Police Transformation and the South African TRC

by

Janine Rauch

Race and Citizenship in Transition Series, 2004.

Janine Rauch is an independent consultant.

Acknowledgements

The author would like to thank Tebogo Umanah, Stef Snel, Piers Pigou, Khareen Pech, Gail Wannenberg and the various CSVR researchers who assisted with gathering the data used in this report.

Thanks are also due to the various respondents who agreed to be interviewed.

Generous information-sharing and feedback from Carnita Ernest, Bronwyn Harris and Nahla Valji at the CSVR's Transition and Reconciliation Programme helped shape and refine this report, especially the elements relating to the TRC and amnesty processes.

Elrena Van Der Spuy and Elaine Atkins at the UCT Institute of Criminology provided invaluable support and research assistance.

The research and production of this report was generously funded by the Charles Stewart-Mott Foundation, Development Cooporation Ireland and the Ford Foundation.

Contents Race and Citizenship in Transition Series Executive Summary Introduction    Methodology and author context Policing Apartheid Negotiating Democracy 1990-1994    Negotiating Amnesty Police Reforms    1994: Democratisation, accountability and human rights    1995-1998: Prevention, prioritization and multi-agency collaboration    Managerialism, service delivery and the partnership with business    2000: Crackdown!    2002-2003: Settling in … The Truth and Reconciliation Commission and the SAPS    1990-1994: Uncovering some truth before the onset of the Commission    1994-1995: Preparing for the TRC – Inside the police    During the TRC: Telling the truth about apartheid policing    1996-2001: Amnesty applications by police officials    The TRC Final Report and Recommendations        Political control at provincial level        The Independent Complaints Directorate        Public order policing        Police involvement in criminal activity        Recruitment and training        The SAPS and Correctional Services        Security Companies        Gun control The impact of the TRC on transformation of policing    The impact of the TRC-the view of police officials        Positive views about the TRC        Negative views about the TRC     Impact and implementation of the TRC Recommendations     TRC impact on human rights adherence and abuses of power Conclusions: Separating out the impact of the TRC from the impact of political transition?    Measuring police reform    Moral and ethical impact References

The Race and Citizenship in Transition Series¹

Bronwyn Harris, Nahla Valji, Brandon Hamber and Carnita Ernest

Race and citizenship are extremely complex concepts. In post-apartheid South Africa, they find expression on many different levels, including identity, conflict, nationalism, history, politics and inter-personal relationships. They occupy a spectrum ranging from everyday practices and interactions, to formal political and macro-economic forces. They also overlap with notions of reconciliation, justice and reparation, and, although they are separate notions with different histories, they overlap with each other. This creates an added dimension of complexity. Both race and citizenship can be (and commonly are) articulated and/or silenced to serve particular interests. Both can also feed into certain forms of violence, including xenophobia and racially motivated hate crime. Any analysis of race and citizenship must therefore acknowledge the complexity of their expression, representation and impact. Such complexity in the South African context must be assessed in relation to the country's apartheid history, as well as the processes of reconciliation best captured by the Truth and Reconciliation Commission (TRC).

Apartheid created race as a mechanism for violence. Race, in and of itself, was the social and psychological reality through which repression and violence functioned. Racism was institutionalised, legalised and internalised. South Africans saw the world in 'black' and 'white' terms and violence was commonly used to maintain this status quo. However, during the Mandela era (1994-1999), a new vocabulary emerged to describe the social order. This vocabulary spoke of nationhood, unity, racial harmony and reconciliation. South Africa was described as a 'rainbow nation'. Reference to race entered a sensitive and delicate terrain. This was a positive attempt to give South Africans a new language for speaking about – and to - each other. But, at the same time, it rendered the real, often violent, consequences of race invisible. In the Mandela era, there was little national debate on how race had influenced past human rights violations. There was also little recognition that race continues to shape identity and interactions – violent or not – within the present.

By contrast, the Mbeki era (1999-ongoing) has been characterised by a 'return to race'. This is partly a consequence of different presidential styles and roles – while Mandela had to stress forgiveness and underplay racial issues in order to consolidate a peaceful (and at times precarious) transition, Mbeki, as he stated in his 'two nations speech', has had to deal with economic inequality rooted in past racial practices. Additionally, the 'return to race' has been forced upon the society by violence: through the actions of white extremists like the Boeremag, as well as less political cases of racist hatred. Less violent expressions of/about race have also re-entered popular and political discourse: in 2000, the Human Rights Commission held hearings into racism in the media, and, in 2001, South Africa hosted the World Conference against Racism, Xenophobia and Related Intolerance.

Although race can be read across these two discrete eras - 1994-1999 and 1999-ongoing - it is important not to oversimplify or reduce the differences to how race has been articulated. Despite a general 'return to race' post-1999, there have been numerous contradictions and striking silences on the issue; for example, within the realm of violence and conflict, as well as Mbeki's own discourse (in 1996, he gave his inclusive 'I am an African' speech, which contrasted with his 'two nations' speech in 1999, but at the opening of parliament in 2001, he seemed to discard the two nations analogy in favour of a 'united' South Africa, irrespective of race). Also, while issues of race have partially emerged in the Mbeki era, the notion of reconciliation – particularly racial reconciliation - has become increasing invisible. The TRC finally completed its work in March 2003. Many have interpreted this as the end of South Africa's reconciliation process. However, incidents of racial prejudice, intolerance and violence, both within South Africa and internationally, suggest that the TRC was just the beginning and not the end of a sorely needed social dialogue about racial reconciliation.

The Truth and Reconciliation Commission (TRC)

South Africa did not 'invent' the truth commission. Since 1974 there have been more than twenty-five truth commissions around the world. But it was the South African Truth and Reconciliation Commission (TRC) that captured the world's attention. This is partly due to international interest in the fight against apartheid. Also, the TRC was the largest and best resourced commission, and it was afforded extensive media coverage, both domestically as well as internationally. This ensured that the world was exposed to the Commission, and the openness of the process meant that the violence of the past could no longer be denied. The South African model also attracted scrutiny because it promised an alternative way of peacefully resolving entrenched difference through the unique 'truth for amnesty'² deal upon which it was premised. Consequently, the notion of using a truth commission to deal with political conflict has gained momentum and many countries are now holding their own Commissions.

TRC Chairperson Archbishop Desmond Tutu said that without the compromises made during the negotiations to ensure majority rule in South Africa, the country would have gone up in flames. From this perspective it follows that the agreement by the African National Congress (ANC) to grant amnesty to perpetrators of apartheid violence was a pragmatic choice. Amnesty was the price, albeit a costly one for victims, for saving the innumerable lives that would have been lost if the conflict had continued. However, unlike in most transitional countries to date, amnesty in South Africa was neither blanket nor automatic. Conditions applied to the South African amnesty and the TRC was the vehicle for this process.

The TRC process began in December 1995 and finished in March 2003, when the Commission handed over the final 2 volumes of its 7 volume report. 7 116 people applied for amnesty. Almost 22 000 people came forward and told how they were victimised under apartheid. The TRC made a number of recommendations to the South African government regarding financial and symbolic reparations, issues of justice and ways to address relationships between South Africans. It is these issues that still need to be grappled with and addressed.

Evaluating the TRC

The public acknowledgement of past violations was perhaps the TRC's greatest success; as the brutal horrors of apartheid found their way, via the media, into the living rooms of every South African. An undeniable historical record has been created. However, apartheid history still remains contested and fraught with racialised interpretations; for example, many white South Africans continue to deny the impact of apartheid and many dismissed the TRC itself as a 'political witch-hunt' (cf. Theissen, 1997³). The role of the TRC - in both writing history and as an historical process itself – demands ongoing scrutiny.

At a narrower, more immediate level, a minority of victims did uncover suppressed truths about the past. In some cases, missing bodies have been located, exhumed and respectfully buried. For others, the confessions of perpetrators have brought answers to previously unsolved political crimes – crimes, which the courts, due to expense and inefficiencies, may never have tried. However, for many, the TRC began a process that it was unable to complete. Many of the victims who went before the TRC, with the hope that their case would be investigated, feel let down and no closer to the truth than before they publicly told of their suffering. Irrespective of the feasibility of investigating every case, victims' high expectations of the TRC have been dashed, and in their eyes, this has undermined its credibility.

Justice also remains a burning issue. Politicians may be able to justify the exchange of formal justice for peace, but it was difficult for victims to watch while the perpetrators received amnesty. Not only were many perpetrators 'let off the hook', victims feel let down and disappointed by the government's response to the TRC. Regarding financial reparations, the Commission recommended that the government should pay those victims identified through the TRC process R3 billion, in annual installments over a 6 year period (this total figure represents 0.001% of the country's annual R300 billion budget, which translates into R136 000 per individual). However, the South African government has only agreed to pay R30 000 per individual, in a once off payment. The Commission also recommended that business and other apartheid beneficiaries should pay a once-off wealth tax and that the country's inherited apartheid debt (which accounts for approximately 20% of the government's annual budget) should be restructured in order to free up money for development and redistribution. Again, the government chose to ignore these recommendations. This has left victims feeling betrayed. It also does not bode well for long-term reconciliation. As CSVR researchers, Polly Dewhirst & Nahla Valji (2003) note,

The 'miracle' of a new SA is hardly sustainable if it is built without restoring the dignity and humanity of the majority of its citizens, nor if it fails to address the economic inequalities which fuel social conflict.⁴

There are also debates about the broader merits of the TRC. At the very least the reconciliation project, with the TRC at the helm, has brought South Africa through the transition period with relative political stability. The humanist approach of Mandela and Tutu brought compassion to a brutalised country. Despite the horrors revealed by the TRC, glimmers of humanity shone through and provided hope for the future.

However for some, despite the merits of the TRC, 'reconciliation' is merely a euphemism for the compromises made during political negotiations - compromises that ensured continued white control of the economy. From this perspective, reconciliation is meaningless without structural change. A related, more cynical view is that the rapprochement between the old and new regimes was a strategy to consolidate a new black elite under the banner of reconciliation.

Many argue that the TRC missed the bigger picture by defining victims only as those who suffered intentional violence. Because the TRC focused on victims of gross human rights violations, such as torture and murder – it did not include the 'ordinary' victims of apartheid – the millions of South Africans who suffered from land removals, forced displacements, the migrant-labour system, Bantu education etc. As such the TRC did not engage directly with the institutionalised, structured ways in which racist policies affected and victimised people on a daily basis. Those who suffered more broadly from the economic ravages of apartheid and were not victimized directly by political violence were excluded from the TRC. An important question to ask is: what mechanisms do those, excluded from the apartheid state and then from the TRC, have for defining and consolidating a sense of citizenship in the 'new' South Africa?

Similarly, the degree to which the TRC used race as an explanatory variable in its understanding of the abuses it investigated remains questionable. In some cases, 'race' was generally collapsed into 'political motive', as exemplified by the amnesty decisions in the Amy Biehl, Chris Hani and St James' Massacre cases. However, this was done inconsistently and the relationship between race and politics was not clearly defined. Overall, the reconciliation process engaged less with 'black and white' issues then with inconsistent 'political' definitions of perpetrators and victims. This has had the after-effect of divorcing race, and racial identity, from the violence of the past. It similarly keeps race separate from understandings of violence in the present.

A related point is that, as a transitional justice mechanism, the TRC accepted and legitimated certain explanations for the violence of the past. In this way, it has played a key role in influencing the society's moral reactions to violence. This is specifically evident in the area of amnesty. The question needs to be asked, despite the compromises made to set up the TRC, has amnesty undermined South African citizens' sense of morality? Has it contributed to ongoing violence and impunity? Has it impacted upon how different race groups see each other? There have been various evaluations of the TRC, but none have taken into account the ways in which it has explicitly addressed race, morality and citizenship as components of past human rights violations and factors in contemporary social relations. It is precisely these questions that the Race and Citizenship in Transition Series has sought to address.

The different perspectives surrounding the TRC demonstrate the complexity of dealing with oppression and violence – and how past events shape the process of reconciliation.

The TRC was not alone in its attempts to build reconciliation in South Africa. A number of other institutions were set up to deal with the legacy of the past. These included for example the Land Claims Court and the Human Rights Commission. Other structures, such as the Independent Complaints Directorate, were set up to monitor ongoing abuses by the police. However the degree to which these institutions, and the TRC can be said to have consolidated reconciliation and effected transformation can, at best, be described as ongoing but desperately incomplete. There are ongoing police abuses, young people still express feelings of marginalisation, racism and racist incidents continue to take place, and the poor have not substantially benefited from the changes in the country.

Levels of Reconciliation

The process of reconciliation can be said to operate on a number of levels, i.e. the political, community and individual levels.

At the political level, reconciliation has been embodied in the compromises that lead to a political peace. This process can be said to be broadly successful, as it has brought political stability to South Africa.

At the community level, despite some successes by the TRC, reconciliation is largely incomplete, with many of the old racial and political divisions remaining in place. This is evidenced through high levels of residential segregation between black and white South Africans residentially. It is also expressed between different groups divided along political affiliation, such as ANC and IFP supporters, and xenophobic hostility between South Africans and foreigners, particularly those from elsewhere in Africa.

At the individual level, the question is far more complex and is bound to how individuals feel in relation to the process of reconciliation. Many individual victims feel that their needs have not been met by the TRC. At the same time, many of those who benefited from apartheid are still denying their complicity status. This is linked to the many who refuse any responsibility for reparations and redressing the past. There is also an expectation that the next generation will somehow begin with a 'clean slate' (Oakley-Smith⁵). The ongoing impact of a racist and violent past continues to play out through incidents of racist hate crimes and expressions of xenophobia. Hostility towards foreigners, particularly black Africans, commonly results in violence and is spurred on by overly zealous views of nationalism in the 'new' South Africa. In addition, many South Africans are finding themselves questioning their role in the country. This could be linked to the many young people who are leaving the country as they feel there is no future for them in South Africa.

A Crisis of Citizenship

We would like to suggest that there is a 'crisis of citizenship' in South Africa at present, which threatens the genuine reconciliation begun through processes such as the TRC. This crisis manifests itself in ordinary people asking where they belong in the new society. This crisis suggests that there is much work that needs to be done to consolidate the process of reconciliation and a sense of inclusive citizenship. The Race and Citizenship in Transition Series is a space for exploring this citizenship crisis, along with the related issues of race, reconciliation, violence and identity in South Africa. Key issues to be examined include:

• Racially motivated violence. To what extent does race continue to impact on patterns and trends of violence? How relevant is the concept 'hate crime' to the South African context? What challenges does the criminal justice system face in dealing with racially motivated violence?
• Race and the TRC. As a key instrument of transition, how did the TRC engage with the racism of South Africa's past? How has this impacted on the telling of history and contemporary understandings of racial relations?
• Lessons from Guatemala. There are many parallels between Guatemala and South Africa: historically, both countries were based on racist political systems, which resulted in racialised inequality and conflict. Both countries set up truth commissions to address their pasts and engage with citizenship in the future. Both countries continue to be marred by violence. What lessons can South Africa learn from the Guatemalan transition?
• Young people and race. How do young people conceive of citizenship, identity and racial reconciliation? The views expressed by the younger generation provide a means by which to evaluate the degree to which South African society has, or has not, transformed; as well as the longer-term influences of transitional processes.
• Institutional transformation and the legacy of racism. What recommendations and findings did the TRC make about transformation in the South African Police Services, schools, and efforts to address racism in South African institutions?

The Race and Citizenship in Transition Series is funded by the Ford Foundation, Development Cooperation Ireland and the Charles Stewart-Mott Foundation.

Series Editors

Bronwyn Harris
Nahla Valji

 

Executive Summary

This report examines the relationship between South Africa's Truth and Reconciliation Commission (TRC) and the reform of the police institution over South Africa's period of transition to democracy. In 1994, a new national police service – the South African Police Service (SAPS) - was created and the newly-elected, democratic government put a great deal of effort into police reform during the immediate post-transition period. At that time, the new police service comprised over 140 000 members, largely made up of 11 former apartheid police agencies, with an appalling history of abusive and biased policing. When the TRC began operating at the beginning of 1996, the new government had already solicited international policing assistance, put new training programmes in place - especially in respect of human rights and use of force - and appointed a new leadership from within the police.

During the pre-election period, fear about the coming TRC created a climate in which police officials started 'spilling the beans' on each other; resulting in a handful of high-profile prosecutions of former security force personnel for apartheid-era abuses. The fear of more prosecutions of police officials was fuelled by the State's success in prosecuting former South African Police (SAP) hitsquad commander Eugene De Kock. De Kock's experience also instilled another fear – that the loyal apartheid police members would be 'hung out to dry', unsupported by their former masters. While prosecutions did not materialize on the scale feared by the police, they were indeed abandoned by their National Party leaders, who used the TRC as an opportunity to distance themselves from police abuses and claim ignorance of the horrible conspiracies carried out in their name by their security agencies in the apartheid government's final years.

The TRC model combined a form of truth-telling about abuses committed during the previous apartheid era – such as those committed by the police – with an amnesty for those who chose to come forward and confess. The amnesty process was a controversial aspect of the TRC's work; and yielded only approximately 1600 meaningful applications, of which approximately 250 were from former police officials. The Amnesty Committee held the first hearing in May 1996 and concluded its work at the end of May 2001, having granted amnesty in respect of more than 60 percent of the applications it fully processed. It is unknown what proportion of the amnesty applications made by former police officers were successful, as this information was not provided by the TRC.

In addition to the work of its Amnesty Committee, the TRC also touched on policing matters in the victim hearings of the Human Rights Violations Committee, which allowed victims of police abuses to tell their stories. These were often broadcast through the TRC's daily media coverage. This process had a significant impact on public perceptions about the police, and on some police officials' perceptions about themselves and the organization for which they had worked. However, despite the strong evidence about apartheid policing that the TRC was able to amass, the Commission's recommendations on policing were weak and inarticulate. The TRC had failed to keep abreast of developments in police reform in the period from 1994 to 1998, when significant changes in the police organization had already taken place (although much of the legacy of apartheid policing were also still evident).

This report concludes that the TRC made little impact on the process of police reform in South Africa; although there may be longer-term intangible effects related to the notion of police accountability and to the moral climate within the police organization.

Introduction

The purpose of this report is to reflect on the relationship between South Africa's Truth and Reconciliation Commission (hereafter referred to as the TRC) and the country's national police agency,⁶ the South African Police Service (SAPS), in order to assess whether the TRC had any impact on the institutional transformation of the police service. This discussion is located within the context of institutional reform of the police as it unfolded after South Africa's first democratic government took power in 1994.

The TRC was a critical device for transitional justice and for coming to terms with South Africa's apartheid past.⁷ Another – equally important – component of transitional justice in South Africa was the use of prosecutions for crimes committed in the course of apartheid 'duties'; but this aspect is not examined here, except insofar as the threat of prosecutions for past violations was intended to encourage police officials to use the TRC as an opportunity to confess and reconcile. Only a very small number of prosecutions for apartheid-era abuses have been undertaken, and many of these have failed to secure convictions. A separate research venture to assess the impact of prosecutions in transforming the repressive institutions of the state will be necessary in order to compare the respective merits of the prosecutorial strategy versus the TRC approach.

Before assessing the relationship between the TRC and the policing institution, it is important to outline the historical context and broad processes of change that shaped the police over South Africa's process of transition from apartheid to democracy. It is through looking at these forces that the impact (or not) of the TRC can be measured and evaluated. This report will therefore focus on the following broad phases before turning to the TRC itself:

• Policing apartheid,
• Negotiating democracy (1990-1994)
• Negotiating amnesty, and
• Police reforms post-1994

Methodology and author context

The author served as an advisor to the Ministry of Safety and Security in the first democratic government in South Africa, and subsequently as a senior public servant in the civilian Secretariat of the Department of Safety and Security. Much of the data used in this report is drawn from first-hand experience, observation and recollection. This report therefore provides a unique perspective on police reform in the period between 1994 and the conclusion of the work of the TRC.

Policing Apartheid

The role of enforcing unpopular laws, as assigned to the police, courts and prisons during the apartheid years, created a profound crisis of legitimacy for the criminal justice system in South Africa. In many respects, the police played a military role, crushing popular protest and engaging in South Africa's domestic civil war, as well as being deployed in support of white regimes in independence struggles in neighbouring states.⁸ Under successive apartheid governments, the police, intelligence service, and the military vied for political dominance and greater slices of the national budget. By the late 1980s, the apartheid state was in severe crisis, forcing the police, the military and the bureaucracy to devise joint strategies and tactics to defeat the liberation movements, and an integrated National Security Management System (NSMS) was established to oversee the successive 'States of Emergency' that were declared after 1985.⁹ By 1986, a new period of state repression further alienated the police from the majority black population of South Africa:

the police and the army began to implement their counter- revolutionary strategy, with the aim of destroying the ANC and its allies, and restoring initiative to the state. The strategy entailed establishing a firm police and army presence in the townships¹⁰ and suppressing protest and or resistance; 'taking out' leadership through mass detentions, trials, harassments and assassinations; re-establishing intelligence networks; the 'counter-organisation' of communities through setting up groups with links to the security forces, the use of vigilantes; and the alleviation of socio-economic grievances through the upgrading of selected troublespots. Police and troops poured into townships around South Africa, setting up joint bases and commandeering facilities. They patrolled the streets in armoured vehicles. Curfews and other restrictions were imposed, meetings were banned and, where resistance had been intense, townships were sealed off by troop cordons and razor-wire barricades, while the police went from door to door searching for dissidents. People were picked off the streets and interrogated, leaders of community organisations, trade unions, religious and other groups were rounded up and thrown into jail or put in cells at police stations. Nearly 30 000 people¹¹ were detained in 1986. (Cawthra, 1993, pp. 32-33)

At the end of 1987, the apartheid state took its first public steps towards a negotiated settlement with the liberation movements, when it released ANC veteran Govan Mbeki from prison. This was a 'test case', and the security elites had not anticipated the massive popular reception that he received. Shortly afterwards, a new wave of bannings and repression began, and

extra-legal methods such as assassination, arson and the bombing of offices became more common. The courts were also kept busy, processing the thousands of people charged with offences arising from the conflict, ranging from arson and public violence through to treason and murder. (Cawthra, 1993, p. 34)

The costs of supporting this intense security strategy were unsustainable, the liberation movements were surviving and re-grouping, and international isolation of the apartheid state was intensifying as sanctions took effect. In early 1988, the South African military was defeated, for the first time, in battle in Angola – a crucial symbolic moment. The influence of the security agencies began to wane within the apartheid government, and a new breed of 'enlightened' Afrikaners began to emerge, advocating reform and dialogue with black leaders. The repressive strategy had resulted in virtual civil war in South Africa's cities and rural towns, high levels of militarisation, fear and conflict between its citizens, and the alienation of an entire generation of black children from schools and most other institutions and symbols of authority, including their families and traditional social structures.

Negotiating Democracy 1990-1994

In 1989, F.W. De Klerk assumed the State Presidency, and, in the following year, unbanned the ANC and the other liberation movements and released Nelson Mandela and his comrades from prison. This marked the commencement of a protracted period of negotiations towards democracy, which was characterised by extreme and brutal conflict within South Africa:

The carnage and destruction, which has left few of South Africa's black townships untouched, is virtually unparalleled in this country's bloody history. The politics of transition have resulted in ruptures in the very fabric of apartheid society. One of the consequences has been a dramatic escalation in violence in almost all dimensions of the society - within the community, the workplace and the home. This has been not only a quantitative increase in levels of violence, but also a qualitative shift in the forms and brutality of the conflict. (Simpson, Mokwena and Segal, 1992)

Simpson and his colleagues argue that, because of the preceding decades of state repression, violence had gained social sanction as a means of maintaining political power. So too, in the resort to armed struggle as a means of resistance, alongside the liberation movements' calls for 'people's war' and 'township ungovernability', was violence accorded pervasive social acceptability as a legitimate means of attaining change. Simpson, et al. (1992) suggest that this acceptance of violence was displaced into all spheres of life. A consequence of the legalisation of the previously banned political movements was to replace the externalised enemies (targets of violence) with the 'enemy within' (targets ever closer to home). The central challenge of the pre-election phase in South Africa was to maintain stability in the midst of the politics of transition. The dramatic increase in violence against women, children and the elderly (as well as in violent crime more generally) in the course of the 1990s, was, according to Simpson, et al. 'an accurate barometer of the social and political dislocation of a society in transition, as well as of the introversion of political conflict in South Africa' (1992).

One of the features of the negotiated transition in South Africa was the breakdown of exclusive and repressive apartheid structures responsible for administering various material aspects of life– notably systems of law enforcement, education and welfare provision - without immediate replacement by legitimate or effective alternatives. This 'deregulation of social control' (Simpson and Rauch, 1993) facilitated increased levels of violence and raised difficult questions about the usefulness of notions of 'democracy' and 'reconciliation' in the process of reconstruction and development (cf. Hamber, 1998 and b; Simpson, 2000).

A cursory examination of the sustained, and even increasing, levels of violence in South Africa - both during and after the negotiation process - demonstrates the error of assuming that the moment of the negotiated political settlement represents the transformation of South Africa to a 'post-conflict' society. In the re-building of fractured societies, attempts at reconstruction and reconciliation have to come to terms with the changing nature of conflict and violence, rather than the simple end to such conflicts. In particular, certain sustained features of marginalisation, impoverishment and relative deprivation remain at the root of ongoing criminal violence in South African society in the post-apartheid era. (Simpson, 2000)

One of the first multiparty negotiations in South Africa resulted in the 1991 National Peace Accord created to deal with the high levels of violence, which manifested from mid-1990 onwards, particularly in the provinces of KwaZulu Natal and Gauteng. The Peace Accord established local and regional multiparty 'peace committees' which comprised the police, army, local branches of political parties, churches and other local community groups. Despite their efforts, violence continued. The sustained levels of conflict shook public confidence in the negotiation process and in the leadership of De Klerk and Mandela; as no amount of agreements seemed to be able to bring it under control.

The transition was meant to heal the country's conflicts, and bring peace. But violence stalked it from the beginning. Even before detailed discussion of a settlement began, violence influenced the pace of negotiations – at times, delaying it, at others, ironically, speeding it up … . [T]he transition may not have created the violence, but it certainly worsened it and made it more difficult to end. As society hovered between the old state and the new, institutions broke down … if violence was to be contained, new structures would have to be built. And this could happen only when the transition was complete. (Shaw, 1994, pp. 182-184)

Against this context of escalating violence in the early 1990s, the future of the security agencies was a key issue. This was highlighted within the political negotiations over the new Constitution. The Convention for a Democratic South Africa (Codesa) negotiation forum was established in December 1991 and ran until May 1992. Part of its structure included a group dedicated to 'Law and Order' matters whose members were drawn from the various parties. When the nineteen parties began negotiating, the South African government, with its homeland progeny, had eleven police agencies, five different militaries, and various intelligence organisations. The African National Congress (ANC), on the other side of the table, had their 'armed wing', and an 'intelligence department', and the Pan African Congress (PAC) a small 'armed wing'. 'Non-statutory forces' was the term coined to describe the structures of these liberation movements, and also, later, to include the Inkatha Freedom Party's (IFP) various semi-military units. Codesa produced a 'Declaration' in which the nineteen participating parties agreed to an undivided South Africa, peaceful constitutional change, universal suffrage in a multiparty democracy, separation of powers and a Bill of Rights – 'in Codesa, the points of departure for the post-apartheid order were recognised, and the transition game began'. (Friedman and Atkinson, 1994, p. 18).

The Codesa talks collapsed in May 1992 for a range of reasons that go beyond the scope of this report. However, in the period between Codesa and the next set of formal negotiations, much work was done bilaterally between the ANC and National Party government; many of the key agreements on South Africa's future were reached in that period and merely needed to be formalized at Kempton Park, the venue for the second round of negotiations. Formal multiparty negotiations resumed on 1 April 1993, with a new structure for the talks. Emphasis was placed on devolving discussions to 'technical committees' whose deliberations would be shaped by bilateral agreements between the two main parties, the ANC and the National Party (NP) Government. By the middle of 1993, the forum had agreed that an Interim Constitution would be produced by the negotiators, to be followed by an election. The 'technical committees' would produce draft sections of the Interim Constitution, each of which would have to be presented to, and agreed by, the plenary of the multiparty negotiating forum. The technical committees of most relevance to this report were those on Police, on Intelligence, and on Violence.

Police reform was specifically discussed at the Kempton Park negotiation forum, and also at bilateral meetings between the ANC and NP, attended by SAP generals and academic experts associated with each party. The chief negotiators on policing issues were Nicholas (Fink) Haysom, an ANC lawyer and academic; and reformist SAP General Andre Pruis. The policing negotiations were fairly narrow and superficial, and less heated than could have been expected given the role of the police in enforcing apartheid and their failure to bring the violence under control. According to Shaw (1994), the tame-ness of the policing negotiations was a result of multiple factors:

• The ANC had no structure or capacity devoted to policing issues; it relied on academics to shape its police policy and negotiate the police arrangements in the new Constitution. They tended to be more free-thinking than regular ANC officials; and the absence of pre-existing ANC 'party lines' on policing facilitated creative thinking and negotiated solutions.
• There was little substantive disagreement between the ANC and the newly-reforming SAP on core principles of democratic policing.
• The SAP was obviously the country's 'only' police force - its only possible rivals were the 'homeland' police forces, which were its offspring. The 'sunset clause' agreement at Codesa, which had secured jobs for civil servants in apartheid's bureaucracies and ruled out the possibility of lustration, meant that the SAP's continued existence was effectively guaranteed. The negotiators only needed to agree on key aspects of institutional reform and their timing. No more radical options were possible.
• The SAP was concerned to present itself as an apolitical, professional police organization, willing to serve the government of the day in order to reduce the incentive for a new government to intervene in police matters in the future. Since Mandela's release in 1990, the SAP had been on a major internal reform drive, seeking to manage change itself in the hope that it would not have change thrust upon it later. SAP negotiators were therefore fairly amenable to ANC positions. (Shaw, 1994, p. 206)

The major agreements on policing were contained in sections 214-222 of the Interim Constitution. Most of these were concerned with some aspect of the central-vs-regional-control debate, which was one of the key issues contested throughout the negotiations process. The ANC held the view that policing should remain a national function (i.e. under the control of the central government) while the National Party (and others) preferred the idea of independent regional police agencies. The SAP Generals tended to share the ANC's view that management and reform of the large police organization would be more effective if centrally- controlled; and this led to some conflict between the SAP negotiators and their National Party principals. The outcome, in the interim constitution, was that South Africa would have a unified, single, national police agency, but with a level of provincial organization, in which provincial governments would have very limited say – a victory for the centralists.

One of the more detailed agreements at the multiparty constitutional negotiation forum was the provision for Community Police Forums (CPF's) – perhaps to balance the strong centralization - these grassroot structures would enable local community representatives to hold local police managers accountable. The SAP (in a department headed by negotiator Andre Pruis) had already embraced a version of 'community policing', the doctrine which had swept through the western policing world in the 1970s and 80s and were thus able to see the CPF model as a linear progression of their own reform initiatives. For the ANC, the CPF model promised an opportunity for its constituency to impact in a real, local and practical way on police performance.

Negotiating Amnesty

It has been argued that, were it not for the early political compromises on the question of amnesty, there may have been no negotiated settlement and no democratic election in South Africa (see for example Simpson, 2003). In this interpretation of history, the amnesty agreement is seen to have been a 'historical necessity' for the peaceful transition to democracy in South Africa. However, there was tension within the ANC over whether or not to also pursue the idea of a Truth Commission (Van Zyl Slabbert, 2000) to deal with past abuses. This ambivalence was perhaps motivated by a desire to downplay atrocities which had been committed by – or in the name of – the ANC itself; or by fears of what might be exposed if apartheid security agencies began disclosing the names of their spies within the ANC.

Amnesty – or, at least, indemnification from prosecution - was a pre-condition for the commencement of the negotiation process from 1990 as many of the ANC negotiators needed indemnity in order to return from exile and participate in the talks. In return, some form of amnesty had to be promised to the apartheid government's security forces, whose assistance would be required in running the crucial first election, and whose loyalty to a new (ANC) government needed to be secured, as some believed that the security forces had the potential to derail the entire negotiations process:

There is no doubt that members of the security establishment would have scuppered the negotiated settlement, had they thought they were going to run the gauntlet of trials for their involvement in past violations. (TRC Final Report, 1998, Vol 1, Ch 1, Para 22)

The failure of the Codesa working group on amnesty to come up with an agreement was perhaps one of the reasons for the failure of the Codesa talks in mid-1992 (Friedman, 1993). The initial positions of the major parties produced no common ground: neither a blanket amnesty, demanded by the NP, nor the prosecution of apartheid criminals, called for by the ANC, was possible. As Bell and Ntsebeza (2001) explain, a blanket amnesty was unacceptable to the ANC's mass constituency, which had 'fought and forced the concessions that resulted in negotiations' (p. 237). Likewise it was 'unworkable' to prosecute the apartheid senior ranks while 'trying to take over the apartheid state machine' (Bell and Ntsebeza, 2001, p. 237). The ANC and the National Party eventually agreed on a basic amnesty deal, based on a model proposed by Joe Slovo in his famous piece on negotiations published in the interregnum between the negotiations processes in late 1992. Slovo proposed a qualified amnesty, in which 'applicants would disclose the terms and extent of their crimes' (Doxtader, 2003, p. 134).

Notwithstanding significant discomfort and debate within each of the main negotiating parties, this model of an amnesty process combined with some form of truth commission was agreed, and its fundamental principles were outlined in the the Post-amble to the Interim Constitution which came into effect in April 1994. The amnesty agreement was binding on the new government, which was required to enact legislation to provide mechanisms and criteria for the granting of amnesty within a specified period after the first election. The question of whether and how to create a truth commission would also be left to the new government.

Police Reforms

1994: Democratisation, accountability and human rights

The government elected in April 1994 faced an awesome task in reconstructing its inherited criminal justice and security agencies into ones which would be acceptable to all South Africans and which could address not just their own inherited problems of prejudice and militarisation, but also the country's changing patterns of crime and conflict.

The transformation taking place is very comprehensive indeed: each state department is being radically restructured, its vision redefined, its staff adjusted to match the demographic realities of the country. All civil servants, including those in the justice system, have to learn, almost from the beginning, what it means to do their jobs in a democratic context. For the functionaries in the criminal justice system, that learning is proving to be very difficult, as they have to move from a confession-based system to an evidence- based system of criminal justice. (Scharf and Nina, 2001, pp. 4-5)

The first wave of criminal justice reform was driven by the imperative to establish a more democratic security system, as envisaged in the new (interim) constitution. The main emphasis in this period was on institutional design¹² and on creating a range of mechanisms to ensure greater accountability of the justice and security institutions. In the absence of a credible, professional police service, or expertise within its own ranks to 'take over' the management of policing, the ANC at least wanted to secure compliance from the justice and security agencies on human rights standards, the law, and the direction of the elected government.

The first tier of democratic control over the police is the appointment of government Ministers (Cabinet Members) responsible for the criminal justice and security agencies.¹³ The Ministers – elected representatives, accountable to Parliament – are able to give direction to the various bureaucrats heading the justice and security agencies. The South African system provides strong powers for the Ministers to impact on day-to-day issues, rather than only setting broad policy guidelines, and this has required significant re-negotiation of the boundaries between setting policy (the role of elected politicians) and 'political interference' in criminal justice institutions which require some degree of professional independence for the exercise of discretion. In respect of policing, which had been profoundly politicized under apartheid, senior police officers accommodated, quite pragmatically, the political imperatives of the new government, and many were (in this early period, at least) ardent supporters of the various efforts towards reconciliation, reconstruction and accountability. In the justice sphere, however, the new government encountered resistance from the ranks of the judiciary and the prosecution service – particularly around government efforts to change the appointment and composition of the judiciary and magistracy. The government was determined to change the race, gender and political composition of the top ranks of the judicial system, to better reflect the new democracy. The battles over appointments, conditions of service and the scope of discretionary decision-making in the judiciary, magistracy and prosecution service still continue – perhaps a healthy feature of a democratic society?

In respect of all the institutions and functions of the justice system, parliamentary oversight committees began functioning in 1994, and Parliament became a second tier of accountability and control over the police.

South Africa has more oversight bodies – whether state or quasi-independent – than most nations, but there are limits to their effectiveness. According to the [official] opposition [party, the] Democratic Alliance, the Ministers of Safety and Security, Intelligence and Defence have responded well to security sector questioning in Parliament – but [other small opposition parties] the [Pan African Congress] PAC and [African Christian Democratic Party] ACDP say too many questions are avoided. Mirroring the decline in the Safety and Security Secretariat, and reflecting a swing against 'human rights policing', the Safety and Security Committee (in Parliament) is reportedly becoming less vigorous in its probing of the executive and the police. (Sammonds, 2000, np)

Some parliamentary committees performed more authoritatively than others, as certain government departments were more amenable to direction and inquiry from Parliament than others (such as Intelligence), and some of the old officials (particularly in the police, intelligence and military) were past masters at making their activities and decision-making processes opaque to parliamentary oversight. A former staffer in the Safety and Security Secretariat was less than enthusiastic about the role of parliamentary oversight:

The parliamentary standing committee on Safety and Security has very little capacity to interrogate what is presented before it. Thus, complex police plans, statistics or performance targets can be presented by the police, but the weight of information and expertise lies with the presenters. That is not to say that parliamentarians cannot interrogate these issues, but that the devil is very often in the detail. (Shaw, 2001, np)

The third tier of institutional democratisation was the redesign of the civil service institutions responsible for justice and security. This included:

• Creating one national Department for Safety and Security, made up of the SA Police Service (SAPS) and the civilian 'Secretariat for Safety and Security';¹⁴
• Creating one national Department for Defence, made up of the SA National Defence Force (SANDF) and the civilian 'Secretariat for Defence';¹⁵
• Creating a National Prosecuting Authority (NPA);
• Restructuring of the Department of Justice to enable a separation between administrative functions and the magistracy;
• Creating various independent oversight or investigatory bodies, such as the Independent Complaints Directorate (to investigate complaints against the police), the Independent Inspecting Judge (to monitor human rights standards and complaints in prisons) and the Human Rights and Gender Commissions.

Another core imperative of the early period of police reform in South Africa was constitutionalism, with a strong emphasis on aligning justice and security systems with the Bill of Rights.

[T]he transition to democracy was a triumph for both legal reformism and a radical criminology imbued with strong communitarian ideals, over the conservative, racially-based approaches to crime that had dominated South Africa for most of the century. For legal reformists, the Constitution, with its Bill of Rights, was a considerable triumph. (Van Zyl Smit, 2001, p. 1)

The Human Rights approach was characterized by a great deal of 'visioning' - the development of visionary policy statements¹⁶ describing the goals and ideals implied in the new Constitution – and by the creation of numerous 'watchdog' oversight bodies to monitor human rights standards and investigate violations. As with other components of the transformation, this approach attempted to ensure that South Africa's criminal justice system conformed to standards appropriate for a modern democracy, with no explicit recognition of the problems created by South Africa's high-crime, resource-poor context. In the police service, elaborate programmes of Human Rights Training and new Codes of Conduct were implemented (for example, training on international human rights norms and standards, and training in how to deal appropriately with child victims and suspects). However, the police, prison officers, magistrates and prosecutors tended to see the new human rights dispensation as a threat to their professional discretion, as undermining of their experience and expertise, and as an obstacle to the effective administration of criminal justice. This argument was quickly formulated in the media and popular discourse as 'the Constitution means that the police can't do their job' and 'the Constitution gives more rights to criminals than to victims of crime', fuelling a popular rejection of human rights values in favour of more punitive approaches to suspects. These included popular support for the death penalty and use of lethal force by the police in dealing with suspects and a rise in recorded incidents of vigilantism.

Given the immense difficulties of transforming inherited state institutions and building formal mechanisms of democratic control, the emphasis of the ANC's policies during this first wave of police reform lay on strengthening oversight and accountability of the police. At the institutional level, numerous bodies were established to oversee the police and to receive complaints or investigate abuses:

• The Human Rights Commission
• The Gender Commission
• The Magistrates Commission (magistrates only)
• The Public Protector
• The Independent Complaints Directorate (police agencies only)
• The Independent Investigating Judge (prisons only)
• The Portfolio Committees in Parliament, Provincial Legislatures and the National Council of Provinces.

While the early emphasis was placed on the creation of external mechanisms for accountability of police and prosecutors, some internal initiatives were also undertaken, as part of the managerialist overhaul of the criminal justice institutions. These included:

• ? the introduction of a rigorous selection system for new staff;
• revision of basic training curricula;
• the development and introduction of human rights training curricula;
• the introduction of various 'Codes of Conduct';
• the implementation of a Special Service Order on the use of force in affecting arrest, intended to bring the regulatory framework relating to police use of lethal force more in line with the Constitution;
• the development and introduction of an anti-torture policy within the SAPS;
• the reorganisation and retraining of public order police units; and
• the introduction of the new, less than lethal weaponry, intended partly to assist police in reducing reliance on lethal force.

Notions of 'community policing', gleaned from contact with the international police fraternity, gained currency in enlightened police circles in the late 1980s, and similar discourses on 'community justice' and 'community prosecutions' resonated well with traditional and customary forms of justice. The first 'community policing' structures were introduced in the early 1990s, and the 1994 Interim Constitution contained a detailed requirement that the new police service should establish a Community Police Forum (CPF) at every police station to enable consultation with the local community. In April 1997, the Department of Safety and Security published the Community Policing Policy Framework and Guidelines. Developed through a consultative process over a three-year period, the Framework defined community policing in terms of a collaborative, partnership-based approach to local level problem-solving. As this was the first explicit expression of community policing as a methodology for reducing crime and improving the service provided by the police, the policy marked a watershed in the development of community policing policy in South Africa. Written retroactively in response to developments on the ground, the policy document was mainly intended to provide direction for police managers. The document therefore provided detailed step-by-step guidelines for establishing CPFs, a guide on change management, guidelines for demographic and local level crime analysis, the development of partnerships and local level problem-solving (see the SAPS 1997 Community Policing Policy and Guidelines booklet).

The five core elements of community policing in South Africa were defined in the policy document as:

• service orientation: the provision of a professional policing service, responsive to community needs and accountable for addressing these needs;
• partnership: the facilitation of a co-operative, consultative process of problem-solving;
• problem-solving: the joint identification and analysis of the causes of crime and conflict and the development of innovative measures to address these;
• empowerment: the creation of joint responsibility and capacity for addressing crime;
• accountability: the creation of a culture of accountability for addressing the needs and concerns of communities.

Policy guiding the implementation of community policing in South Africa has shifted substantially since 1994. Initially focused on establishing community oversight of the police, it has moved through enabling structured liaison and communication, to emphasising collaborative problem-solving and, more recently, to providing a framework for participatory and complementary local-level crime reduction. The state of CPFs varies dramatically in different parts of the country, with some now effectively defunct; though they served a useful purpose in rebuilding relationships between the police and the black majority in the early post-election period. The imperative towards 'community participation' and 'community accountability' was not confined to policing: a system of lay assessors was introduced in the magistrates' courts, to introduce representatives of the community into the adjudication of cases, and community-based corrections were introduced, although on a small scale. Little of the rhetoric of community-based justice was actually turned into policy or implemented in practice; however, as Scharf and Nina point out,

there is no country where civilian adjudication has been incorporated into the formal system without major problems … . At this stage of our transformation, and in the context of the current style of social organization in African townships and informal settlements, there is not yet the capacity within the state to manage and implement a formalized system [of community justice]. Nor are communities ready for a formalized system. (Scharf and Nina 2001, p. 70)

1995-1998: Prevention, prioritization and multi-agency collaboration

By the end of 1995, the government recognized that growing public concern about crime and rising levels of reported crime required new responses – no matter how 'transformed' the institutions of criminal justice became, they were unlikely to be able to deal with the volume of cases. The Department of Safety and Security was tasked with leading an inter-agency effort to find new solutions. In a split that has subsequently come to characterize the government's approach to crime, the SAPS developed a 'high-density' policing intervention strategy, which would focus on 'hotspot' areas and involve a return to more repressive police tactics (including roadblocks and cordon-and-search operations); while the civilian Secretariat for Safety and Security led an interdepartmental team to consider prevention strategies which would ultimately reduce the number of criminal cases entering the justice system. This latter venture resulted, in 1996, in the production of the National Crime Prevention Strategy (NCPS, available online on http://www.gov.za).

The NCPS strategy motivated a shift in emphasis from crime control to crime prevention; that is, a shift towards understanding crime as a social issue requiring a wide array of developmental and preventative measures, instead of just the traditional criminal justice responses. The NCPS comprised four approaches:

• Re-engineering the Criminal Justice System: aimed at providing an efficient and legitimate criminal justice system as the foundation for crime prevention, law enforcement, and the protection of human rights.
• Reducing Crime through Environmental Design: aimed at limiting environmental or situational opportunities for crime and maximising constraints by, primarily, ensuring that safety and crime prevention considerations were applied in new development programmes, and in the re-design and upgrading of existing programmes.
• Community Values and Education: aimed at harnessing community participation and involvement in crime prevention to ensure a positive impact on the way society engages with and responds to crime and conflict.
• Transnational Crime: aimed at addressing the enormous influence of international and regional criminal syndicates and involving improved border control, addressing cross border crime and enhancing regional co-operation (NCPS).

The NCPS outlined a multi-departmental approach to the prevention of crime, which envisaged high levels of co-operation and co-ordination of the activities of different government departments, as well as between the three spheres of government.¹⁷ In doing so, the NCPS – the touchstone of this second wave of justice sector reform - aimed to provide the means by which the police, other government departments, the private sector and the non-governmental community could consolidate their activities and increase their preventive effect.

Some of the reform efforts directly associated with the original NCPS include:

• Integration and automation of the entire criminal justice system, particularly with respect to information management;
• The development and implementation of a Witness Protection Programme;
• The development and implementation of a Victim Empowerment Programme;
• Revised procedures for the administration of bail;
• Completion of an official National Victim Survey in 1998;
• The initiation of a number of 'Safer City' programmes in metropolitan areas;
• Revisions to the juvenile justice system, and, in particular, the provision of secure care facilities for young offenders;
• Finding ways to meet the resource requirements of under-performing courts;
• The initiation of a number of programmes focusing on key crime categories including vehicle theft, narcotics and cross-border crime

Following on from the NCPS, the 1998 White Paper on Safety and Security advocated for targeted, multi-agency crime prevention strategies that would focus on offenders and victims and the environment in which they live, as well as on the particular 'root causes' of particular crime types. This approach was characterized as social crime prevention. Neither of these prevention policies was ever really implemented in practice, at least not by the government's security and justice agencies. The climate of increasing levels of crime and public hysteria has made long-term, developmental and preventive approaches decidedly unpalatable.

[By 1997], there seemed to be a real increase in serious crime, and … there was certainly a massive increase in popular punitiveness … the primary response of the government has been to water down the first wave of constitutionalising legislation, while at the same time to push through the harshest possible new criminal justice legislation that it thinks it can get away with in constitutional terms. (Van Zyl Smit, 2001, p. 225)

The implementation of the NCPS in fact never matched up to its original motive – to initiate prevention efforts as a way of reducing the burdensome caseload which was constantly mitigating efforts at institutional reform of the criminal justice system. Rather than prevention becoming a watchword, the theme of this wave of reform became co-ordination, as part of the desperate attempt to 'fix' the criminal justice system so that it could deal with the massive numbers of victims and suspects, and thereby deter other potential offenders.

The only aspect of the original NCPS that was seriously pursued was that dealing with the 're-engineering' of the criminal justice process and, even then, the emphasis was on co-ordination and reducing duplication, rather than fundamental system change. This aspect was led by Business Against Crime (BAC), a consortium of big business interests who had decided that the severity of the crime crisis required their direct intervention, given that their traditional tactics of advocacy or critique of government had done little to improve the situation. The initial attempts to integrate, rationalize and streamline the administration of criminal justice under the auspices of the NCPS provided the first significant entry point for management consultants, who later came to shape and drive the entire system reform effort. At this early stage, some of the tensions between the need for an 'efficient' criminal justice system, and for a legal process which respects and promotes human rights, became evident. Problems around the administration of bail, prison overcrowding, and the treatment of juvenile suspects threw these contradictory imperatives for criminal justice reform into sharp relief.

Managerialism, service delivery and the partnership with business

The third wave of criminal justice reform in South Africa was dominated by management consultants and businesspeople, who argued that private sector 'efficiency' approaches should be adopted to solve problems in the administration of criminal justice. The vehicle for this approach was the 'Integrated Justice System' (IJS) initiative, which had grown out of the original NCPS work on re-engineering criminal justice procedures. This technicist approach resonated with developments in criminal justice administration in the first world – the 'value for money' drive and the emphasis on customer-oriented public services which had been well under-way in Europe and the US for over a decade. As such, it tended to appeal to donors and also fitted well with the ambitious tone concerning public sector reform that was set by the second democratic administration.

The IJS initiative and other managerial strategies adopted at this time (from 1997 onwards) were primarily focused on law enforcement and the administration of justice, and made little attempt to engage with the social dimensions (neither causes nor effects) of crime and victimization. They appeared to be underpinned by the naïve belief that if the system were 'fixed' and functioning well, it would be able to deal with any amount of crime.

The first National Victimisation Survey in South Africa (Statistics South Africa 1998) revealed that only 26 percent of respondents thought that the effectiveness of the police had improved since the 1994 transition, while 32 percent of respondents thought that they had stayed the same and a majority of 42 percent thought they had become less effective since the 1994 democratic elections. The majority of respondents (in this rather small survey) felt that the police were performing as badly, or worse, than they had under apartheid. Issues of effectiveness and customer service were problems that business people were familiar with, and through Business Against Crime, they were willing to lend their experience to the project of criminal justice reform. In 1998 a senior business figure – Meyer Kahn from the South African Breweries - was appointed 'Chief Executive Officer' of the SAPS to lead the process of making the police more efficient.

Among the managerialist strategies adopted in this period were:

• The development of systems for automated fingerprint identification, which would reduce delays at various stages of the process
• Design and implementation of case management systems, in courts and between police and prosecutors, to reduce delays and to increase prospects for successful prosecutions
• Improved budgeting and financial management
• The introduction of pre-trial services and procedures to manage suspects and witnesses, reduce delays in court rolls, and reduce numbers of awaiting-trial prisoners
• Changing the administration of bail and thus reducing the numbers of prisoners awaiting trial for long periods
• Occasional release of certain categories of prisoners to reduce overcrowding
• Policy or administrative guidance to encourage withdrawal of charges (or of decisions to prosecute) in order to reduce the volume of cases in court, and to increase the rate of successful prosecutions
• Outsourcing of various functions within the criminal justice process, and of some prison-building.

The 'Integrated Justice System' programme became the centerpiece of criminal justice reform, and significant government funds were allocated to it. In a version dated early 2001, the 'IJS 2000 Plus' strategy consisted of the following elements:

• An (information-technology-based) architectural plan clearly defining the integration of the justice system.
• The creation of a virtual private network for the four departments, to provide the necessary security and stability of the network.
• Providing access to electronic (IT) infrastructure in all four criminal justice departments.
• A common method of identifying persons, cases and key business requirements in each of the departments. Central to this is the Automated Fingerprint Identification System (AFIS).
• A court and case management system, that includes docket management, case tracking facilities, event notification, inmate tracking and resource scheduling.
• Data warehousing facilities, supported by business intelligence functions (IJS 2000 Plus, Dept of Justice, 2001).

The emphasis on technology is significant, if somewhat difficult to imagine in practice, in South Africa's highly inequitable context where some police stations and courts do not have electricity, and where one third of members of the Police Service are illiterate.

These managerial and technology-heavy approaches to institutional reform and capacity-building were not unique to the criminal justice sector; they fitted well with trends in public service reform in South Africa more generally. However, after repeated attempts at packaging these reforms as 'the answer' to South Africa's crime problems, the politicians and police leaders of the second post-democracy Cabinet realised that this type of reform was not impressing the voting public. The media, victims and potential victims of crime were ever more vocal, and, although there was some satisfaction with the evident improvements in the criminal justice system, public demands tended to focus on harsher forms of deterrence and punishment.

2000: Crackdown!

By 2000, the South African criminal justice system was not getting any better at dealing with crime, despite new institutional design, basic adherence to the Constitution, improved accountability and better technology.¹⁸ The pressure to deal with the massive caseload provoked a new set of reform initiatives. This period of criminal justice reform was characterized by the imperative to prioritise the most serious, most prevalent or most politically damaging crime problems, and simultaneously, to 'crack down hard' on the factors that cause or facilitate those problems. The Police Service and their new Minister led this approach, as they had already experimented with various prioritization strategies,¹⁹ and the new approach came to be known as 'Operation Crackdown', after the name of a series of heavy policing operations. Broadly put, the 'get tough' policy strategy had two primary objectives:

• To reduce or 'stabilise' crime in a limited number of high crime areas,²⁰ and improve local performance of courts and police stations in those areas
• To improve public confidence in the police and to improve public perceptions of safety.²¹

The new police strategy had three main components:

• A geographical approach - high crime areas (particularly where violent crime was a problem), were clustered into 'crime-combating zones' and then targeted for aggressive, high-density street-level policing.
• An intelligence-driven focus on organised crime syndicates operating in the high-crime areas, which aimed at disrupting syndicate activities by arresting syndicate leaders and 'runners', and by closing down the flow of goods and markets for these goods.
• Medium-term social crime prevention initiatives aimed at addressing the social- economic and development deficits conducive to high rates of criminal activity in the high crime areas. Many of these prevention programmes - like poverty alleviation and urban renewal - would be implemented by government agencies outside the justice and security cluster. (see speeches by Minister Steve Tshwete in early 2001)

Implemented since March 2000, this framework continues to shape government policy and the implementation of reforms within South Africa's criminal justice system. Contest for the leadership of the system is largely a thing of the past – by 2001, the police were clearly in the driving seat.

Improved service delivery was recognized as the key determinant of government legitimacy in the period after the second democratic election in 1999. Most of the criminal justice agencies implemented 'Service Delivery Improvement Programmes' in accordance with public service-wide policies implemented after 2000. A large-scale public survey produced by the Institute for Security Studies (ISS) assessed the impact of the Police Service Delivery Improvement Program (SDIP) in 2000.²² The ISS survey was conducted with a sample of 45 of the 219 'priority' police station areas – the highest crime areas in the country – and it consisted of:

• a general community perception survey of over 13 000 respondents,
• an exit poll of over 2000 people who had been into one of the 45 police stations (67 percent of whom had reported a crime there) and were questioned as they left the stations, and
• a follow-up survey of people who had reported a crime to one of those police stations within the preceding three months (Pelser, 2000, pp. 7-10).

The survey took place in October 2000, five months after the 'Crackdown' police strategy began to be implemented, and five years after the establishment of the new South African Police Service. It found that: members of the public who had visited one of the priority police stations were overwhelmingly satisfied with the attitude of the police officials and the treatment they had received from them;²³ most respondents said that the police treated them well and responded in a helpful, respectful and supportive way.²⁴ Over 70 percent of respondents who had reported a crime to the police were happy with the service they received from the police at the time of reporting the incident (Pelser, 2000).

Following the police's lead, the Department of Justice prioritized certain court sites for major improvement, and certain types of crime for 'fast-tracking' through the courts. The National Prosecuting Authority also adopted the approach of prioritizing certain cases for prosecution, and its Directorate of Special Operations (also known as The Scorpions) was seen to be 'getting tough' with various organized crime bosses and syndicates. The judiciary and magistrates were under enormous pressure to impose heavy sentences, and, as a result, the crisis in the prison system continued to deepen. By the end of 2001, there were approximately 170 000 inmates in South African prisons, which had an approved bed capacity for 101 000 prisoners (IJS Plus, 2001).

Government discourse on 'community participation' was deeply imbued with ambivalence and cynicism by the time the Crackdown approach emerged. In 2000, the Minister (Tshwete) announced the SAPS's intention to 'rationalise existing government liaison structures with communities' creating only one local structure at each locale for community liaison with all the criminal justice departments (12 Feb 2001, Parliament), and then, a few months later, issued draft regulations aimed at reviving CPF's. In respect of courts and the Justice department,

talk of community justice has continued, but there have been only patchy legislative initiatives to involve communities significantly in the formal criminal justice system or to incorporate their informal processes into the overall crime prevention strategy. (Van Zyl Smit, 2001, p. 226)

On the one hand, community consultation, participation and accountability were seen by the managerialist police reformers as onerous and obstructing the development of a more streamlined and effective criminal justice 'machine'; yet, on the other hand, the 'Crackdown' approach relied on populist political support.²⁵

2002-2003: Settling in …

There appears to be some settling of the discourses of police reform in South Africa, perhaps the result of a decade of democratic government. The 1998 White Paper on Safety and Security was envisaged as a five-year policy framework, and is due to be replaced by a new policy document; and revisions to the 1995 Police Act are expected some time after the 2004 General Election. After that election, the possibilities for policy shifts will increase, as a new Minister may be appointed. At the close of the first decade of democracy, the following themes defined SAPS strategies:

• An emphasis on getting the basic aspects of democratic policing functioning well; including adherence to human rights standards.
• Commitment to some form of police-community partnership – most recently described as 'sector policing' – but variations in how this is envisaged.
• Capacity for tough 'fire-brigade' policing operations such as Crackdown, in the highest-crime areas.
• Encouraging communities and other government departments (such as Education, Welfare, Health and Housing) to implement programmes aimed at strengthening social capital and building community cohesion, thus contributing to the long-term prevention of crime.
• An emphasis on the use of volunteers to supplement limited human resources inside the police organization, such as Police Reservists and volunteers from Community Police Forums.

The Truth and Reconciliation Commission and the SAPS
1990-1994: Uncovering some truth before the onset of the Commission

Throughout the negotiations period, the ANC had made clear its intentions to reform the police gradually, rather than radically; and that, although the ANC would not tolerate abuses of human rights, it would not victimise perpetrators of such abuses committed in the past if the perpetrators abided by new government doctrine.

Simpson (2000) describes the fact that the new Government inherited its criminal justice institutions, practices and the vast majority of its personnel - lock, stock and barrel - from its anti-democratic predecessor as a 'liability embedded in the compromises necessary to achieving South Africa's negotiated peace' (Simpson 2000).²⁶ This meant that the new democratic government effectively inherited - largely intact and with their existing personnel - institutions of state from the former regime; and, with them, a legacy of popular mistrust and a crisis of legitimacy for the criminal justice institutions.

In order to build new legitimacy, the police institution had to deal with officials who had been involved in apartheid atrocities. The issues around the amnesty provision contained in the postscript to the Constitution revolved around the vexed question of exactly how to deal with perpetrators of past abuses, who, by virtue of the agreements reached in the negotiations, would retain their jobs in the new South African security establishments. The South African TRC model did not provide for a blanket or unconditional amnesty, but an amnesty available to perpetrators only in exchange for full disclosure.

Many of the 'gross violations of human rights' committed by members of the SAP were known prior to both the 1994 election and the commencement of discussions about the form of the amnesty process. Aside from the public displays of excessive force that often characterized police handling of protest gatherings, torture was commonly reported by ex-detainees, and a number of high-profile activists had died in police custody. Even the less visible side of police abuses had already begun to be exposed: for example, five years prior to the election, former policemen Butana Nofomela and Dirk Coetzee had revealed the existence of the Vlakplaas C1 Unit – a secret assassination unit (hit-squad) of the SAP Security Branch.

The Goldstone Commission was created as part of the 1990 National Peace Accord, and statutorily established in 1991²⁷ to investigate the ongoing violence. During the course of its various investigations in 1992 and 1993, the Goldstone Commission uncovered some evidence of sinister extra-legal activities by the government's security forces. At the same time,

Joe Mamasela, a former ANC operative-turned-policeman, decided to confess, not only to 44 murders he had committed, but also to reveal everything he had learned in more than a decade as a [Vlakplaas] death squad member … . [In response, the SAP] Generals and other branches of the police severed all links with Vlakplaas. The military did the same. Vlakplaas and its commander, Eugene De Kock, would be sacrificed. The records at Vlakplaas had already been incinerated,²⁸ but there were a myriad of ties throughout the security establishment … . So … the government ordered perhaps the greatest destruction of records in the country's history … . In little more than six months in 1993, while the political parties of the apartheid state negotiated with the representatives of the liberation movements [at the multiparty negotiation forum in Kempton Park], some 44 metric tons of records from the headquarters of the National Intelligence Service alone were destroyed. (Bell and Ntsebeza, 2001, pp. 7-8)

Goldstone's team raided military intelligence premises in Pretoria in November 1992 and found evidence of secret 'dirty tricks' projects. The National Party government responded by firing a number of senior defence force generals (instead of protecting them, which is how the National Party had traditionally responded to exposés of its security forces' secret activities), and Eugene De Kock realized that 'the game was pretty much up' (De Kock and Gordin, 1998, p. 244). At the end of April 1993, De Kock and 83 other police officials attached to the Vlakplaas unit took 'early retirement' and large severance payouts from the SAP.

In early 1994, a group of De Kock's former subordinates gave damning evidence to the Goldstone Commission about the activities of the SAP's Vlakplaas unit. They were taken to Denmark in an extraordinary witness protection scheme arranged by Judge Goldstone. On 18 March 1994, just weeks before the country's first election, the Judge announced the existence of a 'horrible network of criminal activity' that operated within the military-security establishment. One week after the election, De Kock was arrested by Goldstone's investigators. His arrest sent shockwaves through the security branch and senior leadership of the SAP – the likes of De Kock had previously been shielded from prosecution by the National Party government; and now the country had new rulers, a Truth Commission was in the offing, and De Kock could provide damning evidence against many of his colleagues.

I could have escaped anytime I wanted to, but I had decided well before my arrest to stay and face the music. This, clearly, was something I was going to have to do largely on my own. I heard that Police Commissioner Johan van der Merwe was telling all his colleagues to keep well away from me. Some people see my trial as the first time the true extent of security force covert actions became known; others see it as a vindication of the Goldstone Commission's findings and a triumph of justice. I see it as two years of betrayal – first, by the state that had given me my orders, and then by my friends, who lined up to testify against me … . The state chose to give indemnity from prosecution to many of my men, so that a bulldozer of a case could be assembled against me – and, in the process, allow other men, just as guilty as I, to laugh in the face of justice. But we at Vlakplaas, and in the other covert units, are by no means the guiltiest of all. That dubious honour belongs to those who assembled us into the murderous forces that we became, and which we were intended to be all along. And most of them, the generals and politicians, have got off scott-free. (De Kock and Gordin, 1998, p. 249)

For the purposes of this research report, a handful of current and former police officials were interviewed about their perceptions and experiences of the TRC and its impact. Some of the interviews were conducted in Gauteng in 2000, and some in the Western Cape in 2003. The sample is comprised of black and white men, most of whom were still serving in the SAPS at the time of the interview, all of whom served in the former SAP. Interview formats were informal and differed in each setting, so as to best enable the respondents to feel comfortable discussing difficult questions. The small size and random nature of the interview sample prevents any quantitative conclusions from being drawn. However, the qualitative narratives of the respondents offer insights into their experiences of the TRC. Their perceptions are presented alongside those of various other role players within civil society and the non-governmental world, who were also interviewed across the 2000-2003 time period for their perceptions of the TRC and policing in South Africa.

Most of the police respondents first heard about the TRC in 1991-2. Some of them learned about it at official SAP briefings held for Security Branch personnel, some through the 'grapevine' inside the police organization, and some through the media. Their common initial reaction was fear.

I was concerned for a witch-hunt, but later I realized that it was no personal threat to me – it was just an opportunity for the ANC to remove some specific individuals. (former SAP Security Branch member, interviewed November 2003)

We were afraid of a witch-hunt. Two of my colleagues died in trying to clean up before the TRC. (former SAP detective, interviewed November 2003)

The former police officers spoke of their initial fear of being publicly exposed, the truth being told about some of the things they had done. Although they believed they had acted under orders in order to fulfill their duty to the apartheid state; in hindsight, some doubts crept in about whether or not their actions were justified or would be deemed socially acceptable.

The pre-election period was a difficult time for the police. They were required to account to a variety of new masters (such as the Goldstone Commission and the various other Peace Accord structures) and to defer to the ANC, who were the government-in-waiting, but just previously had been the hated target of apartheid policing. The negotiation process was unstable, and hence their futures were not clear. The idea of a TRC was mooted from 1990, but only finalized just before the 1994 election. The election itself would hinge partly on effective policing, a huge challenge for the police. At the same time as managing the worst community violence ever experienced in South Africa, they were also expected to practice new doctrines of minimum force and community policing. Many police members took medical discharges for 'stress', and even more took retrenchment packages.

1994-1995: Preparing for the TRC - Inside the police

At the same time as the new police service was being created, the debates around the shape of the truth and reconciliation process were being thrashed out in Parliament and in quiet discussions inside the security forces. The legislation for creating the vehicle for the amnesty and truth-recovery processes was legally required to be finalised within a year of the first election (this was completed with the passing of the Promotion of National Unity and Reconciliation Amendment Act, number 87 of 1995). However, the arrest of Eugene De Kock in 1994 significantly changed the environment in which the final form of the Truth and Reconciliation Commission was debated:

The demand for a blanket amnesty continued to be voiced by the [National Party] NP as it came under increasing pressure from its own constituency. The police and the military feared that they would end up the scapegoats. They began to point publicly to the 'collective responsibility' of the political establishment. Individuals, they argued, should not be blamed for carrying out their orders, even if only under 'implied authority'. The naming of names [of perpetrators of gross human rights violations under apartheid] was of particular concern. … The police had more to fear from the naming of names, since they were at the forefront of the individualised brute maintenance of the system. They handled more of the domestic torture, most of which did not result in death [ie there were survivors who could give evidence about their torturers]. … Under pressure from their underlings, the generals of both the military and the police sought to put pressure on their political masters. They hinted in their submissions [on the early drafts of the TRC Act] that they might really be forced to tell all, to disclose who had given which orders and when. (Bell and Ntsebeza, 2001, pp. 240-241)

The spectre of the TRC loomed large for police officials in the immediate post-election period. Together with the Constitutionally-required amalgamation of the SAP with the ten homeland police forces to create the new South African Police Service, it was probably one of the major drivers of police co-operation with the new government in 1994: the police did not know what might happen to them once the TRC began operating and after the new SAPS was created. The threat of possible prosecutions or lustration generated unexpected compliance from the police leadership with the policies of a new government which had previously been its political foe.

The appointment of George Fivaz as the first National Commissioner of the new SAPS at the beginning of 1995 was a crucial moment in the police reform process. Although he had been a senior police officer in the SAP, Fivaz had never been involved in political policing or human rights abuses, and he was Nelson Mandela's direct appointee. In his first public announcement, he reinforced one of Mandela's key messages to the police - to have faith in their new political leaders:

I would like to send a very special message to the members on the ground. The management team has the fullest confidence in the future, and is committed to improving your working conditions. We need your support. We wish to stress again that no member of the SAPS has anything to fear. We look forward to a bright future (JG Fivaz - first statement as National Commissioner of SAPS, 29 January 1995).

At the time of his appointment, Fivaz emphasised the need to make the new SAPS legitimate and acceptable in the eyes of the majority of citizens, and for the SAPS to make a 'clean and definite break with the past' (first statement as National Commissioner of SAPS, 29 January 1995) - which was to remain one of the themes of his leadership. At this time, the establishment of a Truth and Reconciliation Commission was a foregone conclusion, the new SAPS had been set up, and a number of senior apartheid-era police generals, including van der Merwe, had retired. The ANC government took an opportunity to clarify its position on previous abuses by police officials. Mandela insisted that his government was opposed to, and had no intention of conducting, a witch-hunt against the police as a result of activities arising from orders given to the police by the apartheid regime. He urged police officers not to dwell on possible investigations by the planned TRC, and instead to get on with the job of law enforcement and community policing (Address to the Opening of Parliament, 17 February 1995).

In the latter half of 1994, then-Minister of Safety and Security, Sydney Mufamadi, designated one of his advisors – Peter Gastrow – to assist him in preparing the SAPS to engage meaningfully with the TRC process. This was well in advance of the passing of the TRC act and the commencement of operations of the Commission, indicating the seriousness with which the police's political masters were taking the issue of the TRC. Gastrow was formerly the Chairperson of the Transitional Executive Council Sub-Council on Law, Order, Stability and Security; and the Vice-Chair of the Peace Accord's 'Police Board'. He had entered the peacemaking politics of transition from the ranks of the Democratic Party – and was therefore seen as relatively neutral because he was not associated with the ANC. Mufamadi himself had a long record of involvement in the National Peace Accord and in the pre-election negotiations to reduce violence; as well as a personal interest in truth–seeking about the disappearance of one of his brothers during the apartheid era.

Mufamadi was concerned to ensure that the new police service demonstrated a constructive and positive attitude to the TRC, and that as many police officers as possible went to the Amnesty Committee to make disclosures about their activities in the past. He and Gastrow worked closely with new National Commissioner George Fivaz to develop good co-operation from the SAPS with the TRC. However, there were also countervailing forces at work; most publicly the 'Foundation for Equality before the Law', an association of former senior officers in the SAP and South African Defence Force (SADF), whose mission was to protect the interests of apartheid security force personnel during the TRC process; especially as they felt that the National Party had abandoned them and was letting them take the fall for all the evils of apartheid. Inside the new SAPS, the former Security Branch personnel also organized began to organize their responses to the proposed TRC. Informal meetings were called to prepare collectively for the proposed commission and to foster camaraderie and solidarity among the once-elite security cadre. So, when the legislation establishing the TRC was passed in July 1995, its Commissioners appointed in December, and the TRC began operating early in 1996, the top echelons of the police organization and its most notorious members were relatively well-prepared.

During the TRC: Telling the truth about apartheid policing

The TRC Act was passed in June 1995, and the TRC began operating in early 1996. Its work was conducted through three Committees:

• The Human Rights Violations (HRV) Committee
• Amnesty Committee
• Reparations and Rehabilitation Committee

One of the most significant features of South Africa's TRC process was its provision of amnesty in return for full disclosure about abuses committed under apartheid. Much has been written elsewhere²⁹ about the merits and flaws of the South African amnesty model and the manner in which it was practiced and this will not be reflected on here. Instead, this section will examine the amnesty process as experienced and engaged with by apartheid's police. Earlier within this report, the amnesty deal agreed by the negotiating parties and enshrined in the Constitutional Post-Amble was described. The TRC Act of 1995 (Promotion of National Unity and Reconciliation Act) outlined, in more detail, the mechanisms and procedures that would govern the amnesty process. There were various amendments to the Act; most of which revolved around shifting timeframes for the TRC's lifespan and the amnesty period.

This section of the report focuses on the work of the Amnesty Committee and the impact of the amnesty process on the police institution during South Africa's transition. The amnesty arrangements were as follows (Promotion of National Unity and Reconciliation Act, 1995):

Applications for amnesty had to fall within the following parameters:

• for acts committed between 1 March 1960 (the month in which the Sharpeville massacre took place and the liberation movements were banned) and 10 May 1994 (the date of the inauguration of the first democratic government),
• the incident had to have been associated with a political objective (i.e. common crimes were excluded),
• the applicant had to give a full and truthful account of the incident (i.e. make full disclosure).

Applications had to be made on prescribed forms and witnessed by a Commissioner of Oaths. Applicants were required to provide the following information:

• personal details
• political or other affiliation, or employment by the state (eg SAP)
• particulars regarding the incident
• particulars regarding victims
• particulars regarding the political objective that was being pursued
• whether any benefits had accrued to the applicant as a result of the incident
• whether the act was committed in execution of an order or with implied or express authority
• particulars regarding prosecution or civil proceedings related to the incident.

More than 7000 applications, relating to over 14 000 different incidents, were submitted to the Amnesty Committee before the final deadline on 30 September 1997. (TRC Final Report, 2003, Vol. 6:1:2, p. 24). The Committee developed the following seven-stage process for dealing with applications:

1. Administrative perusal of the application to make sure that forms were properly completed.
2. Perusal by evidence analysts to establish whether the application fell within the various criteria required by the Act; and whether it was likely to require a public hearing or could be dealt with administratively in chambers.
Conducting further investigations when the cases did fall within the criteria.
3. Application dealt with in chambers or at a public hearing. Later in the process, the applications were 'grouped' for purposes of efficiency, and one of the five 'groupings' used was 'members of the former security forces'.
4. Arranging a public hearing, if necessary. This included prior notification to all interested parties.
5. Conducting a public hearing.
6. Delivery of a written decision and notification of all concerned parties about the decision (TRC Final Report, 2003, Vol. 6:1:2, p. 24).

Although there has been much debate on the usefulness of the disclosure-for-amnesty device as a way of seeking truth about apartheid violations, the impact of victims' testimony on the process of police reform has not been examined. The TRC proceedings were broadcast daily between early 1996 and mid-1998, the period of the most intensive reform of the police institution. The intense daily media coverage of the TRC proceedings revealed to many ordinary South Africans the extent of police involvement in apartheid atrocities. Much of the detail was revealed in the 'victim hearings' convened by the HRV Committee which gave victims of apartheid violations an opportunity to tell their stories. The broadcast of this process proved to have an enormous impact on the public psyche, perhaps greater than any other aspect of the TRC. However, against the backdrop of rising crime and declining conviction rates, many police officers believed this material did further damage to public confidence in the police

The TRC almost destroyed the image of the police because when it was revealed that policemen were involved in human rights violations, some people thought that it was everyone in the SAP. (Station Commissioner, Katlehong, Interview, April 2000)

[The TRC] has destroyed the image of the police … by digging up old wounds. (SAPS officer in focus group interview, East Rand, April 2000)

The TRC revelations have contaminated the new SAPS which we were struggling very hard to build … I do not think any of the police involved [then] are SAPS members [now]. (SAPS manager, quoted in Mafokoane, 2001, unpublished)

Although I agree that the TRC has done damage to the image of the police, I still believe that we had to go through that period. We had to bring everything into the open and then start anew. It's no use forgetting the past and try to continue from there. (SAPS officer, quoted in Mafokoane, 2001, unpublished)

It gave us a lot more insight into what happened. The police were no longer heroes. (former SAP Riot Unit member, interviewed Western Cape, November 2003)

My profession lost a lot of credibility and my family and friends were shocked that I was part of that system. (former SAP detective, interviewed Western Cape November 2003)

When asked how the overall TRC process impacted on them personally and on their families and social lives, the small number of former SAP officials interviewed for this study shared the following reflections:

It opened up a part of the police I didn't know much about – it was embarrassing to have been part of the police at that time. (former SAP member, interviewed Western Cape, November 2003)

My kids had a horrible time and my family distanced themselves from me. (amnesty applicant, former SAP Security Branch member, interviewed Western Cape, November 2003)

My wife was concerned that I was as bad as the others, and she still is, to this day. (former SAP Security Branch member, interviewed Western Cape, November 2003)

Ordinary citizens told their stories about the police at victim hearings of the Human Rights Violations Committee, and police officials told their versions in their amnesty applications. Some efforts were also made to submit 'official versions' of police history to the TRC. The National Party made two submissions, which dealt with the role of the SAP in supporting its apartheid policies and blamed some of the violations committed by the police on renegade units or individuals. The Foundation for Equality Before the Law, a network of former members of the SAP, fronted by the old leadership of the SAP, also made a submission to the TRC. This version of apartheid police history portrays the police organization as merely the instrument of its political masters, and ascribes all abuses to the doctrines of the National Party and the context of 'war' in which the police felt they were operating at the time. This submission was subsequently privately published and marketed as a book, under the authorship of former SAP General Hermann Stadler (The Other Side of The Story, 1997).

Police officials who had served under the old SAP felt abandoned by their leaders and angry about the lack of National Party accountability for policing policy and practice during the apartheid years. Giving evidence in mitigation of his sentence at the end of 1996, De Kock expressed the anger felt by many of his former colleagues and 'decided to let rip, pointing my finger firmly in the direction of PW Botha, Louis Le Grange and the Generals who had given me my orders. … I savaged F W De Klerk as best I could, for deserting the men of his security forces in their hour of need, we, who had kept him and his party in power' (De Kock and Gordin, 1998, pp. 273-275). The small sample of former SAP officials interviewed in the Western Cape in 2003 had similar views about National Party politicians and the SAP leadership:

They failed to take responsibility for the actions of the foot-soldiers they commanded. (former SAP Security Branch member, interviewed Western Cape, November 2003)

The politicians and generals sold us out. (former SAP Security Branch member, interviewed Western Cape, November 2003)

They did not support us. A lot of us felt destroyed. (former SAP Security Branch member, interviewed Western Cape, November 2003)

Nobody held them accountable for anything. (former SAP management services official, interviewed Western Cape, November 2003).

It proved to us that we were abused by the people in power. (former SAP detective, interviewed Western Cape, November 2003)

Except [former SAP Commissioner] Van Der Merwe, there was no effort to support Security Branch members. By supporting us after 1994, they would have exposed themselves and threatened their own careers (former SAP Security Branch member interviewed Western Cape, November 2003).

1996-2001: Amnesty applications by police officials

The Amnesty Committee convened its first public hearing in May 1996. The significant media coverage that amnesty hearings received disguised the reality that only a small proportion of cases were dealt with at public hearings – the vast majority (5489 of the 7115 applications) were dealt with in chambers (Ernest, 2003). The hearings – at which 2548 applicants made 1626 applications - were an important 'public face' of the Amnesty Committee's work. The hearings took place in front of a 'panel' of adjudicators drawn from the Committee – all lawyers, some of whom were judges. The Committee

endeavoured to steer a middle course between a purely inquisitorial and an adversarial procedure at its hearings. The guiding principle was to allow every interested party the fullest possible opportunity to participate in the proceedings and present a case to the panel. Every party that participated in the hearings had the right to legal representation, and even those who were indigent were always afforded some form of legal representation. (TRC Final Report, 2003, Vol 6:1:2, p. 49)

There has been some criticism of the TRC for 'juridicalising' or over-legalising the amnesty process and ultimately resembling a court procedure (Simpson, 2002). This criticism came largely from the victim constituency. The police applicants, by contrast, relied heavily on legal representation and on the formal nature of proceedings to afford them some protection in their journey through the TRC process. Legal fees for most of the police applicants were paid by the (new) government, in line with longstanding police practice.

In an attempt to evade the amnesty process, former members of the SAP Security Branch put together a 'collective application' for amnesty, on behalf of the entire group of Security Branch personnel. This was similar to efforts made by other parties – such as the ANC – to submit 'collective applications' rather than go through the amnesty process in the individualised manner envisaged by the legislation. Among the small sample of seven former SAP Security Branch members interviewed in 2003 for this research, five had been represented in the 'collective application'. One was not included because his Security Branch work had been focused outside of South Africa (and foreign acts were not covered by the TRC); the other was excluded because he was seen as a 'sell-out', no longer part of the brotherhood of the Security Branch. The TRC refused to consider the 'collective applications', thus presenting individual police officers with a personal choice about whether or not to apply and confess.

Few police officials voluntarily came forward to seek amnesty during the early period of the TRC's operations, but a significant (if small) number of applications for amnesty – mainly from police officers - was received by the TRC at the end of 1996, after judgment was given in the De Kock trial. Somewhat unexpectedly, perhaps, De Kock himself subsequently applied to the TRC for amnesty, submitting a 1200-page application in which he set out to show 'that I had not acted alone out of some depraved personal motive, but that my orders had come from the highest political levels' (De Kock and Gordin, 1998, p. 270). According to one staffer of the Commission,

De Kock made an extraordinary contribution to the TRC process, a point that is acknowledged in the Commission's final report. Most of the security police and, certainly all the police Generals, who applied for amnesty, were forced to do so as a result of De Kock's disclosures to the Commission. (Godobo-Madikizela, 2003, p. 137)

Although the amnesty process generates a great deal of discussion among those who study and reflect on South Africa's TRC, it affected only a small proportion of police members directly. As the following table shows, of approximately 1500 individuals³⁰ who applied for amnesty, approximately 300 applicants came from the security forces (a term denoting both the police and defence forces).

Table 1: Number of individual applicants who applied for amnesty, with political affiliation (from Ernest, 2003)

Political Affiliation of Applicant	CSVR Figures Number percentage		TRC Figures Number percentage
ANC and affiliates	753	53%	998	61%
State Security Forces (incl SAP)	334	23%	293	18%
Other Liberation Movements	130	9%	138 PAC; 1 AZAPO = 139	8%
IFP and affiliates (incl. KZP)	89	6%	109	7%
Community Organisations	38	3%
Right Wing	53	4%	107	6%
Other	15	1%
Unknown	15	1%
Total	1427	100%	1646	100%

Research conducted at the Centre for the Study of Violence and Reconciliation (CSVR) suggests that approximately 250 of those 300 security force members who applied for amnesty (for approximately 827 incidents of human rights violations³¹) were former members of the SAP (Ernest, 2003).

The vast majority of the respondents interviewed for this report did not apply for amnesty and did not believe that they had any reason to do so. They suggested that their former colleagues who did apply for amnesty did so for a variety of reasons:

• 'they realized that they had committed criminal acts [in the past]'
• 'they couldn't live with their actions'
• 'they were concerned that they would lose their jobs or go to jail'
• 'they knew they were involved in atrocities in their work; or they were afraid of being implicated'
• 'they knew they would be exposed [by other testimony to the TRC]'
• 'Security Branch Head Office persuaded some people to apply for amnesty, to secure your future in the SAPS. Others had conscience issues and felt it was the right thing to do.'
• 'Those that applied were told to. Some were ordered to.'[emphasis added]

While the successful prosecution of De Kock, and a related fear of prosecutions among more former Security Branch officers may have been one important factor generating amnesty applications, this has not actually been researched. The participants in this project list a range of other possible incentives in addition to fear of prosecution. This suggests the need for further research into the factors and conditions that motivated decisions either to apply or not apply for amnesty. These might include 'negative motivators', for example, fear of prosecution, fear of job loss and being ordered to testify, and 'positive factors', such as conscience and 'doing the right thing'.

During primary research for this report, only one former police officer who had applied for amnesty was interviewed. His decision to apply was motivated by his knowledge that former colleagues had given information to the Attorney-General's investigation unit and therefore he faced possible prosecution. Despite the negative incentive, he found the process cathartic:

Initially I was very negative of the process but I was treated with amazing understanding and compassion. It was a good process, better than a Nuremberg trial kind of thing. (interview, former SAP Security Branch officer, November 2003)

Although his experience of the TRC was ultimately positive and healing, this officer also reflects that there were negative costs of his truth-telling:

But we were regarded [by our colleagues] as sellouts for applying for amnesty. (interview, former SAP Security Branch officer, November 2003)

From the side of the TRC, fairly limited resources were put into the process of investigation and prosecution, and this meant that the direct threat of prosecution against former police officials who had abused their powers was actually very limited (cf. Bell and Ntsebeza, 2001; Pigou, 2001; Pigou 2002). During and after the TRC, the government mandated the National Prosecutions Authority to consider prosecuting perpetrators who had been denied amnesty or who had not approached the TRC. Recent trials initiated by the Prosecutions Authority include those of SADF chemical warfare programme head Wouter Basson, Ferdi Barnard of the SADF's Civil Cooperation Bureau, and individuals named in the 1992 Bisho massacre. The failure of earlier prosecutions (notably against former Minister of Defence Magnus Malan³²) as well as the failure of the NPA's case against Basson could strengthen the view (among security force officers) that prosecution was not a real threat, and therefore that confessing through the amnesty process was unnecessary. A fatal flaw in the prosecutions' strategy is the fact that the Unit does not have any full-time investigators, nor can it afford to mount any trials (Wannenberg, forthcoming, p. 16). This reality was noted by the Chairperson of the TRC, at the time of its Final Report:

There are very many, I agree, who should have applied for amnesty and who didn't … [but, if they were charged], the burden on our [justice] system would be quite intolerable and the cost astronomical.. (Tutu, 2003)

In 1991 it was estimated that there were approximately 4000 members of the former SAP Security Branch (Cawthra, 1993, pp. 55, 58). Even if all of the estimated 250 former SAP members who applied to the TRC for amnesty (see above) were once members of the Security Branch (which is not necessarily the case, as the TRC was also interested in other aspects of policing, such as 'Riot Policing'), this means that fewer than 6 percent of the Security Police went to the TRC to apply for amnesty for acts committed in the course of their duties during the apartheid years.³³ This small number may appear to indicate the failure of the new government's efforts to get the police to participate meaningfully in the TRC process, until it is contrasted with the number of amnesty applicants from the former military, which are estimated at approximately 50 soldiers in total. The new government, and most notably first Minister of Defence Joe Modise, made little effort to get the military to participate in the truth commission process. The perceptions about the former South African Defence Force (SADF) among the former SAP officials interviewed for this report were interesting:

• 'The military considered it a war and so they decided they were not going to apply for amnesty.'
• 'The SADF officers were loyal to their troops, the SAP officers were not loyal' [i.e. they let their foot-soldiers take the rap].
• 'They closed ranks'.
• 'They were soldiers, it was war, and that was that'.
• 'They trusted that their operations would not be exposed'.

In its final report, the TRC remarked on how obstructive the South African National Defence Force (SANDF) was to the process and concluded that its 'research and investigation into the former [South African Defence Force] SADF were insufficient and constrained by a lack of access' (TRC Final Report, 1998, Vol. 5:8, p. 330). However, it went on to conclude that the former SADF had been 'significantly' involved in the perpetration of gross human rights violations.

Of those approximately 250 police officials who applied for amnesty, it is reasonable to surmise that most are no longer in the SAPS, although the reasons for their departure may have been unrelated to the TRC; as a result of either the offer of generous retrenchment packages (for example the payments made to De Kock and his colleagues on their departure in 1993) or the scrutiny associated with the senior appointments process in the new SAPS in 1995 and thereafter (which saw the exit of former Commissioner Johan Van Der Merwe and dozens of his contemporaries).

The TRC's Amnesty Committee completed its mandate at the end of May 2001 and published its final report early in 2003, as part of the TRC's Final Report. A breakdown of its work shows the following:

Table 2: Total numbers of Amnesty Applications and numbers processed by the Committee (taken from Coetzee, 2003, p. 193):³⁴

Amnesty Applications	Number	% of total number of amnesty applications	Applications that did meet the criteria	Proper applications that went through the full amnesty process
Total number of amnesty applications	7115	100%
Amnesty refused administratively because application did not meet the criteria defined in the Act	5143	72%
Number of applications that did meet the criteria	1972	18%	100%
Application withdrawn	258	4%	100%
Duplicate applications	40	1%	13%
Remaining applications that did meet criteria	1674	24%	85%	100%
Amnesty granted	1167	16%	60%	70%
Amnesty refused after a public hearing	362	5%	18%	21%
Amnesty granted for some incidents applied for, but not all	145	2%	7%	9%

Overall, this data shows that where the amnesty application was sound (a total of 1674 cases), at least 70 percent of such applications succeeded (although this number also equals 16 percent of the overall total of amnesty applications made to the TRC, because so many (72 percent) of the total number of applications made (7115) were refused on administrative grounds). In only about one-fifth of the 'proper' applications, amnesty was refused after the full process (including a public hearing) had been followed. It is not known how many of the 362 refused ap