Newham, G. (1995). Keeping the Wolves at Bay: Issues and concerns in establishing a witness protection programme in South Africa. Research report written for the Centre for the Study of Violence and Reconciliation, October.


by Gareth Newham
(edited by Graeme Simpson & Paul van Zyl)

Paper commissioned by Justice in Transition, October 1995.

Gareth Newham is a former Project Manager in the Criminal Justice Programme at the Centre for the Study of Violence and Reconciliation.


The issue of witness protection will undoubtedly be an important arrow in the quiver of the investigations of the South African Truth and Reconciliation Commission (TRC). Even more significantly, considering the historical absence of an effective witness protection system in South Africa, the development of such a mechanism within the TRC could have vital implications for the criminal justice system a whole.

Although South Africa has statutory provisions for the protection of those who may face danger as a result of coming forward to testify, these provisions fall far short of what is required if witness protection is to function as an effective tool in this country's criminal justice system. The current statutory provisions are rooted in legislation enacted to give the Goldstone Commission the legal capacity to protect witnesses testifying before it. The centrality of witness protection to the enterprise of the Goldstone Commission has not gone unnoticed by those drafting the Promotion of National Unity and Reconciliation Act, with the result that provision has been made for the TRC to establish a 'limited' witness protection programme of its own.

Recognising that there are substantial improvements to be made in witness protection in South Africa, the Act merely stipulates that:

The Minister shall, in consultation with the Commission, promote the establishment of a witness protection programme in order to provide for the protection and safety of witnesses in any manner when necessary1 (author's emphasis).

However, there has been very little research done on the issue of witness protection in South Africa and it is therefore the object of this paper to briefly examine some of the important issues surrounding the establishment and functioning of such a programme.

Despite the expense and complexity of adequate witness protection, at a recent one day conference attended by a wide range of delegates from both the criminal justice system and civil society, it was concluded that a well structured national witness protection programme is indispensable.2 Were such a programme already in existence, it would negate the need for a specific programme under the TRC. However, the amount of time that will be needed to plan and implement a national witness protection programme, makes it improbable that this will occur before the establishment of the TRC. This implies that the TRC witness protection programme will act as a precursor to a future national system. Indeed, it is consistent with a view of the TRC as being 'forward looking' in its objectives (rather than merely retrospective), to argue that the witness protection programme to be established under the Commission should act as a pilot project for a nationally administrated initiative. Although a limited witness protection programme run by the TRC will differ significantly in scope when compared to a more general national programme, the TRC's concern to secure the protection of witnesses will nonetheless engage with many of the same issues. For this reason, this paper will also concern itself with the similarities and differences between such a limited enterprise, when compared to an enduring national programme.

The first part of this report will pose and answer the question as to why a witness protection programme is necessary. This will be followed by a brief evaluation of witness protection in South Africa up to and including the Goldstone Commission's witness protection programme. Thereafter, the paper will turn to an examination of some of the broader issues that need to be taken into account with the conceptualisation and establishment of such an initiative. The aim of this latter section is to bring to the fore the kinds of problems and experiences that a witness protection programme will encounter, based on the experiences of previous initiatives. The ultimate object of this paper is to provide concrete recommendations, suggestions and potential solutions to the various problems raised.

Why a Witness Protection Programme

South Africa's criminal justice system demands that a court of law establish beyond a reasonable doubt that an accused person is guilty of the offence with which s/he has been charged, before such a person can be convicted. One of the most powerful vehicles to link a perpetrator to a crime is the testimony of a reliable witness, especially if other direct evidence is absent or inadequate. Most witnesses are called upon to testify in respect of crimes committed by other individuals or a small groups of individuals. In other cases however, witnesses may have to testify against a large group of people such as a gang or an organisation. If the witness is a past insider of such an organisation, it is likely that such a witness will be regarded as a traitor by his or her former colleagues and may be at risk of retributive actions from these former associates. It is under such circumstances that protection for the "whistle blower" is crucial. Having said this, it is important to recognise that the dangers posed by such highly organised criminal elements (whether political in nature or not) are not exclusively to former insiders, but may equally threaten former victims who come forward to testify, or mere third party witnesses to criminal acts.

Fear of reprisal is thus a powerful tool in preventing people from coming forward to testify. If a person is not offered any protection or support, it is unlikely that they will put their lives or well-being at risk by incriminating a criminal suspect. A relevant indicator of the magnitude of this problem is to be found in the attempts being made to combat criminal gang activity in parts of the Western Cape, where the National Institute for Crime and the Rehabilitation of Offenders (NICRO) cites intimidation of witnesses by gang members as the primary problem in the quest to curb these criminal activities.3

Apart from criminal youth gang activity, the problems of organised or syndicated crime appear to have increased significantly since the erosion of South Africa's international isolation. The unshackling of legitimate political activity which went with the negotiated transition of the early 1990s, was ironically accompanied by the broader deregulation of social control and the consequent creation of greater space for the operation of criminal interest groups.4 There is evidence to suggest that international crime syndicates, have consequently targeted South Africa as a lucrative market and transit point to other parts of the world, particularly in the illegal drugs trade.5 Sometimes the only way to successfully combat this problem is through police infiltration and the use of insiders who are willing to testify - often in exchange for their own indemnity from prosecution.

A witness protection programme should therefore not merely be seen as a mechanism for protecting people who are already in danger - it should be seen as an integral part of the criminal justice system and of crime prevention more generally. The criminal justice system needs to send out a message to the general public that they will receive protection and support if they are willing to come forward and help combat crime through giving evidence in court. At the present time this message is not being effectively conveyed and there can be little doubt that potential witnesses are not coming forward for fear of reprisals.

By contrast, in the United Kingdom excellent results have been obtained through an effective witness protection programme. In addition to the success of the programme in never having lost a witness, the calibre of witnesses attracted to the programme has resulted in a conviction rate of eighty eight percent where protected witnesses have given evidence in court.6 In the United States of America the results have been equally impressive, with an overall conviction rate of eighty nine per cent obtained as a direct result of testimony provided by protected witnesses.7 The witness protection programme run by the United States Marshall Service is "widely recognised as providing a unique and valuable tool in the government's war against major criminal conspiracies and organised crime".8 Through its potential to provide valuable witnesses with the confidence to come forward and testify, the establishment of a well structured, nationally run witness protection programme in South Africa is, therefore, essential as part of a comprehensive anti-crime strategy.

The general need for witness protection (as well as the need for sophistication thereof) is most apparent when examining the various creative instances where attempts have been made to use it. These have usually involved "politically-related" instances of criminal activity. Most noteworthy in this regard was the innovation of The Goldstone Commission of Enquiry, which was eventually involved in protecting thirty seven witnesses and forty of these witnesses' relatives.9 In another instance, the witness protection programme set up by Lawyers for Human Rights (LHR) was involved in the protection of thirty witnesses in the first twelve months of this programme's existence.10 At the present time there are other criminal investigations and prosecutions being conducted that will have a direct bearing on the work of the TRC and which also highlight the efficacy of effective witness protection. One of the most notable is the Internal Task Unit (ITU) investigations into hit squad activities in KwaZulu Natal. Currently the ITU has a witness protection scheme of its own operating under the present statutory provisions.11 Furthermore, in the high profile case of Colonel Eugene de Kock, the former C10 Vlakplaas commander, five witnesses who were to testify against the accused were flown to Denmark "because of fears by the men that they could be intimidated by serving police officers".12 Perhaps equally instructive are the contra-indications implicit in recent allegations of the professional murder of a Military Intelligence officer whom police believe was killed because of his access to sensitive information.13

It is clear that witness protection is a particularly important tool when dealing with highly organised covert operations. Whilst much of the illustrative focus thus-far has been on criminal gangs or syndicates (due in part to the "forward-looking" imperatives intrinsic to the enterprise of this report), it is perhaps most significant to highlight the relevance of the above to the state-sanctioned operations which systematised criminal activity under apartheid and which relied on covert activities by members of the security forces, bonded by their silence and by the belief that what they were doing was in some way authorised. Witness protection, combined with the prospect of amnesty in exchange for full disclosure, may be the only effective mechanism for breaking these bonds of silence in the truth recovery process. Consequently, there is every indication that the TRC will have to make use of witness protection to garner crucial evidence.

The Promotion of National Unity and Reconciliation Act has provided the TRC with the necessary statutory power to establish a limited witness protection programme. In itself, this reflects the general recognition that such a programme is necessary to elicit existing - yet hitherto uncovered - sensitive information about past human rights abuses during the apartheid era. It remains highly likely that many perpetrators of these past crimes will resist appearing voluntarily before the Commission to apply for amnesty. In this respect, their position remains secure as long as they are sure that information relating to their past crimes does not emerge from any other source. The intimidation of witnesses to prevent them from appearing before the TRC is consequently a very real possibility. Witness protection must therefore be seen as functional to the objectives of the TRC not merely in order to protect those who are willing to testify, but also to firmly encourage those who are resistant to the TRC enterprise to do likewise.

Witness Protection of the Past

Until 1992, statutory provision was made for the protection of witnesses under section 185 of the Criminal Procedure Act. In accordance with the Act, witnesses who were deemed by the Attorney General to be in need of protective custody, were kept in the same conditions as prisoners awaiting trial. The protection was not voluntary and the witness could be held incommunicado. This meant that the witness could be kept in a police or prison cell to which only state officials had access. The witness could only see a legal representative with the permission of the Attorney General and could be kept in detention until the relevant trial ended. As such "protective custody" was usually involuntary and was used mainly for state witnesses in political trials, the conditions proved to be restrictive and repressive in nature and unconducive to the voluntary cooperation of potential witnesses.14 These shortcomings of the old statutory witness provisions are captured by Judge Didcott in a judgement where he states:

When one looks at … section 185 of the Criminal Procedure Act, I am disturbed by what I see. I fear that the legislature may not have fully grasped the implications of some of the provisions of the section … and it seems to me that legislative reconsideration of some parts of the section may well be necessary.15

However, it took considerable lobbying and representation before new legislation was finally passed by the Minister of Justice in terms of the Prevention of Public Violence and Intimidation Act, under which the Goldstone Commission had been established. The new legislation made some important changes allowing for witnesses to voluntarily withdraw from the programme. Moreover, witnesses were also to be visited daily in order to have their psychological and physical needs attended to. Nevertheless, despite the improved conditions under which witnesses could thus be protected, these provisions (which currently remain in place) still have serious shortcomings. Some of these shortcomings are dealt with briefly below.

It is problematic that witnesses entering into the state programme are still compelled to deal directly with the police. They have to present themselves to a police station,16 are compelled to be kept in police or prison cells,17 and are looked after by security officers.18 These regulations may act as a disincentive for witnesses who do not trust the police or who are involved in matters which may require them to give evidence against the security forces. An even more problematic shortcoming of the existing system resides in the manner in which the programme has thus-far been operated by the Department of Justice. There has been no specialised training given to those administering the present programme and it is consequently still inadequate in providing comprehensively for those in need of protection. Given the small number of witnesses presently receiving protection, it is apparent that the current programme lacks credibility and therefore falls well short of being an effective tool in the South African criminal justice system.19

Even before the Goldstone Commission established its witness protection programme, another independent programme had been created by Lawyers for Human Rights (LHR). Aware that many witnesses were reluctant to be in protective police custody, LHR began to offer witnesses access to places of safety. It was predominantly due to the efforts of this organisation that recommendations were eventually forwarded to the Commission, leading to the formation of the Goldstone witness protection programme.

The Goldstone Commission Witness Protection Programme

A brief examination of the Goldstone witness protection programme provides useful insight into the immediate problems that such an initiative confronts under the auspices of a commission of inquiry. The Goldstone Commission did not initially make any plans for the establishment of its own protection programme for witnesses. The Goldstone Witness Protection Programme only came into operation almost a year after the establishment of the Commission. A press release announced the establishment of such a programme on 26 August 1992, only once it became absolutely obvious that there was an urgent need to protect witnesses who were willing to testify. A Security Director was appointed whose responsibility was to ensure the safety of witnesses. Until that time, the protection of witnesses who were reluctant to go to the police, had been reliant on non-governmental organisations, lawyers and journalists.

The functioning of the Goldstone programme, was framed on the assumption that witnesses needed to be encouraged to come forward, and that their security was paramount. The procedure for those seeking protection within the Goldstone witness protection programme was set out in a press release.20 Applications were to be made in writing to the Security Director who, in consultation with the Chairperson of the Commission, would then decide if the applicant was to be accepted onto the programme or not. If the witness was accepted onto the programme, the Chairperson, the Security Director, and the witness would agree on a place of safety. A Security Officer would then be appointed to look after the protected person. Any suitable person could be appointed as a Security Officer, whether a public official or a private individual. This person would be chosen by the Security Director in consultation with the Chairperson and the Commissioner of Police.

Initially there was a considerable suspicion of the Goldstone Witness Protection Programme. The main reason for this was that in May 1992, the administrative structure of the Goldstone Commission was transferred from the National Peace Accord to the Department of Justice. This move undermined some of the credibility that the Commission had sought to establish through having been part of the National Peace Accord rather than a government department. Under the Department of Justice, the administrative infrastructure of the Goldstone Commission was linked to a discredited state bureaucracy. The result was that the Goldstone witness protection programme became highly dependent on the SAP and the Department of Correctional Services. This suspicion and lack of credibility of the programme was silhouetted by former security policemen who were willing to testify before the Commission, but who refused to be protected inside of the country.21 Ironically, it is unlikely that the Goldstone Commission would have been able to sustain a witness protection programme had it not been for the resources of the Department of Justice. Not only did the Justice Department provide the Security Director for the programme, but it also contributed significantly to the financing of this extremely expensive witness protection enterprise. Nonetheless, problems relating to the location of the programme within the Department and doubts as to its independence continued to plague its effective operation.

The fact that none of the witnesses on the Goldstone Witness Protection Programme was killed, has been cited by some as a testimony to its success. However, a brief examination of some of the incidents that occurred, suggests that this was more by luck than by design. Paul Erasmus, a former security policeman and a key witness on "third force" activities, was attacked in the "safe" house where he and his family was staying while on the programme.22 This is but one of several examples where the security of protected witnesses were severely compromised. In some cases, this was a result of the ill-disciplined actions of witnesses themselves, but in others it was clearly due to the inexperienced operations of those engaged in running the programme.

Ultimately, it should be acknowledged that, formative as it may have been, the Goldstone Witness Protection Programme suffered from a lack of adequate planning and experience. The programme was managed and run by people who had had little or no previous experience in directing such an endeavour. Much of the time, those saddled with the responsibility of ensuring the witnesses' safety and well being had to rely on common sense and quick thinking. It very quickly became clear that the staff of the Goldstone Witness Protection Programme were inadequately prepared for the taxing and often complex demands that such an initiative would place on them. The result was a programme that operated in a predominantly ad hoc fashion and which was primarily concerned with the narrow enterprise of short term physical safety for those being protected. In addition, when the Goldstone Commission itself concluded its operations, there was little forethought given to the future of some of the key witnesses who would potentially be involved in the subsequent prosecutions emerging from the Commission. It is therefore true to conclude that despite its limited innovations, the Goldstone Witness Protection Programme did not result in any substantial improvement in national witness protection.

Contextualising Witness Protection

The aim of this section is simply to highlight some of the important broader dilemmas that must be confronted in fashioning any effective future witness protection programme. The first basic issue that has to be dealt with is the ultimate objective of the witness protection programme. Ostensibly it is to protect witnesses from any danger or intimidation arising from the decision to come forward to testify. Indeed, the Goldstone programme took this as its point of departure, with the result that the primary focus was simply to remove the witness to a place of safety. It quickly became apparent, however, that far more than this was needed to provide for the well being of witnesses. Whereas the Goldstone programme was established as a reaction to witnesses' needs for protection, it is imperative that any future programmes adopt a different approach based on a wider vision. A witness protection programme needs to be conceptualised as a pro-active tool in fighting crime or, in regard to the TRC, to help achieve the broader aims of the Commission.

The TRC is concerned with "truth recovery" and with securing a process of national reconciliation, rather than merely with prosecutions, revenge, victimisation or legal retribution for past politically-motivated violence acts. This agenda imposes specific needs on the process of witness protection which may distinguish it from the broader system primarily intended to service effective prosecutions and criminal justice. There is a need to thoroughly debate the implications of this in forging the "limited" TRC programme. On one hand, within a prosecutorial process, the need for ongoing protection of witnesses may or may not cease with the successful conviction and incarceration of the perpetrator/s. On the other hand, in the TRC, where no such prosecutions will result, the suggestion is that, depending on the success of the reconciliatory enterprise (which is extremely difficult to anticipate in the context of individual experience), witness protection may consistently demand mechanisms which are more enduring. Furthermore, rather than merely demanding protection from fellow perpetrators or co-conspirators, the TRC witness protection programme may have to seriously consider requests to protect perpetrators who testify in order to receive amnesty, from the victims of their past acts, as well as protecting survivors who testify from possible recriminations of those they may incriminate.

Although none of these specific demands are entirely strange to witness protection more generally, they nonetheless demand a flexibility (in policy and practice) which can grapple with both the practical and moral dilemmas which result. Despite its "limited" span, therefore, the TRC witness protection programme may have to do more - rather than less - than an ordinary programme servicing criminal justice. Perhaps one of the greatest dilemmas which will be confronted by the TRC programme resides in the fact that it will be operating in the context of a simultaneous attempt to rehabilitate and possibly compensate victims. One grave moral danger which must be avoided, is the prospect that the benefits received by perpetrators who are placed on the witness protection programme outweigh those available to their victims through the reparations process. This would not only serve to discredit the embryonic witness protection programme, but would have severe implications for the credibility of the TRC itself.

Nonetheless, it is indispensable that any witness protection programme must win credibility in the eyes of the wider society. The message must be conveyed that there is support and protection for those who are willing to testify regarding crimes which they have witnessed or to which they have been party. If the object of such a programme is to encourage those with valuable information to come forward, then it is imperative that it cannot result in witnesses ending up in a worse off position than they were before they decided to testify. However, it is equally important that the programme is not seen as providing rewards for those testifying, or as an easy mechanism for simply avoiding ordinary criminal prosecution. If this were the case, it would both substantially undermine the credibility of the witnesses' testimonies, as well as serving to attract opportunists who merely hope to improve their personal circumstances.

However, in its conceptualisation, the programme must primarily focus on the needs of the protected witness. In other words, a witness protection programme has to be "people orientated" and not "trial orientated" (particularly in respect of a non-prosecutorial process such as the TRC), as seems to be the case at the present.23 To date, the state witness protection programme has been administered by the Department of Justice, the primarily focus of which remains the pursuit of criminal convictions. As a result, the Department has tended to conceive of the limits on its responsibility to protected witnesses as being defined only by the duties called for under the Criminal Procedure Act.24 It will be demonstrated below that this is clearly not adequate.

The South African Social Context

A difficulty that South Africa faces in establishing a witness programme has to do with the historically generated inequities that exist within the country's population. The consequent racial, economic and cultural disparities between South Africans has plagued previous witness protection programmes in South Africa and limits the applicability of the various comparative international experiences that may be used as examples for a future South African witness protection programme.

By way of example, the National Coordinator for the Network of Independent Monitors (NIM), highlighted the severe problems confronted in removing witnesses from one community and relocating them in another.25 In many cases this could be achieved without generating suspicion, but in other cases it was argued that the cohesive nature of local communities, coupled with the existence of broad national networks, sometimes resulted in it taking only a couple of weeks before the whereabouts of witnesses would be known to their original communities. In terms of the size and mobility of the population, there is much that mitigates against the success of witness relocation. However, this dimension should not be excluded as a possibility, as the Goldstone witness protection programme managed to successfully relocate a number of witnesses.

More obviously, the difference in class and culture also presents obstacles to using a single system of witness protection for all witnesses. As has been asserted, the lifestyle of the witness must remain similar to that which to which s/he has previously been accustomed. Cases did arise on both the Goldstone and LHR programmes where witnesses were moved out of the social context to which they were accustomed, with the result that the stress endured by them increased substantially. This inevitably compounds the problems which already highly stressed witnesses are experiencing.

The lessons from witness protection programmes in other countries do not readily provide mechanisms to overcome the problems presented by the South African situation. Few other developing countries boast effective witness protection programmes and the successes of programmes in countries such as the US, have generally been achieved through trial and error. Given South Africa's unique situation and historically shaped racial and class-based social divisions, it is likely that the development of an effective witness protection programme will only be achieved over a substantial period of time, based on constant re-evaluation and regular improvements.

This incremental view does pose problems for a "limited" witness protection programme under the TRC, which would have to be established rather hastily. However, it also implies that witness protection strategies under the TRC should not be seen in isolation, but as a vehicle for testing and developing a more enduring and nationally coordinated programme.

Reluctance of Witnesses to Testify

Apart from the historically-rooted climate of fear and mistrust which continues to pollute community-police relations in South Africa, the criminal justice system also lacks popular credibility. Neither of these arms of the criminal justice system has effectively checked the high level of intimidation which continues in South Africa and which is symbolised by the ongoing gang and taxi wars. Even the courtroom has been a forum for such intimidation (as in the trial of the killers of US exchange student, Amy Biehl), and ill-applied bail regulations have, in the eyes of many South African victims and survivors, rendered the court complicit in their ongoing intimidation.

More to the point, coming forward to testify as a witness in a trial also presents as a major inconvenience to people who have been called upon to do so. The absence of adequate facilities, repeated delays and postponements of trials, all result in frustration and inconvenience for witnesses. Given the importance of witnesses' testimony in the criminal justice system, it would appear that what is needed is a general paradigm shift in terms of the treatment of witnesses.

As noted above, a particular problem that the TRC may have to face in encouraging witnesses to come forward, has to do with the likelihood that under the Promotion of National Unity and Reconciliation Act, many perpetrators will be granted amnesty. This may impact on potential witnesses who may be reluctant to come forward if they feel that nothing will happen to those that they give evidence against. There is simply no guarantee that witnesses will be safe from retribution after testifying.

Critical Issues and Recommendations

Who Requires Protection

As noted above, the testimony of witnesses in courts of law and commissions of inquiry are of prime importance in reaching judgements and making findings. For a witness protection programme to effectively encourage people to come forward to testify, it is critical that cognisance is taken as to the kind of people that will need to be protected under such a programme.

Some of the witnesses will be innocent bystanders or direct victims of crimes, whilst others will be insiders to criminal activities who may have been significantly de-socialised due to the past experiences that have resulted in their behaviour. This means that a witness protection programme will have to be prepared to deal with all types of people - including some who may be criminals themselves. This has important implications at a number of different levels. Firstly, witnesses who are relocated may present a danger to the new communities into which they are settled. Experiences in other countries have demonstrated that a significant proportion of such witnesses commit further crimes while on the protection programme. Secondly, because of their past criminal activities, as well as the fear of further such criminal behaviour, few host communities are willing to offer protection or re-integration to those who have been implicated. In particular, the creation of job opportunities presents serious problems for former criminals - and even for former victims.

A resultant problem is that witnesses on protection programmes often experience severe stress. Most of them face uncertainty over what the future holds and loss of the normality of their everyday lives. Witnesses who are relocated our of their communities generally also experience a great amount of frustration and alienation. Where such witnesses have been victims of violent crimes, they often have to simultaneously deal with the consequences of violence-induced trauma symptoms - at precisely such a time as they are deprived of any community or family-based support systems. All these and other stressors can often manifest themselves in anti-social behaviour. It is not uncommon for such witnesses to become prone to alcohol or drug dependence, or to act violently. Witnesses frequently experience marital problems while on the programme, due to the disruption in their lives. Others may become very depressed or uncooperative.

From all this, it is thus apparent that a witness protection programme needs to provide far more than mere physical protection, but must also cater for the social assistance that witnesses may require as a direct consequence thereof. However, most of those interviewed who have had to deal directly with witnesses in either the LHR or Goldstone witness protection programmes, felt that they were unable to provide the witnesses with the kind of counselling and emotional support required. In the LHR programme it was usually the lawyers who were representing them before the Goldstone Commission that had to take care of the witnesses. These lawyers related that witnesses generally became highly dependant on them - to the extent that looking after a witness became a full time job. On some occasions witnesses threatened or tried to blackmail these lawyers by refusing to testify unless they had their demands attended to.

Some opportunists may also lie or exaggerate their claims in order to get onto the witness protection programme in order to improve their lifestyle or evade their obligations. Experiences in other countries show that some people attempt to get onto witness protection programmes so that they can escape responsibilities such as debts or maintenance payments. Even witnesses who legitimately need protection may try and exploit the system.

A significant aspect of witness protection is that in many cases, protection will have to be offered to the whole family of the witness - therefore including a substantial number of people who may not themselves be direct witnesses. This dramatically impacts on the costs of the entire enterprise. In the US, the witness protection programme has to protect on average 2.5 family members for every witness protected.

Regarding the protection programme that the TRC seeks to establish, the nature of the witnesses seeking protection will most probably be highly specific in that they will usually be testifying against people in the security forces and possibly against powerful people in political parties and/or government. Nonetheless, the problems of stress and anti-social behaviour in these circumstances know no boundaries based on such distinctions, and it is therefore important that even a limited protection programme be prepared for all eventualities related to the trends outlined above. An understanding of these trends is an essential first step in recognising the types of services that such a programme will have to offer to those who are being protected. It is self-evident that there will be a need for a large contingent of trained experts to deal with the various problems that will ensue.


Witness protection must be open to everyone, provided that they can provide valuable evidence as witnesses and that they and/or their families are placed in danger as a consequence of so-doing. Witness protection programmes must offer more than mere physical protection and must cater for the wide range of psychological, emotional and social problems which will inevitably result within diverse groups of potential protected individuals and their families. This imposes substantial costs and difficult management issues on any such witness protection programme.

In some other countries, such as Australia and the US, background screening is done on witnesses before they are allowed full access to the programme. Factors such as personality, debts, addictions, family, medical and other problems, are assessed so that the witnesses liabilities and needs can be accurately determined and predicted in advance.

It is proposed that in establishing a witness protection programme in South Africa, some early evaluation systems should be integrated to facilitate planning which effectively considers the following factors:

  • An assessment of the fears and problems of witnesses, so that the necessary support and help can be offered.

  • Who would be best placed to work with the witness, in terms of language, culture, experience, trust etc.

  • Witnesses should be protected as far as is possible under the conditions to which they are accustomed.

  • A contract or "Document of Understanding" should be drafted between the witness and the Directorate which spells out what is expected from the witness, as well as outlining penalties for non-compliance under the agreement.

  • It is also vitally important that such a contract gives witnesses an accurate and thorough assessment of what they can expect from the programme before they agree to participate. No undertakings should be made that cannot be kept.

  • The onus should expressly remain on the witness in respect of the resolution of any outstanding debts or obligations accruing to such a witness at the time of entering the programme. The specific limitations on the programme's liability for debts and obligations incurred by a witness after his/her incorporation under the programme should also be spelled out in no uncertain terms.

Structure of the Witness Protection Programme

The structure of a witness protection programme has to take into account every aspect of dealing with witnesses who may be under threat, and must simultaneously delineate and integrate its various functions. The overall effectiveness of the structure of the witness protection programme will ultimately determine the quality of the witnesses that will come forward.

It is argued that one of the biggest problems of the Goldstone witness protection programme was the fact that it was not adequately professionally structured. Throughout the life of the programme this lead to problems which could have been avoided with proper planning and foresight. This was ultimately due to the fact that the Goldstone witness programme came about as a creative response to the existing need for witness protection, rather than as a preemptive policy aimed at encouraging witnesses to come forward.

It must be acknowledged that some of the structural problems confronted by the Goldstone programme could not have been avoided due to the circumstances at the time. The programme was administratively bound to the Department of Justice. However, the programme's staff - including the Security Director - were, as a matter of convenience, chosen from those already attached to the administrative section of the Commission, rather than for their skills in the area of handling witnesses. Many of the problems experienced were relatively minor, but disruptive and efficiency could have been improved significantly had a proper structure been in place.

Initially, the Lawyers for Human Rights (LHR), witness programme appeared to be more structurally efficient than the Goldstone Commission programme, accommodating as many as thirty witnesses in its first 12 months, compared to only six under Goldstone. However, rather than being an indicator of structural efficiency, this appears to have been related to an initial issue relative credibility. LHR in fact states that it was intended to terminate its programme with the inception of the Goldstone programme, but that the suspicion of the Goldstone programme, combined with the fact that it only protected witnesses who were to testify before the Commission, imposed on LHR the need to sustain its own witness protection mechanism.26


Despite any of the criticisms above, there appears to be a strong consensus amongst stake-holders that a South African Witness Protection Programme should be located within the Justice Department.27 However, it is suggested here that the problems associated with the bureaucratic mind-set of the Department of Justice which have been apparent thus far, can only be overcome if the witness protection programme has its own dedicated staff, administrative support and funding. A witness protection programme should be established by a Directorate consisting of individuals who would have to be as independent and politically neutral as possible. If possible it should be directly accountable to parliament.

The Directorate of a national witness protection programme should:

  • Establish the aims and functions of the Witness Protection Programme.

  • Make decisions regarding the which applicants are taken onto the programme and under what terms.

  • Establish the necessary infrastructure at a national level, and possibly regional level, including the employment of the necessary administrative and support staff.

  • Set up sub-committees for the different localities who will be responsible for the witnesses in their areas.

  • Determine the links and relationships between the programme and other support structures both within and outside of the state.

  • Be responsible for the budget of the programme.

  • Establish a system of checks and balances within the programme to enhance accountability.

An effective witness programme would, of necessity, have be coordinated at a national level. However, a relatively decentralised structure would be equally important. It would therefore be important to have regional sub-directorates who would coordinate the protection of witnesses in their areas, but which could, through national coordination, move people in and out of a particular region without any problem. It is difficult at this stage to say wether it would be viable to have permanent structures in every region, or to have mobile or collapsible structures that can move about to each locality within a particular region.

One of the clear advantages of establishing an embryonic witness protection programme under the TRC, resides in the structural concern with the independence of this initiative. Such a limited protection programme would most likely be coordinated and accountable to the TRC. This may preclude the issue of location of the programme within the Department of Justice. If it is acknowledged however, that a programme under the TRC could be used as a base for a national initiative, it should be designed and established in a manner conducive to effective integration into such a state structure at some future point.

Financial Control

One of the most important issues in the establishment of a witness protection programme, is that of financial control. Witness protection programmes have to deal with a wide range of different types of expenditure, and it is imperative that there is a system of strict control in place. The structure must be set up with accountability and efficiency in mind, but equally importantly, it must also be flexible enough to meet the unusual requirements of witness protection. Maintaining a witness protection programme can be a very expensive endeavour and it is for this reason that careful financial planning occurs and regulatory mechanisms be put in place. There are various ways to cut costs, but the innovative strategies and foresight needed to do so entails careful assessment of every aspect of a witness protection scheme. The difficulty in sustaining a balance between flexibility and efficiency on one hand, and adequate regulation and transparency on the other.

Because the Goldstone programme was financially administered by the Department of Justice, it was bound by the Department's financial procedures. The main drawback, was that these procedures proved to be too labourious and bureaucratic for the needs of the witness protection programme. If, for example, a witness had to be moved in the middle of the night to a different location such as a hotel, there was no mechanism whereby immediate funds were available to cover transport or accommodation costs. In many cases, the staff of the Goldstone programme had to pay for certain requirements themselves and later seek to reclaim the money from the Justice Department. As they were frequently unaware of the precise procedures governing this process, such claims often became complex and inconvenient, leading to time wastage and extreme frustration on the part of either the staff or the witnesses or both.28

This experience demonstrated that the issue of money was frequently central to generating severe tensions between the witnesses and their 'minders'. On occasion 'minders' even had their lives or physical safety threatened by irate witnesses.29 The Goldstone experience clearly indicates that such a witness protection programme requires a specialised financial structure capable of overcoming the bureaucratic constraints which typically characterise other governmental departments.

The present legislation is markedly inadequate in providing explicit directives regarding the financial implications of being a witness on a protection programme. Questions over the taxable status of monies paid out to witnesses, such as any financial remuneration provided to compensate witnesses for loss of income, the status of any debts owed by such witnesses, the ordinary remuneration of witnesses while on the programm, etc. must be clearly set out so as to avoid confusion and tension as may occur at a later stage. A balance has to be found which seeks to sustain witnesses at the standards to which they are accustomed, without over-compensating them in such a manner as might appear to be rewarding them for giving testimony. Not only could this discredit the witness' testimony, but it may also inadvertently encourage the witness to remain on the programme for excessive periods of time. Once again the moral dilemma of the comparative treatment of perpetrators who enter the programme, relative to the benefits available to their victims through compensation before the TRC, must weigh heavily in the policy decisions underpinning such financial arrangements. At the level of policy, questions as to whether all protected witnesses should receive equal benefits or treatment, must be weighed against the potential that if entry to the programme appears to prejudice the interests of particular people, there will be less incentive for such witnesses to come forward to testify.

Policy must also be clearly established and communicated in respect of witnesses who owe debts to creditors when they enter the programme. They may have to pay off bonds or loans, pension schemes, medical aid, or they may be paying maintenance. To ensure that such creditors are not prejudiced, this sphere will need considerable attention - it has proven to be problematic in other witness programmes such as in the US. It may be necessary to find a mechanism for payment of such debts through the programme, on condition that contractual obligations are established for repayment either during or after the period of witness protection.

It is argued by some that witness protection programmes, if effective, can ultimately pay for themselves through the capture of the illegitimate earnings of criminal cartels.30 Indeed, money may be put back into the witness programme through the confiscation of illegitimate capital and assets that are uncovered by investigators, who use witnesses on the programme to secure convictions. The US drug enforcement agency could provide a model for the channelling of such funds. Whilst idealistic, this is not always easily realisable. In any event, given that money for a national witness protection programme would probably have to be approved by parliament, a long term cost-benefit analysis would have to be undertaken to establish the utility of such an approach.

Such perspectives on self-financing witness protection may impact on a wider national witness protection programme, but will certainly not effectively service the limited witness protection programme of the TRC which is simply not concerned with this type of criminal activity. In terms of the TRC, the finances would simply have to be provided by the state through the Commission. Considering the expense of the Goldstone witness protection programme, this may prove to be a very significant problem for the TRC. It is thus of primarily importance that limitations are imposed on the TRC witness protection programme, both through restrictive entry criteria onto the programme, as well as through establishing cost effective expenditure policies which are set out clearly before protection is offered.


The priorities of cost minimisation will demand careful research into the most cost effective methods to be used and will necessitate mechanisms of oversight and scrutiny of expenditure in each case. Most of the recommendations forwarded in this report have attempted to take cognisance of the financial implications involved, but this requires specialised expertise and an independent focus. Whilst the general assumption behind the establishment of a witness protection programme may be that it saves money in the long term through curbing the costs of crime, this is not as easily arguable in the case of the TRC witness protection programme, which is at best concerned with broader objectives of reconciliation, rather than with criminal convictions.

Further recommendations include:

  • There should be a full-time financial controller at both national and regional level. The financial side of the programme should not be managed by the Security Director.

  • A strict financing structure should be established so that all monies spent can be accounted for.

  • An account with limited funds should be established to allow for immediate access, so that the people working with the witnesses can see to minor matters such as accommodation costs, food, transport, etc. without unnecessary delays.

  • Regulations or legislation will need to spell out policy considerations relating to all the financial implications of being a witness on the programme, including the financial obligations of potential witnesses, issues of tax, and civil claims against witnesses, etc., as well as the financial obligations of the programme in respect of such witnesses.

Staffing Considerations

As noted above, a significant problem which affected the Goldstone witness protection programme was the absence of any coherent staffing plan. The Commission simply had to appoint employees of the Commission on an ad hoc basis in order to fulfil certain functions in regard to the witness protection programme. This meant that there were no full time staff on the witness protection programme and those who were appointed generally had other pressing Commission responsibilities. The Chief Director of the Department of Justice was appointed as the Security Director of the Goldstone witness protection programme. However, in effect, much of the day to day administration and running of the programme, including such tasks as locating places of safety for the witnesses, transportation of witnesses and looking after the general well being of those on the programme, ended up on the desk and shoulders of the Goldstone Commission's secretary, Glen Cuthbertson.

Given the various social, emotional and other problems that experienced by protected witnesses, the responsible Commission staff were often forced to play all the roles from social worker to marriage counsellor or financial advisor. Although the initiatives taken by many of these staff under extremely difficult conditions can only be commended, the lack of experience and of effective structures meant that in some cases serious mistakes were made which could have had disastrous consequences.

The Goldstone Commission experience therefore brings to the fore some of the important priorities in terms of staff requirements and support structures needed for such a witness protection programme. The range services that have to be provided to witnesses effectively means that a self-contained protection programme would demand a substantial full time staff. Whereas this may be viable in the long term within a coordinated national witness protection programme, it is unlikely that this would be practically possible within the ambit of a limited programme under the TRC - which is unlikely to command the resources or staff required. This would mean that much of the service provision (in the form of psychological or vocational counselling, social work assistance etc.) may have to be provided from outside the TRC and its protection programme. This raises critical security questions and places a premium on the quality and experience of the limited number of staff who will nonetheless have to be employed full-time to run the TRC witness protection programme.

The Goldstone Commission experience demonstrated that the staff of a witness protection programme need to have some specialised training. Necessary qualities and skills include an ability to be highly empathetic and able to deal with the stress and concerns of those under protection. The Goldstone Commission seconded people from the Department of Justice because it proved to be most convenient. However, it became apparent that those appointed as Security Directors often received these briefs due to the authority which they were required to have in order to handle and administer the finances involved, but they lacked the necessary skills, insight, empathy and approach that such a position required. Even more critically, they tended to be inflexible in regard to the rigid financial procedures to which they were accustomed in other contexts and insensitive to the specific needs of witness protection in this regard.


The staffing of the witness protection programme is a crucial factor to be carefully considered. Any successes of the Goldstone witness programme can be attributed to the dedication and innovation of the particular people who were responsible for the day to day running of the programme, and who worked with the witnesses. However, any future programme will need to more systematically recruit well trained, committed staff at all levels. Specific training may also be necessary to equip these staff to anticipate the nature of the demands placed on them and the problems they will confront. These recruitment and training priorities provide an important motivation for a sustainable and enduring national witness protection programme. It will be a waste of human and other resources if such a programme does not emerge from and capitalise upon the development of the more limited TRC initiative. Ultimately, there is a need for a contingent of full time staff who will be able to build on the experience that they acquire through the process of developing a protection programme.

The following additional recommendations should be considered in the staffing sphere:

  • Securing the assistance of international experts who have had valuable experience in establishing or running witness protection programmes elsewhere in the world, to assist in setting the TRC programme.

  • The Directorate should consist of between three and five people with at least one person at the level of an Attorney General.

  • There should be a strict separation of functions and the duties and responsibilities of each position within the Directorate needs to be defined.

  • Any staff working on the programme will need to be screened and obtain a security clearance - and should preferably also undergo random follow-up security screening.

  • The safety of the staff administering the programme must also be assured.

  • The staff working with the witnesses need to be objective, have good "people skills" and should be specifically trained to deal with specific exigencies.

Support Structures

In terms of support structures, it appears that a limited programme will not have the capacity to provide witnesses with all the support they will need in terms of social and psychological services. During the life of the Goldstone Commission, Judge Goldstone realised that there was a need for witness support and rehabilitation. As a trustee of NICRO, an organisation which ran programmes to rehabilitate ex-convicts back into society, Judge Goldstone secured the assistance of the organisation in this regard, specifically to assist witnesses with reintegration into society once they had come off the programme. In some cases this was reportedly extremely successful. The link between the Goldstone witness protection programme and NICRO can provide valuable lessons for the TRC in regard to outside support structures servicing such a limited witness programme.

The issue of rehabilitation and termination of witness protection will be dealt with in greater detail in the next section. However, some of the structural issues regarding outside links to a witness protection programme will be now examined. One of the most important issues in this respect is that of security. This relates both to the security of the witness, as well as to the security of the social worker or service provider. In terms of the former, tensions may arise between the needs of the witness and his or her security. During the Goldstone Commission it did occur that a witness' access to a social worker was curtailed because of security considerations. To overcome these tensions, it was imperative to ensure effective coordination and consultation between the coordinator of the service providers and the person responsible for witness security. In the case of the Goldstone Commission, this demanded that the "rehabilitation services" of NICRO was highly centralised. The coordination of all the social workers was determined by a single coordinator from within NICRO.

This served a number of important functions. It meant that if there were problems in regards to the security of a witness and access to a social worker, each case could be worked out between the security coordinator and the social worker coordinator. This freed the relevant social worker or service provider from being caught up in the specific problem which was resolved at a centralised level. Despite some of the practical difficulties which this introduced, this system ultimately enhanced the security of the witness as particular social workers or service providers would only know of those witnesses with whom they were working and not of the others on the programme. Lastly, this centralised system helped ensure the security of the social worker/service provider as well, as only the coordinators of security and of NICRO knew which social workers were working with which witnesses.

It is likely that the witness protection programme under the TRC is going require similar links with various service provider organisations which can offer the skills and support necessary for witnesses within the programme. It is therefore crucial that these linkages are clearly delineated and carefully organised. It is critical however, that these service organisations are chosen carefully, especially as the commitment needed will be substantial. Organisations will have to be highly credible with a strong record of work in the relevant fields.

The comprehension of these intricacies within the Goldstone protection programme, reflects the commendable attributes of the individuals involved, but also reveals the need for a careful analysis of the dynamics of keeping witnesses on protection. These factors need to be translated into the various staffing functions and responsibilities within a coherent programme structure.


For the programme to deal effectively with the various needs and social services required by witnesses, it is going to need to secure the services of social workers, welfare organisations, medical practitioners, psychologists (both for children and adults), various counsellors, and ministers of religion, etc.

In terms of such support, the following should be considered:

  • The programme need not have to include all these services within its structure, but can set up consultancy schemes with organisations that offer these services.

  • All social service consultants will have to be screened and given security clearances.

  • The Directorate or sub-committee may have to assist in negotiating arrangements to organise employment for witnesses on the programme and once such witnesses come off the programme. This is a controversial issue which must be weighed against the policy considerations of how far the TRC can go in offering victims comparable assistance. It is least controversial if such negotiations merely revolve around the possibility of witnesses returning to the jobs they occupied prior to entering the protection programme.

  • "Rehabilitation" of witness in order to assist them to reintegrate into society, will need to be a sustained process throughout duration of the witness' participation on the programme.

  • Provision should be made for occupational therapy and witnesses should be encouraged to acquire skills while on the programme which will improve their employability and thus help with their reintegration back into society.


This is a crucial issue that is often one of the most difficult to deal with. Once a witness who is under serious threat enters the witness protection programme, they are effectively removed from normal interaction with society. They often cannot do the simplest things, such as going to the shops or paying bills, without a guard or someone from the programme present. This has commonly resulted in the witnesses becoming highly dependant on the programme and those involved in protecting them. This can occur to such an extent that witnesses start to feel incapable of doing anything for themselves. The longer a witness has been on the programme the more intricate this problem becomes. Once the witness has testified and is no longer under any threat, he/she may find it very difficult to adjust back to normal life. At the same time, the person may not be able to go back to the original community from whence they came or, as in the case of policemen, may often not be able to go back to their previous jobs. This may be the result of social ostracism that they face for testifying against colleagues or powerful people, or as a result of their own humiliation. This may compound feelings of insecurity about being accepted back into society.

The insecurity felt by these witnesses is usually apparent for the entire time that the witness is on the protection programme. It is thus important that the rehabilitation of the witness begins from the moment that they contract into the programme. The Goldstone Commission only implemented these rehabilitation measures two to three weeks before the witness was to leave the programme. This proved problematic for a number of reasons. In some cases witnesses had been on the programme for a substantial amount of time. This often demanded more sustained counselling for witnesses who had become highly disillusioned after spending most of their time extremely anxious about an uncertain future. In some cases it became obvious to the social workers that witnesses were in need of professional psychological help as a result of trauma. Coming in at such a late stage compounded the difficulty in dealing with this trauma.

The rehabilitation programme did however, manage to successfully relocate and integrate some witnesses back into society. A scheme was worked out with the witnesses as to where they felt safe to relocate, and attempts were made to accommodate their desires for the future. In some cases witnesses even moved back to their original communities. In other cases witnesses were encouraged to choose areas where they were not known, and this his usually entailed being relocated to totally different parts of the country. Witnesses were also encouraged to examine their options for future employment. Some witnesses could not go back to their previous jobs and opted to take training courses to increase their employment potential in other fields. A careful programme of re-integration was supposed to be worked out with each witness, which would set out the various steps that the witness would undergo to become self-sufficient in society. This programme was to be monitored for six months, during which time the social workers would sustain regular contact with the witnesses to ensure that the integration was going well.

The single most significant factor that contributed the Goldstone rehabilitation programme's relative success, was its access to financial resources.31 Once witnesses left the protection programme it was claimed that they were offered financial support until they could start working. Skills training was sponsored and, in some cases, a small capital outlay was provided to assist ex-witnesses to start their own businesses. By way of overall evaluation of the Goldstone programme, it must be acknowledged that none of the witnesses who have been relocated have come to any harm, and most have successfully managed to start new lives. As contact usually ceased after the six month period, it is difficult to assess the success of the programme beyond this. It must also be recognised that some witnesses also sought to take unfair advantage of the situation. In one instance a witness was claiming money for his family whom it was later discovered, were not benefitting from it. The need for a strictly accountable financing system must also regulate expenditure on rehabilitation measures.


A limited witness protection programme under the TRC would provide the perfect opportunity for creative attempts at rehabilitation of witnesses as part of the process of building reconciliation. Experience in the USA has revealed that witness protection programmes can play an important rehabilitative function. Of those witnesses on the US Marshals protection programme who had extensive criminal records, the recidivism rate (the percentage of those who return to crime), was half that of criminals who were released from state penitentiaries.32 This is however, contingent on the ability of the protection programme to offer witnesses a comprehensive rehabilitation programme.

It is crucial that any rehabilitation programme begins as soon as the witness enters the programme. Furthermore, witness evaluation done to establish the suitability of a witness to such a programme should include a psychological assessment and evaluation. This will not only help to determine the kinds of support necessary, but will also help to determine the nature of the rehabilitation programme needed.


Probably the most important aspect of any witness programme is its security arrangements. Security has to be seen in a holistic sense that encompasses every aspect of the witness protection programme. Although the primary focus of security surrounds the protection of the witness from harm, security arrangements also have to take into account the safety of the witness protection programme staff and, in some cases, the security of communities where witnesses may be relocated.

The Goldstone programme aptly highlights the kind of problems that an unplanned and limited protection programme can expect to encounter at the level of security concerns. The most immediate form of security that can be offered to witnesses is to relocate them to a safe environment. This has to be done efficiently and as quickly as possible. Once they have been moved, it is then vitally important that their place of safety remains a secret for as long as necessary. It was in this respect that, due to a lack of planning and experience, the security of witnesses on the Goldstone programme was occasionally compromised. A particular example concerns a witness who was given a safe house which was rented out in the name of the Goldstone Commission. Evidence suggests that it was popular knowledge within the surrounding community that the family living in the house was connected to the Goldstone Commission. There were also reports that police officers who were transporting witnesses informed unauthorised people of their activities and acted in such a manner as to both increase the stress of the witness and to compromise their security.

Witnesses may also compromise their own safety. Once again the Goldstone experience is revealing. In some cases witnesses contacted their family and friends and revealed where they were living. This is a difficult problem to overcome as witnesses often feel un-threatened by their friends and family. Any breach of security however, may ultimately result in the witness being endangered. It is therefore imperative that the witnesses are thoroughly educated and trained to ensure that they are aware of the guidelines that need to be followed to ensure their safety. This is also important because of the danger which the witness may present to the staff of the protection programme, or to others. This danger is increased when the witness is a perpetrator of past criminal behaviour. Indeed, some of the people who worked with witnesses on both the LHR and Goldstone protection programmes were threatened by the very witnesses they were ostensibly protecting.

A problem also exists regarding who is given responsibility for the safety of the witnesses. During the Goldstone Commission, police often could not be used to protect witnesses who were testifying in respect of police abuses of power. Some of the initial witnesses on the Goldstone Commission were reportedly more confident of their ability to protect themselves than they were of the Goldstone Commission's capacity to ensure their safety. The time consuming nature of witness protection also poses some dilemmas if it is to remove police personnel from an already under-staffed and under-resourced institution for sustained periods of time.

Although many of the security problems inherent in witness protection relate to the relocation of witnesses to places of safety, the witness is in fact most at risk when s/he has to appear to testify. It is at this time that the witness will be the most exposed to intimidation and danger from those who they are to testify against. It is thus crucial that in order for a witness' safety to be enhanced, legislation is adopted which seeks to specifically prohibit the various forms of intimidation that may occur and which sets out the relevant penalties for contravening the law.

Ultimately the safety of witnesses is dependant on a good security plan. As few people as possible must know the identity and location of any witness. All staff working on the programme must be properly trained in the relevant security procedures and given security clearances. At the same time, the identities of the staff need to be kept secret so as to protect them from harassment or intimidation.


Security measures should as a minimum include the following:

  • Tight internal security measures within the programme and the TRC, so that the information on witnesses is highly confidential. The database system which may contain information about the witnesses must be inaccessible to those without high security clearances and must be very tightly secured.

  • The programme should be able to make use of police and private individuals as guards for witnesses depending on the specific circumstances.

  • Although dependent on government resources and infrastructure, the witness protection programme must retain a high degree of independence of government and should not be subject to government probes or investigations except through its financial accountability to parliament.

  • Investigation units need to be set up to assess the threat to witnesses and establish where such threats may come from.

  • There needs to be some internal intelligence capability to strategically plan and to assess the potential areas of risk for any particular witness.

  • The investigation units also need to be able to investigate the witnesses to determine their credibility and reliability.

  • The potential threat of a witness to the community in which relocation has occurred needs to be evaluated.


The establishment of a witness protection programme demands careful deliberation and planning. Before attempts are made to extend or copy any particular model of witness protection, it is vital that more research is conducted and a strategy worked out as to the process involved in the establishment of such an initiative. All the various organisations and individuals who have worked on witness protection, including past witnesses, should be consulted during the process of conceptualising the programme. In this way, a comprehensive understanding of the issues involved and practical methods of dealing with these issues would be forthcoming.

Although foreign models of witness protection can be valuable in highlighting potential problem areas, as well as in offering practical solutions, it must be stressed that the specific objectives of the TRC witness protection programme demand a differentiated approach which is sensitive to this unique social and political project, as well as to such factors as the paucity and unevenness of infrastructure in some areas, the highly specific population mobility and networks in the country, as well as the striking socio-economic and cultural disparities which will impact on the programme.

It must be accepted that not all witnesses will need the same support and protection while on the programme, and provision must be made for innovative and cost effective alternatives. One method is to work out a categorisation system whereby witnesses are offered a certain type of protection programme determined by the threat against them, and their various needs. Witness protection does not necessarily mean that every witness has to be removed from their communities or place of abode. Depending on circumstances, if there is not a high risk factor involved, witnesses may be protected at their homes. Full time guards may be deployed, or home security improved and extra police patrols provided. Often, the most important thing is that the witness feel safe.

The ITU in Natal provides an innovative way of conducting witness protection, which may provide a working model. This programme is structured and operationalised in four distinct phases:

Phase 1
  • This is the initial phase where the witness comes forward, makes allegations and demands protection, or is offered such protection.

  • This phase should last for about two weeks during which time the witness and the allegations made are assessed.

  • The allegations are investigated to determine if they are valid or not.

  • The danger to the witness is assessed to establish what type of protection is needed.

  • This phase can be extended or shortened depending on the specific situation.

Phase 2
  • During this phase the terms of the witness protection programme for the particular witness is negotiated.

  • Experts in the fields of social work, psychology etc. are introduced during the negotiation phase.

  • During this phase the witness is given an accurate picture of what he/she can expect from the programme in terms of security, finances, future opportunities etc. No promises should be made that cannot be kept, the witness should be given the whole truth about their situation.

  • The witness must also be told what is to be expected from them in no uncertain terms.

  • These expectations can to a large degree be negotiated to allow the witness to feel part of the process and thus feel more secure and in control.

  • The third phase is the longest and coincides with the time that the witness is on the programme according to the conditions that have been mutually determined. Throughout this period there should be ongoing assessment of how the witness is fairing and relevant changes can be negotiated and made as determined by necessity. The witness can voluntarily come off the witness protection programme at any time, but should be debriefed before this happens.

Phase 4
  • This is the rehabilitation phase.

  • Although the witness should be prepared for rehabilitation from the very beginning, it is at this phase that the witness is de-briefed and actively integrated back into society.

  • Skills training and intensive counselling could occur during this time.

  • The witness should be assessed to establish if s/he is prepared for the future and capable of looking after him/herself independently of the witness protection programme.

  • An idea could be to give the witness a full salary for the first few months, and then slowly decrease this amount until s/he is totally independent.


A professionally run National Witness Programme which has credibility in the public eye would be a significant asset to the criminal justice system. For such a program to be effective, much planning and additional specialised research has to take place. The programme has to take into account the costs involved as well as the unique South African social context. Although much work will have to be done in order to establish an effective programme, the experiences of witness protection that have already taken place in South Africa can provide valuable lessons as to the various difficulties involved.

It is clear that a limited witness protection programme is critical to the TRC. There is no other similar structure which could effectively cater for witnesses who will be testifying before the commission. Given the importance of a national witness protection programme, the TRC offers the perfect opportunity upon which to begin the inevitable process of trial and error upon which the development of all other witness protection programmes have been based.

If a witness protection programme is to be successful it must be framed within the larger context of crime prevention. This means that it has to gain the institutional support of all involved in the criminal justice system. Police training should include a component concerned with witness protection and its role in effective crime prevention. This obviously has to be contextualised within the broader ambit of community-police relations. Every police station must be able to provide potential witnesses with the relevant information. There should also be a non-policing vehicles of access to such a witness protection programme. It is important that those administering the protection programme can move quickly and without hindrance. The witness protection programme should be able to supply the necessary protection and support to all valid candidates no matter from where the threat comes. If the establishment of an effective national witness protection programme can be realised, South African policing, and the society at large, will be the true beneficiaries.


1 s 35 (1) Promotion of National Unity and Reconciliation Act, no. 34 of 1995.

2 The one day conference was held in Pretoria on the 26th of July 1995 by the Institute for Democracy in South Africa (IDASA). Delegates included representatives from: The Ministry of Justice, the Attorney Generals Office, the South African Police Service, the legal profession, former covert operatives, academic institutions and non-governmental organisations.

3 Veriava, F., "Thrown to the Wolves", in Democracy in Action, Vol. 9, No. 3 (June 1995), p.9.

4 See Simpson, G., & Rauch, J., "Political Violence: 1991", in Boister, N. & Ferguson-Brown, K., South African Human Rights Yearbook 1992, Vol. 3, Oxford University Press: Cape Town (1993), pp. 212-239.

5 In fact the police have reported that they have identified approximately five hundred organised crime syndicates operation in South Africa. Sunday Independent(9 September 1995).

6 Lawyers for Human Rights (LHR), Representations to the Goldstone Commission on Violence and Intimidations: Establishment of a National Witness Protection Programme, Unpublished (1993), p. 38.

7 United States Marshals Service, Fact Sheet Prepared by the Office of Congressional and Public Affairs, Revised January 1992.

8 A statement released by the Office of Congressional and Public Affairs of the United States of America, January 1992.

9 Final Report of the Goldstone Commission, 27 October 1992.

10 LHR, opcit.

11 Colonel Frank Dutton, Private Interview, 10 July 1995.

12 The Star, 21 July 1995.

13 Sunday Independent, "MI officer's murder was a 'professional hit'", 11 August 1995.

14 Lawyers for Human Rights, opcit., p.21.

15 Grant, E., Cited in South African Journal on Human Rights, Vol. 5, Part 2, Juta and Co.: Johannesburg (1989), p. 221.

16 s 185A Criminal Procedure Act, No. 51 of 1997.

17 Regulation 7 of R2204.

18 Regulation 15 of R2204.

19 Although there was a reluctance on the part of the authorities to release specific figures of those presently receiving protection, an approximate figure was given as twenty witnesses. This seems to include protected witnesses involved in the present De Kock trial, as well as those concerned with prosecutions arising from the ITU investigations.

20 Goldstone Commission press release, 26 August 1992.

21 Major Klopper, Private Interview, 18 July 1995.

22 Paul Erasmus made mention of a number of occurrences where there were indications that his life was in danger, or that his security had been compromised while of the Goldstone witness protection programme.

23 Dr. Toerie Pretoias, Private Interview, 26 June 1995.

24 Nick Grobelaar, Private Interview, 21 July 1995.

25 Jenny Irish, Private Interview, 11 July 1995.

26 LHR, opcit., p.25.

27 This was generally agreed upon at the Conference on witness protection, 26 July 1995.

28 Glen Cuthbertson, Private Interview, 4 July 1995.

29 Aubrey Lekwane, Private Interview, 27 June 1995.

30 Toerie Pretorias, Private Interview, 6 July 1995.

31 Private Interview with Gladys Sibeka, coordinator of the NICRO rehabilitation programme for witnesses, (25 July 1995).

32 Morris, S.E., "Testifying Without Reprisal" in The Pentacle, Winter February 1998, p.3.

© Centre for the Study of Violence and Reconciliation

Copyright © 2019 CSVR. All Rights Reserved.