its power from surveillance’ (in Berg Reference Berg2010: 289). Although physical reactions were alienable to contributors, this surveillance power was not, creating quiet tensions between the ‘bad ‘uns’ who were free-loading and ‘good ‘uns’ who were not.Footnote ³⁸

In many cases, however, informal collectivization was more complicated. Tom and Sarah, for example, lived at the tail end of a T-junction in Berea. The road that intersected with theirs had hired a private security guard to watch the street. Tom and Sarah did not contribute to the service (approximately £110 per month) because their house was not visible from the guard hut. The couple had two young children and employed Cindy, a live-in maid, to help with childcare and the upkeep of the house. Innocent, the private security guard who worked the day shift, often came to speak with Cindy in his breaks and during his patrols.Footnote ³⁹ One afternoon, when Innocent was looking for Cindy, he saw two men trying to lift Tom and Sarah's security gate off its hinges. He approached their property to confront the men, phoning Cindy as he did so. She called Tom and their armed response firm. Realizing that they had been seen, the suspects escaped.Footnote ⁴⁰

Nancy, who lived two doors down from Tom and Sarah, contributed to Innocent's wages. The two households were acquainted, largely because Nancy was civically active and worked to raise awareness about crime in the area, sending email alerts to them and other residents. When I spoke to Nancy, she was ambivalent about the incident. At one level, Tom was part of her social–civic network, so she was glad to help. Moreover, the incident had reduced crime in the area, which she determinedly supported. Ultimately, however, she was ‘really irritated’ by the informal collectivization of Innocent's services. Paying for a private security guard was no small financial commitment for Nancy.Footnote ⁴¹ This, then, was a case of contested collectivization. Significantly, Nancy had not mediated between her neighbour and her private security provider: as Innocent was permanently posted on their road, he had formed his own social networks, which had shaped his actions in this instance. This contrasts with Geoff and Rachel's case, where a subscriber played a gate-keeping role to private security services.

Thus, informal collectivization was a highly contingent act. It did not spring from any shared vision of alternative citizenship or avid communitarianism.Footnote ⁴² In this sense, we are seeing a fundamentally unstable act that was subject to contestation: while free-riding in its most passive sense was difficult to eliminate, anything involving more proactive responses had to be offered or negotiated between subscribers, contractors and would-be recipients. Informal collectivization was the product of friendship and favour, not obligation. But, as such, these acts were revealing of the social networks that operated in Berea and became entangled with the private provision of policing.

‘IT WOULD NOT HAPPEN IN THIS SECTOR’: PARTICIPATION IN POLICING ORGANIZATIONS AND FORMATIONS

In their work on everyday policing, Buur and Jensen (Reference Buur and Jensen2004) distinguish between participation in ‘policing organizations’ and participation in less structured, more ephemeral, ‘policing formations’. Participation in formations and organizations in predominantly white areas has received little academic attention and was often overlooked by residents themselves.Footnote ⁴³ This was particularly true when participation involved violence: ‘taking the law into your own hands’, many argued, was something that happened elsewhere. Implicitly and explicitly, this ‘elsewhere’ was the country's black townships, which were still associated with the ‘mob violence’ and ‘kangaroo courts’ of apartheid (Buur Reference Buur2008a; Cooper-Knock Reference Cooper-Knock2014). In contrast, the white residents of Berea portrayed themselves as privileged, elitist and isolationist but not civically active, nor unlawful or violent. A typical portrayal of Berea was given by Carol, a self-employed resident in her forties who lived with her family in the sector:

Carol raises two insightful points here: first, many Berea residents could afford to buy their way out of civic responsibility and compensate privately for failures in state delivery. Second, hardly any residents regularly participated in the local CPF. However, this did not mean that policing was utterly outsourced, nor that residents lacked what might be termed a ‘civic reflex’.

There were two key policing organizations in Berea, which tried and failed to attract participation from residents. The first was a neighbourhood watch, which predated the transition and was somewhat side-lined by the advent of CPFs. Previously, its members had patrolled the area in cars, but now, as violent crime was perceived to have increased (along with the price of petrol), the patrols had ceased. Aside from issuing a quarterly newsletter for local residents, and running ‘maid watch’ sessions, the watch was largely inactive and had considered dissolution. Although local residents encouraged it to stay in operation, there was no increase in the watch's membership or activities and, as I left the field, its future seemed uncertain.Footnote ⁴⁵

The second body was the local CPF. As mentioned above, CPFs emerged in the transition from apartheid. They were initially envisaged as citizen–police interfaces at which security could be co-produced and police legitimacy restored (Brogden Reference Brogden2002; Reference Brogden2004; Dixon Reference Dixon2007; Gordon Reference Gordon2001; Reference Gordon2006). Policing approaches have since shifted towards state-centric models, but police stations remain legally obliged to establish and run these forums. Berea's CPF had a local reputation as a ‘talk shop’, or, in the words of one past executive member, ‘a melting pot of nothingness’.Footnote ⁴⁶ While the police representatives were amicable, they saw the forum as a space for receiving information, not direction, and for ‘educating’ residents; however, attendees had little useful information on crime. Usually, CPF attendance rose when a crime occurred (personally or locally) – it represented a reflex response to reach out to the state when the limitations of private security became apparent. However, like a reflex, this outreach was short-lived and CPF attendance dissipated just as fast, as citizens became disillusioned and resentful of police efforts to promote ‘responsibilization' (Garland Reference Garland2001) by the police, taking this as evidence of further state inaction.

Carol herself had been a CPF co-chair for almost a year, but had recently resigned when we spoke. Like many others, she had begun attending after falling victim to crime (in her case an armed house robbery). This crime had convinced her husband that they needed to move into the secured estate, Mount Edgecombe. Carol had joined the CPF to seek alternative strategies for everyday policing, so that her family might feel safe remaining in Berea. When her efforts to mobilize local residents and shape police performance failed, she resigned. The couple has since moved north to a secured golfing estate. Carol's story reminds us that, while security estates can be spaces in which residents retreat from civic engagement (Durrington Reference Durrington2009), we cannot assume that these residents have been on a long or inevitable path towards such disengagement. Nor was the failure of structured activities in relatively affluent, predominantly white areas inevitable. In similar areas across Durban, I interviewed groups that actively patrolled on a regular basis, with scores of active members. In each of these cases, patrols operated in addition to, rather than instead of, private security. The failure of these structures in Berea, then, is contingent on more than the variables of ‘culture’ or class.

Having explored policing organizations, let us turn to policing formations. A shared discourse existed on policing formations in Berea, which I will first explain and then critique. According to this public narrative, ‘taking the law into your own hands’ was something that happened elsewhere, namely in formerly black townships, which residents associated with the operation of the ‘sovereign mob’ (Buur Reference Buur2003). As David argued, ‘there is a different conditioning in this sector that would allow the law to take its course … the average man is going to kick a would-be thief … on the property whereas he would probably be kicked quite badly elsewhere’.Footnote ⁴⁷ This reticence emerged from residents’ unwillingness to ‘get their hands dirty’ – as we heard above – and their fear of being on the ‘wrong side’ of the law. Residents believed that, while the SAPS was failing to tackle crime, the police would successfully prosecute those who broke the law to tackle it themselves. In a few cases, these beliefs were bolstered by personal experience.

Nancy, for example, was a middle-aged business owner in Berea. Driving home one afternoon, Nancy saw her private security guard – Innocent – being attacked by two men who had been trying to steal a car from the street. Having ‘hooted’ to attract her neighbours' attention, she fled to her house to summon her armed response unit. When she returned, two male neighbours were ‘helping the guard’ by hitting the suspects with their golf clubs until they fled. When Nancy took Innocent to the station to open charges against the suspects, she was told she had insufficient evidence against them. Instead, Nancy was ‘horrified’ to learn that her neighbours were themselves facing charges. ‘The victim is the citizen … not the baddy,’ she concluded.

Undoubtedly, the SAPS did sometimes seek to prosecute vigilante citizens, and not all acts of policing by residents were violent, but the dominant discourse that Nancy's narrative supported did not reflect the complex and ambiguous reality of everyday policing in Berea. Other encounters suggested different possibilities. Graeme, for example, was a pastor at a local church. Like Nancy, he witnessed a public display of violence against a criminal suspect. The outcome, however, was very different. On his way into work one morning, Graeme had stopped at the traffic lights behind another male driver and a woman on her bike. As the traffic lights changed, a young, black male knocked the woman from her bike and attempted to cycle away, with Graeme and the other driver giving chase. Ultimately, Graeme's car blocked the would-be thief's escape as the other driver left his car, chased the suspect down, knocked him from the bike, and began to assault him. When a policeman arrived at the scene, he also gave the thief ‘a couple of slaps’. Graeme saw both responses as problematic but ‘understandable’ and did not believe that either would face further repercussions.Footnote ⁴⁸ The discrepancy between Nancy's and Graeme's stories reminds us of the inconsistent reactions of state actors to everyday policing. In interviews and informal conversations, several local police officers appeared quietly frustrated by the limitations of legal violence. In the words of one relatively senior white police officer: ‘I would like to see more sjamboks [whips] used, but I cannot say that out loud.’

Graeme's account mirrored the accounts of other interviewees, who reported assaulting suspects before the police arrived, in front of the police, or with the police, without legal consequence. So, why did this shared narrative persist? As well as providing a flattering self-image, grounded in tropes of race and class, there were several reasons for its resilience.

Firstly, when assaults did occur, they usually took place at the scene of the crime, inside people's private ‘fortresses’.Footnote ⁴⁹ Consequently, many acts of everyday policing could be framed as lawful: under South African law, citizens may ‘forcibly confine’ a suspect given ‘reasonable suspicion’ that they are committing, or have committed, a robbery.Footnote ⁵⁰ They may also use proportional force in self-defence. These laws are open to fairly broad interpretation. In both cases, there is space for ‘deviance allowed by the rules’ (Brogden and Nijhar Reference Brogden and Nijhar1998: 90) – behaviour that would be considered illegal were it not permitted by a legal loophole or a stretched legal definition. When does defence become aggression? When does confinement become assault? Most of the violence exercised by citizens against criminals occurred in this contested area. Geoff provides a good example of this ambiguity. Reflecting on his experiences, Geoff claimed that, while he would not consider himself a vigilante, his actions had borne the marks of vigilantism. The police, however, did not question his claim of self-defence.

Secondly, contrary to the experience of Nancy and Graeme, most violence was hidden from view inside people's houses. This invisibility could also strengthen people's negotiation with the police: there were no witnesses presenting and supporting competing interpretations of events or pushing for alternative outcomes (Cooper-Knock Reference Cooper-Knock2014). By the same logic, however, such instances of violence and negotiated impunity were not heard of by other residents, making them appear less prevalent and less possible. Instead, they were left reading of more publicly controversial cases in the media, where residents had been arrested.Footnote ⁵¹ This left residents such as Lawrence with the impression that, in Berea, the police ‘say you cannot take the law into your own hands. But the police in the townships they don't do anything.’ Only those with long-running relationships with local police, it seems, had a sense of the parameters of impunity they could negotiate.

Edmund, a middle-aged resident, was one such individual. He had recently arrived in Berea from another predominantly white suburb to the north of the city, where, he claimed, ‘the only reason that there was no crime … was that I was like a vigilante. If people come and cause trouble you just come with a paintball gun, shoot them with a paintball gun, beat them with a baseball bat, do whatever.’Footnote ⁵²

In contrast to the ‘apathetic’ CPF, Edmund claimed that he had taken action against the ‘vagrants’ who were the source of local crime. Edmund acted ‘like a vigilante’ in the absence of police action with no legal consequences, but he also recalled times when he had acted in close cooperation with them. Sometimes, his actions were reactive. Once, for example, Edmund had found two people trying to break into his car. Having shot ‘over their heads’ from the window, he called the police. The K9 (dog) unit arrived and together they cornered the suspects, at which point, Edmund claimed, ‘the cops left the dogs to bite them from [sic] a long time before they went and got them’.Footnote ⁵³

Elsewhere, he took a more proactive approach. For example, he conducted ‘raids’ on his local beach to ‘confront’ those who had built ‘shelters in the bushes’, explaining: ‘You have to be that aggressive and that frightening that that person will not come back.’Footnote ⁵⁴ Once Edmund and his colleagues had confronted the ‘vagrants’, the police would take over. In South Africa, a court order is required to evict people from structures such as those on the beach (RSA 1998; Cooper-Knock Reference Cooper-Knock2008). By allowing Edmund to enter into the area first, the police were outsourcing both illegal violence and illegal eviction to citizens who would be far harder to trace and hold accountable than uniformed officers.Footnote ⁵⁵

Clearly, then, some police allowed or facilitated space outside the law in which citizens could act. However, it was hard for citizens without working relationships with the police to know where these opportunities might emerge and how secure they might be. This served to reinforce a public narrative that depicted police officers as being eager to prosecute vigilante citizens and residents being law-abiding as a result. This narrative was an unreliable guide to everyday policing in Berea, but an insightful window into subjective understandings of statehood, citizenship and everyday policing.