Twenty years of punishment (and democracy) in South Africa: The pitfalls of governing crime through the community

 

 

Gail Super

PhD in Law and Society from New York University, an MSc in Criminology from the London School of Economics and Political Science and a BA LLB from the University of Cape Town. She has practised as a human rights lawyer in Namibia, where she conducted extensive research on the Namibian criminal justice system, focusing on restorative justice, children in prisons and in pre-trial detention. She is currently an honorary research associate at the Centre for Criminology, University of Cape Town. gjs220@nyu.edu

 

 

---

ABSTRACT

This article examines how the ideology of 'community' is deployed to govern crime in South Africa, both by marginalised black communities and by the government. Although the turn to 'community' started under the National Party government in the late 1970s, there is no doubt that as a site, technology, discourse, ideology and form of governance, 'community' has become entrenched in the post-1994 era. Utilising empirical data drawn from ethnographic research on vigilantism in Khayelitsha, as well as archival materials in respect of AN policies and practices before it became the governing party, I argue that rallying 'communities' around crime combatting has the potential to unleash violent technologies in the quest for 'ethics' and 'morality'. When community members unite against an outsider they are bonded for an intense moment in a way that masks th very real problems that tear the community apart. Because violent punishment is one of the consequences of the state's turn towards democratic localism, we should question the way in which the 'community' is deployt as a tool of crime prevention, and subject it to rigorous scrutiny.

---

 

 

With the advent of formal democracy in South Africa in April 1994 one might have been justified in expecting that the criminal justice system would become less punitive and that this would entail less reliance on imprisonment as a punishment par excellence.¹ However, although the numbers in custody have been reduced since an all-time high in 2004, South Africa has the highest incarceration rate in Africa and one of the highest in the world.² In 2013, the number of people serving life imprisonment stood at 11 000, as opposed to 400 in 1994.³

Democratisation has thus brought with it a dramatic increase in long-term prison sentences, ranging from seven years to life. One of the consequences has been an escalation in the number of maximum security risk category prisoners. Prison overcrowding is rife (albeit unequally distributed) and social workers, psychologists and other professionals who are key for rehabilitation are in woefully short supply.⁴ Indeed, in overcrowded prisons, prisoners only have 1,2 m² in which to eat, sleep and spend 23 hours of the day.⁵ The conditions are, to say the least, appalling.⁶

Yet, in 2002, precisely when the South African rate of imprisonment had almost peaked, the Department of Correctional Services introduced a restorative justice approach 'aimed at facilitating the mediation and healing process between offenders, victims, family members and the community'.⁷ This article discusses the apparent contradictions in, and consequences of, the state embracing 'community' based criminal justice initiatives in tandem with long-term imprisonment. Although the turn to community started in the late 1970s, under the National Party government, there is no doubt that as a site, technology, discourse, ideology and form of governance, the 'community' has become entrenched in the post-1994 era. It is, of course, one of those terms that is so vague and amorphous as to be capable of many different meanings. As such, it appeals to all parts of the political spectrum.

The paper focuses on the punitive underside of community, which, at its most extreme, manifests in the form of vigilante killings. Drawing on ethnographic and documentary research into non-state crime prevention and punishment practices in Khayelitsha, conducted between 2012 and 2014, I argue that vigilantism is in fact part of a continuum of violent technologies that are both connected with, and imbricated in, this shift to governing through the community.⁸ Not only is the term 'community' a discursive construct but, as I argue, because of its distinctively punitive iterations, both in the present and historically, it manifests as a peculiar mix of socialist-based grassroots activism, coupled with violently exclusionary attitudes towards those accused of criminality. I show how violent punishment is one of the consequences of this turn towards democratic localism. As such we should question the way in which the 'community' is deployed as a tool of crime prevention, and subject it to rigorous scrutiny.

I start by discussing the argument that community/ democratic policing is opposed to the principles of liberal minimalism enshrined in the South African Constitution.⁹ I then discuss the consequences of the ANC's version of 'mass political culture'; how the 'cultural patterns' inherited from the struggle against apartheid combine with the way in which the problem of crime, and what to do about it, is framed by political rhetoric, resulting in an approach that significantly stigmatises deviance.¹⁰ Deploying Zimring's concept of 'symbolic transformation' I argue that the distrust that poor communities have of the police translates into a call for more (and not less) punitive treatment of criminals.¹¹ Finally, I make some recommendations for future research.

 

Contradiction or coherence?

One explanation for the apparent contradiction in embracing the benign-sounding 'community', together with 'the prison', is that in fact these are not simultaneous, but consequential, penal developments. Thus, although the state embraced community policing and restorative justice in the early days of the heyday of rainbow nation democracy, in fact it soon thereafter shifted into a more punitive gear due to the panic about crime, and pressure to do something about it.¹² The argument, brought to its logical conclusion, is that community/democratic policing is somehow opposed to the punitive style of policing recently adopted by the South African state. In a challenging and provocative article Hornberger counters this analysis, arguing that the 'current [punitive] changes' should be interpreted as 'popular rather than elite or autocratic'.¹³ This helps to make sense of what might otherwise appear to be 'contradictory, incoherent trends'.¹⁴

According to Hornberger, community policing, which she refers to as 'a policing of proximity', seeks legitimacy from the community. This involves a 'penetration of policing by forms of local justice'.¹⁵ These forms of justice, which include the call for 'illegal violence', are removed from what Hornberger refers to as 'the civility of the law'.¹⁶ I argue that this 'illegal violence' is in fact part of a continuum of community-based crime prevention and punishment practices, where the legal and illegal are blurred, and where the state is complicit in the construction of vengeful 'communities'. In a context of great scarcity and rampant social and economic inequality, 'mob' justice serves as an occasion for victims, and the communities with which they are linked, to proclaim the extent of their suffering and seek punitive redress. It is structured by the state insofar as victims (and their communities) have, in the past 20 years, discursively at least if not always in practice, come to assume a central role in the criminal justice system.

Instead of seeing vigilantism as a form of 'mob justice', as a scourge, as inimical to 'civil' society and as being somehow outside of and opposed to it, I argue that we should acknowledge how vigilantes, or at least their supporters, are in fact part of 'civil society'.¹⁷

 

'Mass political culture' and the community

Rooted as it is in 1980s notions of people's power and the left-wing notion of grassroots democracy, the term 'community' has great rhetorical purchase in South Africa today. It encapsulates the communitarian emphasis of the 'people shall govern' clause of the Freedom Charter, echoes the global embracing of informalism, is presented as an effective means of combatting crime, and of course dovetails with the neoliberal shift towards greater responsibilisation - across all fields of government - by shunting responsibility from the state to the 'people' (a.k.a. 'the community').

This form of governance, one that operates in terms of a 'liberation paradigm', valorises local level initiatives by constantly seeking to mobilise communities on the ground.¹⁸ Thus, in 1992, the ANC stated that it was 'the community who are largely responsible for prosecutions [and] ... not the police alone who combat crime'.¹⁹ One of the objectives of Community Policing Forums (CPFs) is to 'enhance the ability of the police to combat and prevent crime, disorder and fear, in partnership with the community', and parole boards are meant to give the 'community' a special say in release decisions.²⁰ Not only has democratisation ushered in a growing discursive emphasis on giving crime victims a role to play in sentencing, bail and parole decisions, but the South African Police Service (SAPS) measures the success of its 'social crime prevention strategy' in terms of the number of crime awareness programmes, neighbourhood watches, business forums and street committees that are established to deal with crime.²¹ In some instances the notion of partnership policing even includes 'mobilising the community to oppose bail' via collaboration with the CPF.²² Indeed, the community is so fundamental to policing in democratic South Africa that police will, in future, be subjected to a 'stringent new recruitment process' that includes 'being paraded in front of community members', via a 'community parade'.²³

A 1997 amendment to the Criminal Procedure Act 51 of 1977 provides for crime victims and/or the community in which the crime occurred to play a role in bail decisions. In particular, a court may refuse bail where the release 'will disturb the public order or undermine the public sense of peace or security'.²⁴ The criteria that it may take into account relate entirely to how the community will react to the release. Thus bail may be refused where:

• The nature of the offence is likely to induce a sense of shock or outrage in the community where the offence was committed²⁵

• The shock or outrage of the community might lead to public disorder

• The release might jeopardise the safety of the accused

• The release will undermine or jeopardise the sense of peace and security among members of the public

• The release may undermine or jeopardise public confidence in the criminal justice system²⁶

In this way then, the door is opened for an ambiguous and vengeful 'community' to play a central role in the supposedly neutral criminal justice system. Pratt refers to this as a 'decivilising process' in terms of which the 'liberal notions of unemotive sentencing and bail decisions' are undermined.²⁷

 

History of the present

Marginalised communities in South Africa have a history of self-policing. Moreover, political violence has always played a prominent role in local politics.²⁸ Both the National Party government and the ANC liberation movement used the death penalty against their enemies, and the radical traditions of people's power and ungovernability sometimes resulted in violent punishment.²⁹ As such, punishment in South Africa has historically been relatively unconstrained by the minimalist considerations associated with liberalism.³⁰

In the turbulent 1980s the apartheid state depicted township activists as violent criminals and terrorists, stripping their acts of any political dimension. At the same time the 'comrades' committed violent acts in the name of politics while accusing South African government officials and their lackeys of being the real criminals. At times, when criminal gangs looted trucks driving into the townships, some political activists claimed that this was part of the struggle.³¹ Similarly, certain acts - such as 'necklacing' - that might otherwise be regarded as gratuitous violence, assumed some kind of political salience, due to the fact that the targets of these acts were accused of being spies or apartheid collaborators, and thus on the wrong side of the 'just war'. In fact, there was some substance to the claims of both the state and political movements. Among and alongside the 'comrades', activities and people emerged to take advantage of these township struggles to wreak violence for their personal gain - hence the appearance of the label 'comtsotsis', meaning criminals masquerading as 'comrades'.³²

With the transition from a white minority government to a black majority government in 1994, the ANC had to transition from a liberation organisation calling for ungovernability in the black townships, to a governing party. As such it had to govern and demonstrate control over a crime situation about which citizens were becoming increasingly vocal, and which it had hitherto ignored. Tensions arose between its previous pronouncements on how it would deal with crime, the punitive practices that it now adopted, and its attempts to legitimate the police via community or 'democratic' policing. As crime became increasingly politicised - the subject of many parliamentary debates and a general consensus on the need to treat criminals harshly - so too did the new ANC government seek to simultaneously legitimate the previously vilified police and prove that it was not soft on crime. It did so via a two-pronged strategy of implementing community policing and uncoupling criminals from a political and social context, presenting them as a threat to the country's young democracy. Thus, at the opening of Parliament in 1995, then President Nelson Mandela blamed crime and violence for 'eroding the foundation of our democracy', necessitating a 'harsher approach'.³³ A decade later, political leaders have called on police to 'kill the bastards',³⁴ to 'teach them a lesson'³⁵ by means of the use of lethal force and to show 'no mercy'.³⁶ This is a far cry from the 1992 ANC discussion document 'Crime and crime control' which, to give one example, ascribed gang formation to 'structural and political reasons' and presented gang members as having 'legitimate economic needs'.³⁷

As such, there is no doubt that crime and punishment have assumed an ideological importance in the 'new' South Africa that is markedly different from the situation during apartheid. This 'hyper-politicisation of penal policy' frames the field of punishment (and other 'self-help' initiatives) in Khayelitsha and other marginalised South African townships.³⁸

 

The legitimacy conundrum

Community solutions to crime and policing are attractive, not because they actually reduce the incidence of offences but because, like punitive punishments, they reassure people that something is being done to collectively prevent crime.³⁹ De Klerk argues that the ideology of collectivism encourages vigilantism because the raised expectations generated by institutions of partnership policing, such as CPFs, are inevitably not met.⁴⁰

During evidence given at the Commission of Inquiry into Policing in Khayelitsha (hereafter referred to as the Khayelitsha Commission), a senior police officer stated that he knew about 'formal meetings' that had resulted in a decision to evict people from their homes due to a crime.⁴¹ A resident testified that when the 'community' had called her to a meeting, demanding that her nephew leave the area due to his alleged 'criminality', she did not argue, and did not think of other options: 'Our main concern was that he should leave the house so he wouldn't be harmed. We could see the mood of the residents and it appeared that they would do something.'⁴² Another witness stated that the boys who had allegedly stolen his niece's money, leather jacket and cellphone climbed into his car voluntarily because 'they were asking us not to assault them, saying that their parents were going to pay back the money'.⁴³ Similarly, a teenager living in a tin shack in Enkanini told me that when her blankets and hair-iron were stolen during a break-in she did nothing, because she did not see the thieves - but that if she had seen them she would have alerted the 'community' to assist her in retrieving her goods.⁴⁴ In this sense then, there is a desire for criminals to be banished and for the victims to get their goods back.

Banishment in informal settlements sometimes takes the form of demolition of the dwellings of suspected criminals. In some instances a decision taken at street committee level to banish an 'offender' by demolishing his/her shack is taken too far when the enforcers not only destroy the dwelling but also assault (or kill) the resident, sometimes destroying other homes in the purging process.⁴⁵ One of my interviewees was a member of three different committees, all with varying relationships with the state: the local neighbourhood watch, a less formalised community patrol group, and the Khayelitsha CPF. He had also been a participant in the demolition of three shacks in an informal settlement after a decision to this effect was taken at a 'general council' gathering of the street committees in his area.⁴⁶ Thus, members of neighbourhood watches and street committees may be, or may have been, the same people who become part of a 'mob' - the point being that there is fluidity between structures. As such it becomes difficult to distinguish the mob from the community, the unlawful from the lawful.

Where the police encourage the public to engage in partnership policing, join neighbourhood watches, establish community patrols, and observe and report crimes, they create the expectation that they will be available to assist in instances when crime is detected.⁴⁷ However, given a scarcity of resources, particularly in poorer communities, this promise cannot be fulfilled. There are widespread allegations of police complicity and police have themselves admitted that they tacitly permit violent community-based ordering processes.⁴⁸ Surprisingly, some members of the Social Justice Coalition (SJC) also expressed support for the beating of 'criminals' by 'community members' as a technique to retrieve stolen goods, even though they stated that they themselves did not participate in violent activities.⁴⁹

Dominant public discourse in South Africa is replete with allegations that criminals are released on bail when they should not be, that prisons are like five-star hotels, that the criminal justice system is too slow, that the police do not do their jobs properly, that there are too many acquittals, that there are too many early releases; the list is endless. One only has to peruse the record of the evidence given at the Khayelitsha Commission to find overwhelming evidence in this regard. The point of this article is not to prove/disprove the veracity of these claims but to note their prominent presence and, at the same time, to observe how, in the quest for harsh treatment of criminals, the negative consequences of imprisonment are almost entirely excised from the debate.

Despite this lack of trust in the criminal justice system, in what Zimring refers to as a process of 'symbolic transformation',⁵⁰ the consistent call from residents in marginalised communities is for a more intimate relationship with a punitive state. This translates into a call for more arrests, punitive punishments, including long prison sentences, and the reinstatement of capital punishment.⁵¹ Even though the state is largely absent/failing/distant, however one describes it, citizens of 'frontier societies', lacking the resources that the well-off have to deploy private security, translate an exercise of police power (in other words a punitive criminal justice system) into a service for victims.⁵²

 

Conclusion

As Garland points out, 'liberal institutions' such as the rule of law and Bill of Rights are not the same as democratic institutions.⁵³ Whereas the former tend to constrain state punishment, the latter are not concerned with restraint but with punishing in accordance with what the majority wants. In South Africa the history of the liberation struggle, combined with apartheid-sanctioned self-help township initiatives, people's courts, notions of community empowerment and grassroots localism, have given popular democracy an exceptional flavour. Looking back from the vantage point of 2014, it is clear that the South African version of mass democracy exists uneasily, side by side, with the liberal minimalism of its Constitution.

While the idea of community sounds progressive and inclusive, important questions should be asked about the power relations between different communities, within communities, and between communities and state agents. The danger of rallying communities around crime combatting is that it can, and does, unleash violent practices in the quest for a 'moral community'.⁵⁴ The ironic twist is that 'mob justice' is in part a mass technology to protect private property in a context of endemic inequality.⁵⁵ When community members unite against an outsider they are bonded for an intense moment in a way that masks the very real problems that tear the community apart and/or separate it from other less or more socially connected and empowered communities.⁵⁶

I have argued in this article that violent community-based punishments are constitutive of, and constituted by, violent state punishments and crime prevention practices (both lawful and unlawful). For this reason South African researchers and policymakers would do well to heed Johnson's exhortation to 'examine the conditions that determine whether and to what extent participants in community-based crime prevention initiatives are prepared to act outside the law in given circumstances.⁵⁷

In particular, a research agenda in marginalised communities should include the phenomenon of punitive populism by mapping the field of crime prevention and punishment practices - asking how penality is constituted within specific localities. Instead of presenting 'communities' and 'mobs' as self-evident binaries we should interrogate the power relationships, local politics, societal networks, social capital and other dynamics that constitute these transient and non-homogenous groupings.

 

Notes

1 In 1992 the ANC stated that 'our crime problems are NOT being solved by large-scale imprisonment' and that 'however much one condemns those deeds' the state response should show compassion for the perpetrator [emphasis as in the original]. See African National Congress (ANC), Discussion document: crime and crime control. What role should the police play?, Centre for Applied Legal Studies, File AK 2195, P2 Police, South African history archives, University of the Witwatersrand, 1992, 7.

2 Judicial Inspectorate for Correctional Services, Annual report 1 April 2009 - 31 March 2010, 11, http://judicialinsp.dcs.gov.za/Annualreports/Annual%20Report%202009%20-2010.pdf (accessed 15 April 2014).

3 Judicial Inspectorate for Correctional Services, Annual report 1 April 2012 - 31 March 2013, note 44, 34, http://judicialinsp.dcs.gov.za/Annualreports/ANNUAL%20REPORT%202012%20-%202013.pdf (accessed 15 April 2014). Between 2000 and 2010, sentences of life imprisonment increased by 572% and those serving more than ten years by 128%. See Judicial Inspectorate for Correctional Services, Annual report 1 April 2009 - 31 March 2010, 15.

4 Ibid., 25; Judicial Inspectorate for Correctional Services, Annual report 1 April 2012 - 31 March 2013, 81.

5 Judicial Inspectorate for Correctional Services, Annual report 1 April 2006 - 31 March 2007, 16, http://judicialinsp.dcs.gov.za/Annualreports/ANNUAL%20REPORT%202006.pdf (accessed 16 April 2014).

6 The conditions, as is the case with overcrowding, vary from prison to prison.

7 Department of Correctional Services, Annual report 1 January 2000 - 21 March 2001, Foreword.

8 My research methodology included selective in-depth interviews with residents of Khayelitsha: a member of Mayitshe (a community patrol group), the Chairperson of the 'Father's Committee' in Enkanini (an informal settlement), members of the Social Justice Coalition (SJC), community journalists who were following vigilante incidents, five street committee members, four residents who openly identified themselves as 'vigilantes', and three mothers of victims of vigilantism. I also conducted participant observation in the SJC 'Campaign for Safe Communities' meetings, attended the first ten days of the proceedings of the Commission of Inquiry into Policing in Khayelitsha (hereafter referred to as the Khayelitsha Commission) and the proceedings in The State vs Mziwabantu Mncwengi, Mzimasi Mncwengi, Buyelwa Mncwengi, Lumnko Babalaza, Xolani Makapela, Mawende Siboma, case number SS03/2013, Western Cape High Court, South Africa (part-heard), where the accused were charged with the kidnapping and murder of four youths who were alleged to have stolen the plasma television set of one of the accused. This was supplemented by documentary research (including the official records of the Khayelitsha Commission, the judgements in court cases that dealt with vigilante incidents, and documents that the Western Cape Department of Community Safety made available for my perusal).

9 I use the term 'liberal minimalism' to refer to those Chapter Two rights that constitute the essence of liberal democracies, namely life, equality, freedom and security of the person, access to courts and the rights of arrested and detained persons. However, it should be noted, as one of the anonymous reviewers pointed out, that because the South African Constitution protects civil, political, economic and cultural rights there is debate as to whether it is minimalist or not.

10 The term 'mass political culture' comes from Vincent Darracq, The African National Congress (ANC) organization at the grassroots, African Affairs 107(429) (2008), 589-609, 595.

11 According to Zimring the 'symbolic transformation of capital punishment' entails its transformation from being a symbol of a powerful state into a form of victims' redress. This is crucial in constructing popular support for the death penalty in the American South where distrust of government is the norm. Zimring argues that '"vigilante" cultural traditions sustain the idea of harsh punishment as a communal ritual on the victim's behalf and counteract the inhibiting effect of distrust of government on support for the death penalty'. See Steven Messner, Eric Baumer and Richard Rosenfeld, Distrust of government, the vigilante tradition, and support for capital punishment, Law & Society Review 40(3) (2006), 559-590, 560.

12 See, for example, Janine Rauch, Criminal justice after apartheid, in C Call (ed), Constructing justice and security after war, Washington DC: United States Institute of Peace Press, 2007; and Dirk van Zyl Smit and Elrena van der Spuy, Importing criminological ideas in a new democracy: recent South African experiences, in Tim Newburn and Richard Sparks (eds), Criminal justice and political cultures, national and international dimensions of crime control, Cullompton: Willan, 2004.

13 Julia Hornberger, From General to Commissioner to General - on the popular state of policing in South Africa, Law & Social Inquiry 38(3) (2013), 2.

14 Ibid.

15 Ibid., 12. See also Fouchard, who argues that community policing does not have a uniform definition but may be used as a 'euphemism for a particular concept of police-civil society relations which include different local structures such as Community Police Forums, ad hoc anti-crime campaigns and neighbourhood watch schemes'. Laurent Fouchard, The politics of mobilization for security in South African townships, African Affairs 110 (441) (2011), 608-609.

16 Ibid.

17 See Anthony Minnaar, The new vigilantism in post-April 1994 South Africa: crime prevention or an expression of lawlessness, Institute for Human Rights and Criminal Justice Studies, May 2001; Benjamin Háefele, Vigilantism in the Western Cape, Department of Community Safety, Western Cape Provincial Government, 2006; Makubetse Sekhonyane and Antoinette Louw, Violent justice, vigilantism and the state's response, Monograph 72, Institute for Security Studies (ISS), 2002, http://www.issafrica.org/publications/monographs/monograph-72-violent-justice-vigilantism-and-the-states-response-makubetse-sekhonyane-and-antoinette-louw (accessed 15 April 2014). These authors argue that vigilantes are outside of and opposed to 'civil society'. See, however, Fouchard, The politics of mobilization for security, 607-627, 609. Fouchard argues that there is 'no clear distinction between vigilantism and community policing and instead [they] should be analysed as two sides of the same mobilization process'.

18 Darracq, The African National Congress (ANC) organization at the grassroots, 595.

19 ANC, Discussion document: crime and crime control, 8-9.

20 See Department of Safety and Security, Community policing, policy framework and guidelines, 1997, 8; and Department of Correctional Services, Correctional supervision and parole boards, http://www.dcs.gov.za/Services/CorrectionalSupervisionandParoleBoards.aspx (accessed 14 April 2014).

21 Khayelitsha Commission, Affidavit of Provincial Commissioner Western Cape Arno Heinrich Lamoer, Western Cape annual performance plan 2012-2013, Annexure PC1, http://www. khayelitshacommission.org.za/images/witnesses/83._Lieutenant_General_Arno_Lamoer.pdf (accessed 15 April 2014). Chandre Gould notes that giving victims a role in the criminal justice system is 'stuck at a policy level with crime victims often complaining that they are in fact not consulted'. Personal communication, 10 April 2014.

22 Khayelitsha Commission, Affidavit of Colonel Michael Francis Reitz, SAPS Lingelethu West performance plan 2010-2011, 33, http://www.khayelitshacommission.org.za/images/witnesses/58.%20S tatement%20of%20Col%20Reits%20 SAPS%20statement%20file%201.pdf (accessed 15 April 2014).

23 Michelle Jones, New breed of police recruits on parade soon, Cape Times, 28 March 2014, 7.

24 Criminal Procedure 2nd Amendment Act 1997 (Act 85 of 1997), Section 4(d).

25 Section 4 (e) inserts Section 8A into the original Act, i.e. into Criminal Procedure 1977 (Act 51 of 1977), Section 60 (a).

26 Section 4(d).

27 John Pratt, Punishment and 'the civilizing process', in Jonathan Simon and Richard Sparks (eds), The Sage handbook of punishment and society, Los Angeles, London, New Delhi, Singapore, Washington DC: Sage, 2013, 97.

28 See Jan de Klerk, Community Policing Forums - inducing vigilantism, Servamus (October 1999), 48-52. He refers to the Sophiatown Nightwatch, the Christmas Watchmen, the Bangalalas, Makgotlas and SDUs that all operated at various points pre-1994. See also Mike Brogden and Clifford Shearing, Policing for a new South Africa, London: Routledge, 1993; Nicholas Haysom, Mabangalala: the rise of rightwing vigilantism in South Africa, Centre for Applied Legal Studies, 1986; James Martin, Vigilantism and state crime in South Africa, State Crime 1(2)(2012), 217-34; Gary Kynoch, Urban violence in colonial Africa: a case for South African exceptionalism, Journa of Southern African Studies 34(3) (2008), 629-45.

29 Lars Buur and Steffen Jensen, Introduction: Vigilantism and the policing of everyday life in South Africa, African Studies 63(2) (December 2004), 139-152; Gary Kynoch, Of compounds and cellblocks: the foundations of violence in Johannesburg, 1890s to 1950s, Journal of Southern African Studies 34(3) (2011); Benedito Machava, State discourse on internal security and the politics of punishment in post-independence Mozambique, Journal of Southern African Studies, 37(3) (2011), 593-609; Jeremy Seekings, The revival of 'people's courts': informal justice in transitional South Africa, in G Moss and I Obey (eds), South African Review 6, Johannesburg: Ravan, 1992, 186-200; Gail Super Governing through crime in South Africa: the politics of race and class in neoliberalising regimes, Surrey, Burlington: Ashgate, 2013; Richard Wilson, The politics of truth and reconciliation in South Africa, legitimizing the post-apartheid state, Cambridge, New York, Melbourne, Madrid, Cape Town: Cambridge University Press, 2002.

30 See David Garland, Penality and the penal state, Criminology 51 (3) (2013), 475-517, where he discusses this in an American context.

31 Steffen Jensen, Gangs, politics and dignity in Cape Town, Oxford, Chicago, Johannesburg: James Curry, University of Chicago Press, Wits University Press, 2008.

32 Bronwen Harris, 'As for violent crime that's our daily bread': vigilante violence during South Africa's period of transition, Violence and Transition Series 1, May 2001; Jensen, Gangs, politics and dignity in Cape Town; Minnaar, The new vigilantism in post-April 1994 South Africa; Michael Morris and Douglas Hindson, South Africa: political violence, reform and reconstruction, Review of African Political Economy 53 (1992); Gail Super, The spectacle of crime in the 'new' South Africa: a historical perspective (1976-2004), British Journal of Criminology 50(2) (March 2010).

33 Department of Safety and Security, Policing priorities and objectives for 1999/2000, 1999, 9.

34 'Kill the bastards' - Minister's astonishing order to police, The Star, 10 April 2008.

35 SAPS must fight fire with fire, The Citizen, 12 November 2008.

36 Johann Burger, 2009, cited in David Bruce, An acceptable price to pay? The use of lethal force by the police, occasional paper 81, Open Society Foundation, Criminal Justice Initiative, 2010, 9, http://osf.org.za/wp/wpcontent/uploads/2012/09/CJI_Occasional_Paper_81.pdf.

37 ANC, Discussion document: crime and crime control, 8-9.

38 The phrase 'hyper-politicization of penal policy' comes from Marie Gottschalk, The carceral state and the politics of punishment, in Simon and Sparks (eds), The Sage handbook of punishment and society, 217.

39 See Nicola Lacey and Lucia Zedner, Discourses of community in criminal justice, Journal of Law and Society 22(3) (September 1995), 301-25; Joanna Shapland and Jon Vagg, Policing by the public, London: Routledge, 1988.

40 De Klerk, Community Policing Forums - inducing vigilantism.

41 Established by the premier of the Western Cape in 2012 and mandated to investigate 'allegations of police inefficiency and of a breakdown in relations between the community and the police in Khayelitsha'. See Khayelitsha Commission, Record, 31 January 2014, 4636, http://www.khayelitshacommission.org.za/2013-11-10-19-36-33/hearing-transcriptions/public-sitting.html (accessed 18 May 2014).

42 Personal notes taken during Nomakuma Bontshi's testimony, 24 January 2014.

43 Personal notes taken during Mr Simelele's testimony, 27 January 2014.

44 Interview, 16 October 2013.

45 Interview with the chairperson of the 'X' South African National Civic Organisation (SANCO) branch, 19 October 2013; interview with representative of the Mayitshe Group, 1 February 2014; interview with the chairperson of the Fathers' Committee, Enkaninini, 1 February 2014; testimony by Nomakuma Bontshi, Khayelitsha Commission, 24 January 2014, testimony by Mr Hendrickse, Khayelitsha Commission, 31 January 2014; interview with a former member of Amadlozi (a vigilante organisation in Port Elizabeth), 29 January 2014; interview with witnesses in The State vs Mziwabantu Mncwengi, Mzimasi Mncwengi, Buyelwa Mncwengi, Lumnko Babalaza, Xolani Makapela, Mawende Siboma, 26 August 2012. See also Steffen Jensen, Policing Nkomazi: crime, masculinity and generational conflicts, in David Pratten and Atreyee Sen (eds), Global vigilantes, New York: Columbia University Press, 2008, 47-68.

46 It should be noted that, although interviewees consistently referred to street committees as making the banishing decisions, this does not necessarily mean that these bodies are affiliated to SANCO, since the term is a code for many types of gatherings. See Bongani Tshehla, Non-state justice in the post-apartheid South Africa - a scan of Khayelitsha, African Sociological Review 6(2) (2002), 47-70.

47 De Klerk, Community Policing Forums - inducing vigilantism, 51.

48 See Major General CP de Kock, Serious crime in Khayelitsha and surrounding areas, Crime Research and Statistics, Crime Intelligence (AL 30), 3 August 2012, iv. In this report De Kock, after referring to taxi associations as being 'in control of vigilante actions', states: 'They [the taxi associations] even went as far as imposing a curfew after 21:00 in the evening. Everyone (including the police) was very happy with the decrease in crime at the time ...' Although this was admitted in the context of what happened six or seven years ago, police complicity is still evident. See in this regard the testimony of school principal Mr Mjonondwana, Khayelitsha Commission, 28 January 2014. This was also the sense I got from chatting to police officers during court adjournments at the Western Cape High Court in August 2013.

49 This was communicated to me during interviews with four SJC members on 4 February 2013, during informal conversations I had, and during a group meeting facilitated by the Centre for the Study of Violence and Reconciliation on 13 March 2013. Given the SJC's left-wing, human rights-based credentials and the fact that it is sensitive to the social, economic and structural circumstances that might lead some people to committing offences, one might have expected SJC members to be less supportive of the harsh treatment of criminals.

50 Messner, Baumer and Rosenfeld, Distrust of government, 561-562.

51 Ironically, even Advocate Masuku, acting for the SAPS in the Khayelitsha Commission, referred to the desirability of the death penalty when, while cross-examining a witness, he stated: 'So if for example a person was arrested and evidence was properly collected and the prosecution was properly done and there was a conviction and the person went to jail and in certain instances there was the death penalty . if that were to be done people would feel a little more confident, confidence in the whole system.' Khayelitsha Commission of Inquiry, Record, 31 January 2014, 958, http://www.khayelitshacommission.org.za/2013-11-10-19-36-33/hearing-transcriptions/public-sitting.html (accessed 15 April 2014).

52 The term 'frontier society' comes from Craig Little and Christopher Sheffield, Frontiers and criminal justice: English private prosecution societies and American vigilantism in the eighteenth and nineteenth centuries, American Sociological Review 48(6) (1983), 796.

53 Garland, Penality and the penal state, 506.

54 The term 'moral community' comes from Buur and Jensen, Introduction: Vigilantism and the policing of everyday life in South Africa, 139-152, 144.

55 Violent retrieval of goods is clearly one of the goals of 'mob justice'. This is clear from a perusal of the evidence given at the Khayelitsha Commission, where numerous people testified to 'mobs' having beaten alleged thieves to death in the quest to retrieve their property. It is also borne out by all of the interviews that I conducted. According to a police report, between April and June 2012 there were 78 recorded 'vigilante incidents' in Khayelitsha. Most of the victims were young men, between the ages of 18 and 30, with at least half having either been caught in the act of stealing/robbing/ housebreaking and/or suspected of the same. See South African Police Service, Bundu courts, 01 April to June 2012, Khayelitsha Cluster, unpublished report.

56 See George Herbert Mead's classic piece on the psychology of punitive justice where he states: 'The criminal does not seriously endanger the structure of society by his destructive activities, and on the other hand he is responsible for a sense of solidarity, aroused among those whose attention would be otherwise centered upon interests quite divergent from those of each other.' George Mead, The psychology of punitive justice, American Journal of Sociology 23 (1918), 577-602, 591.

57 Les Johnston, What is vigilantism?, British Journal of Criminology 36(2) (Spring 1996), 232-233.