From: A Crime Against Humanity - Analysing the Repression of the Apartheid State edited by Max Coleman

The era of destabilisation entered its third year in July 1992 with the memory of the previous month's Boipatong massacre still fresh and negotiations having broken down. An atmosphere of profound mistrust of the apartheid government's real intentions now prevailed and serious doubts were being raised as to whether they had lost control of the elements used in the strategy of destabilisation.

Nevertheless, the year was marked by strenuous efforts by the liberation forces to pin the government down to a programme eliminating political violence and these efforts resulted in a Record of Understanding signed in September 1992 and ultimately in getting talks back on track in April 1993. The ultra right wing did their best to sabotage these efforts through the assassination of Chris Hani in April 1993 and to invade the venue of the now succeeding negotiations in June 1993. In spite of these setbacks, the third year of destabilisation closed on a high note with agreement to democratic elections now clearly in sight and the announcement of the date imminent.

The pattern of violence during the third year was largely an extension of the pattern during the second, namely, a war of attrition in which there were fewer major high-profile attacks (massacres declined to 12) but a high number of incidents (nearly 20% more) producing about the same number of deaths (3096) and exerting maximum disruption of township communities.

However, there was an important shift in the regional focus of the political violence, away from the PWV and into Natal from whence much of it had come originally. Natal accounted for 53% of all deaths, up from 33% for the previous year. PWV accounted for 35% of all deaths, down from 56% for the previous year. A shift of this magnitude suggests a deliberate decision to marshal forces on a provincial level.

The events and implications of this third year of destabilisation are described in the HRC publication entitled:

Three Years of Destabilisation (August 1993).

A record of political violence in South Africa, July 1990 - June 1993
HRC, August 1993

1. The measure of the violence
2. Regional analysis of the violence
3. Components of the violence
4. Victims of the violence
5. Conclusions


The early months of 1990 held out great promise for the dismantling of apartheid; the Nationalist government, under intense pressure from the majority population of South Africa and from the economic crisis stemming from international isolation, seemed to respond to the demands for free political activity as a precursor to negotiating a political settlement. However, before long a two- pronged strategy began to make its appearance -that of destabilising the now-unbanned liberation movement while simultaneously negotiating with it. The destabilising prong, utilising an array of forces with an interest in defending their apartheid power, whether initiated or inherited, began to manifest itself in July 1990 with a massacre in Sebokeng in the PWV region, and by the following month was in full cry with the deaths of over 700 people in a single month.

Three years on, we are still in the grip of destabilisation. It may well be that the original authors of destabilisation, namely, the Nationalist government, are no longer as enthusiastic about the prospects of benefiting from the strategy (since it has now become counter-productive) but nevertheless they now seem unable to control their own creation. It also seems apparent that there are significant elements within the security apparatus that continue to pursue this strategy.

1. The measure of political violence

Incidents of political violence

The number of incidents of political violence recorded by HRC during the 3-year period is as follows:
Year 1 - 2166
Year 2 - 3534
Year 3 - 4178
Total - 9878 incidents

The sharp increase in the number of incidents over the years is an indication more of the widespread nature of the violence over time rather than the intensity. As seen below, the greater number of incidents did not result in a greater number of deaths or injuries.

The death-count in political violence

The deaths in politically related circumstances recorded by HRC during the 3-year period are as follows:
Year 1 - 3190
Year 2 - 3039
Year 3 - 3096
Tota - 9325 deaths

This total represents an average of 259 deaths each month or 8.5 deaths each day.

The death toll for each of the 3 years has been remarkably constant with differences of less than 5%. However, this is simply coincidental, with extreme fluctuations taking place within the course of each year, as can be seen from the chronologies at the beginning of Part C.

Year 1 - lowest month 105 - highest 709 (6,8 times)
Year 2 - lowest month 110 - highest 437 (4,0 times)
Year 3 - lowest month 175 - highest 3 73 (2,1 times)

As time has gone by, the fluctuations have become less extreme, with the peaks reducing and the troughs rising, confirming the broadening of the violence as indicated by the increase in the number of incidents.

The major explosions and lulls which characterised year 1, came to be replaced by a different style of violence in year 3, which, while it produced much the same number of deaths, involved a much greater number of incidents, each of them claiming fewer victims on average.

This observation is confirmed by the decline in the number of major massacres, involving the deaths of 10 or more people which were recorded during the 3-year period as follows:
Year l - 34
Year 2 - 15
Year 3 - 12
Total - 61 massacres

A notable feature of the month by month death toll over the 3-year period has been the surges, which have taken place during months in which important political events have occurred. The correlation of the death toll with the political calendar is shown in the chronologies at the beginning of Part C.

The injury-count in political violence

Injuries were recorded as follows:
Year 1 - 6855
Year 2 - 5033
Year 3 - 5085
Total - 16 973 injuries (an average of 471 per month or 15.5 per day)

Injuries are very approximately proportional to deaths at almost 2:1 but injuries are far more difficult to assess with any accuracy since so many go unreported. The true figure must certainly be in excess of 20 000 in total with no way of knowing how many are permanent injuries.

2. Regional analysis of the violence

The regional breakdown of the deaths for the whole 3-year period is as follows:

Natal region 3653 (39.2% of total)
PWV region 4756 (51.0% of total)
Elsewhere 916 (9.8% of total)

Natal region

Deaths recorded in the Natal region during the 3-year period were as follows:
Year l 1004 (31% of total)
Year 2 1004 (33% of total)
Year 3 1645 (53% of total)
Total 3653 deaths (an average of 101 per month)

The Natal region is different from the rest of the country in that the current wave of political violence cannot be said to have started in July 1990 in this region. In fact, the political conflict in Natal dates back to at least 1980 when resistance emerged to incorporation of townships within the homeland of KwaZulu and educational and other conflicts arose. The emergence of the United Democratic Front (UDF) in 1984 as a strong unifying political force with popular support throughout the country, including Natal, intensified the conflict. This popularity was perceived by elements within the KwaZulu homeland as a threat to their thus-far unchallenged control. A struggle for political allegiance ensued and developed into a conflict in which both aggressor and victim were Zulu-speaking; clearly a political conflict and not an ethnic conflict as some have attempted to portray.

The 'Natal War' has since 1984 claimed the lives of around 7500 victims including 3653 during the 3-year period covered by this report. As can be seen from Fig. 18, the conflict in Natal has an unremitting character with deaths occurring month after month without a lull. In years 1 and 2, the average monthly death toll was constant at 84 with relatively little fluctuation from month to month. In year 3, a shift in gear took place with the average jumping to 137 (an increase of 64%) but still with little monthly fluctuation. The Natal region accounted for one third of the countrywide deaths in years 1 and 2 but jumped to over half in year 3. It can be concluded that an intensification of the 'Natal War' took place in year 3. Nineteen major massacres occurred in Natal during the 3 years

PWV region

Deaths recorded in the PWV region during the 3-year period were as follows:
Year 1 - 1982 (62% of total)
Year 2 - 1688 (56% of total)
Year 3 - 1086 (35% of total)
Total 4756 deaths (an average of 132 per month)

Unlike the pattern in Natal, the character of the political violence in the PWV region is a highly volatile one with considerable fluctuations from month to month (see Fig. 19). After its ignition in July 1990, the PWV violence followed a roller-coaster course, evident to this day, and is clearly the underlying cause of the fluctuations in the national profile of political violence described in an earlier section. In fact the PWV region can be described as the barometer of political activity and is the region most affected by political events. The volatility of the PWV region is evident from the following death toll figures:
Year 1 - lowest month 13 - highest 570 (43.8 times)
Year 2 - lowest month 40 - highest 281 ( 7.0 times)
Year 3 - lowest month 36 - highest 182 ( 5.1 times)

Furthermore, a high proportion of major massacres has taken place in the PWV region (41 out of 61).

Noteworthy is the drop in the death toll during year 3 by 36% from the previous year. After accounting for around 60% of the total countrywide deaths in years 1 and 2, the PWV region accounted for only 35% of the country total in year 3. This drop occurred simultaneously with the escalation in Natal and raises questions about a switch in tactics and resources.

The East Rand is the most unstable sub-region within the PWV and towards the end of year 3 was contributing over 60% of PWV deaths. The Vaal also has a history of instability, which is showing no signs of improvement. By contrast the situation in Soweto and Alexandra has improved greatly and towards the end of year 3 both were dipping below 10%.

Hostels have played a key role in the PWV political violence. An important factor in the abatement of violence particularly in the Soweto and Johannesburg sub-regions during year 3 has been the initiatives of hostel dwellers themselves to reach peace accords. Another has been the pressure placed on the government to make proper use of the security forces to secure the hostels.

Other regions

Deaths recorded in regions other than Natal and PWV during the 3-year period were as follows:
Year 1 - 204 (7% of total)
Year 2 - 347 (11% of total)
Year 3 - 365 (12% of total)
Total 916 deaths (an average of 25 per month)

These figures show a trend towards a broadening of the violence to regions outside of Natal and PWV as time goes on and reached 15% of the countrywide total in the second half of year 3.

3. Components of the violence

In terms of HRC categories of components of violence, the following deaths were recorded during the 3-year period:

General incidents 8580 (92.0%)
Security forces 518(5.6%)
Hit squads ' 173(1.8%)
Right wing 54 (0.6%)
Total 9325 deaths

General incidents

This category includes political violence, which is based within township, rural and city communities, frequently hostel-related and frequently involving the activities of vigilante groupings against anti-apartheid political activists in particular, but also against whole communities at large. For the purposes of this report it also includes a range of actions emanating from elements normally associated with the liberation struggle but which can often only be seen as impeding progress towards democracy; such actions include sectional conflict, violence arising from industrial and educational conflict, attacks launched upon members of the security forces, attacks on white civilians, and the so-called taxi wars. While several of these actions can be understood as a backlash against destabilisation, they are frequently counter-productive in neutralising destabilisation.

Incidents recorded within this category were:
Year 1 - 1093
Year 2 - 2385
Year 3 - 3283
Total 6761 incidents (68.4% of total)

Deaths recorded within this category Were:
Year 1 - 2903
Year 2 - 2806
Year 3 - 2871
Total - 8580 deaths (92.0% of total)

Injuries in this category were:
Year 1 - 4315
Year 2 - 3864
Year 3 - 2806
Total - 10 985 injuries (64.7% of the total).

It can be seen that while the number of deaths has not differed much from year to year, the number of incidents has climbed appreciably while the number of reported injuries has decreased.

Security forces

The security forces include the South African Police (SAP), the South African Defence Force (SADF), homelands police and armies, municipal and council police and other parastatal law enforcement entities. Actions involving the security forces during the 3-year period in a political context were as follows:

Incidents recorded:

Year 1 - 881
Year 2 - 909
Year 3 - 779
Total 2569 incidents (26.0% of total)

Deaths recorded:
Year 1 - 238
Year 2 - 114
Year 3 - 166
Total 518 deaths (5.6% of total)

Injuries recorded:
Year 1 - 2248
Year 2 - 1033
Year 3 - 2061
Total 5342 injuries (31.5% of total)

While the proportion of deaths inflicted by the security forces is relatively low (5.6%), the involvement in incidents of violence is high (26.0%) and the infliction of injuries is higher still (31.5%).

In addition the security forces effected the following number of arrests during political activity:

Arrests recorded:

Year 1 8211 Year 2 8725 Year 3 9137 Total 26 073 arrests (an average of 724 per month)

The security forces are also alleged to have been directly responsible for, or directly involved in a number of massacres:

• SAP for the deaths of 12 in Daveyton on 24 March 1991.

• SADF for the deaths of 11 in Sebokeng on 4 September 1990.

• SADF for the deaths of 12 in Esikhaweni on 16 February 1992.

• KZP for the deaths of 18 in Umlazi on 13 March 1992.

• Ciskei Defence Force for the deaths of 28 in Bisho on 7 September 1992.

The above record does not take into account the numerous allegations and reports of security force complicity in fuelling or condoning political violence.

Hit squads

The activities of professional hit squads located primarily within state structures has been recorded as follows:

Incidents recorded:

Year 1 - 99
Year 2 - 126
Year 3 - 34
Total 259 incidents

Assassinations recorded:

Year 1 - 28
Year 2 - 96
Year 3 - 49
Total 173 deaths

Injuries recorded:

Year 1 - 46
Year 2 - 35
Year 3 - 18
Total 99 injuries.

After a major increase in activity in year 2, there has been a dramatic fall-off in year 3, especially in the second half (January to June 1993) when only 2 incidents were reported, resulting in 2 deaths and 1 injury. There may, therefore, be some validity in the government's claim that all such operations have been terminated.

Right wing

Right wing violence is of 2 kinds, the one perpetrated indiscriminately against blacks in a more or less spontaneous manner by individuals or small groups of individuals; the other involving organisation and planning. Together they have produced the following:

Incidents recorded:

Year 1 - 93
Year 2 - 114
Year 3 - 82
Total 289 incidents

Deaths recorded:

Year 1 - 21
Year 2 - 23
Year 3 - 10
Total 54 deaths

Injuries recorded:

Year 1 - 246
Year 2 - 101
Year 3 - 200
Total 547 injuries

The ratio of injuries to deaths is much higher than for any other component of violence, indicating the level of assault and aggression employed. The killing of Chris Hani in April 1993 was an act with devastating consequences to the levels of violence throughout the country.

4. Victims of the violence

The overwhelming majority of the nearly 10 000 dead and 20 000 injured have been the ordinary residents of black townships throughout the country. Their lives during these 3 years of destablisation have been turned upside down and in so many cases, destroyed. Until year 3, the white community was virtually unscathed and largely unaware of the suffering being endured by their black neighbours. But as time went on, the violence began to permeate almost every corner of South African society in one way or another. Some specific groups or categories of victims are examined below:


Many township communities, particularly in the PWV region, in addition to attacks upon their homes, have been exposed to serious dangers while travelling to and from their places of work by train, bus or taxi.

Train commuters:

Year 1 - 67 deaths - 284 injuries - 16 incidents
Year 2 - 227 deaths - 566 injuries - 230 incidents
Year 3 - 107 deaths - 217 injuries - 110 incidents
Total 401 deaths - 1067 injuries - 356 incidents.

Clearly this form of violence peaked in year 2 and showed signs of abating in year 3 in the face of a concerted clamour for countermeasures to be instituted. However, attacks on train commuters still continue.

Bus and taxi commuters:

Year 1 no figures available
Year 2 - 119 deaths - 230 injuries - 72 incidents
Year 3 - 84 deaths - 161 injuries - 53 incidents
Total - 203 deaths - 391 injuries - 125 incidents

These attacks serve to supplement train attacks in terrorising and destabilising communities.

Women and children

The ultimate in terror-tactics is to kill women and children and thereby strike at the very heart of the community. While some of these women and children will have been simply 'cross-fire' victims there have been numerous instances where their slaughter was deliberate. The record is as follows:


Year 1 no figures available
Year 2 - 189 deaths - 227 injuries
Year 3 - 253 deaths - 315 injuries


Year 1 no figures available
Year 2 - 106 deaths - 87 injuries
Year 3 - 58 deaths - 211 injuries

The combined totals for women and children represent 10% of all deaths. In the notorious Boipatong massacre, 25 out of 46 dead were women and children, a clear example of deliberate intent.

Security force members

Attacks on members of security forces and upon their vehicles and bases have increased in frequency and intensity, as the following figures show:
Year 1 no figures available
Year 2 - 68 deaths - 279 injuries - 205 incidents
Year 3 - 200 deaths - 264 injuries - 569 incidents

Nearly 14% of all incidents of political violence in year 3 were attacks on the security forces.

White civilians

Attacks on white civilians in a politically related context only began to make their appearance in year 3, during which HRC recorded the following:
34 deaths - 43 injuries - 52 incidents

This represents only around 1% of the casualties for the year but has raised concern in most quarters that efforts to carry the violence into the white community can only impact negatively towards an abatement of the violence.

5. Conclusions

Faced with the staggering record of political violence outlined in this report, the 2 questions most frequently asked are:
• What is the identity of the real perpetrators of this violence?
• What does it take to bring this violence to an end?

The perpetrators

In broad terms, all South Africans today sooner or later have to choose between belonging to the pro-democracy forces or the anti-democracy forces in the country. Those who have a vested interest in apartheid power and are reluctant to relinquish that power, place themselves squarely in the anti- democracy camp and collectively constitute the base of the so-called 'third force'. They are collectively responsible for the violence of destabilisation.

The past beneficiaries of apartheid power are not difficult to identify and this has been done in some detail in the previous chapter. In summary they were the stakeholders in the 3 primary structures in which apartheid power was vested:
1. Tricameral parliament
2. Homelands
3. Black local authorities (BLA)

The tricameral parliament stakeholders include:
• The white electorate, white parliamentary parties, white government, the security establishment, the civil service and the business community;
• members and supporters of the 'coloured' House of Representatives;
• members and supporters of the 'Indian' House of Delegates.

The homelands stakeholders include the governments, administrations, security forces and supporters of 6 'self-governing' and 4 'independent' entities.
The black local authorities stakeholders are those who took office as black councillors, along with their hangers-on.

Some erstwhile stakeholders have fallen away, either through the collapse of their structures (as with most black councils) or their displacement from power (as with coups against homeland leaders). Other stakeholders have publicly committed themselves in advance to abrogating their power in favour of a unitary democratic constitution while elements within white, 'coloured' and 'Indian' structures and constituencies have openly committed themselves to a negotiated democracy and thereby renounced their claims to the privileges and benefits of apartheid power.

However, there remains a hard core of stakeholders who cling to their past power, have no intention of relinquishing it if they can possibly help it and actively oppose the arrival of a non-racial democracy. They are to be found within all of the structures and constituencies listed above, but particularly within the white right wing, the security establishment and a number of homelands. Together they constitute the anti-democracy camp in South Africa and together they are responsible for the activities of what is referred to as the 'third force'. Who else could possibly benefit from the apparently mindless and indiscriminate destabilisation of whole communities through such untargeted violence as the random slaughter of train commuters, taxi commuters, tavern patrons, beer hall patrons and funeral mourners?

Before leaving the issue of the identity of the perpetrators of the political violence, a word needs to be said about the criminal factor in the violence. There can be no doubt that some of the violence projected as politically inspired involves the hand of apolitical criminals but at 2 different levels. The one level is simply criminal activity for personal gain under the smokescreen of political violence; taking advantage of a turbulent situation and perhaps even promoting it; engaging in protection and extortion rackets; and even seeking cover within hostels and other community structures, often in 'alignment' with one or other political grouping. The other level is availability for hire, for involvement in planned assassination and massacres. Such hitmen are simply hired guns with special expertise paid by politically motivated sponsors.

Ending the violence

Three critical phases are crucial in bringing political violence to an end:

• the control of violence during the election campaign;
• the election itself;
• acceptance of the election results.

Violence during electioneering

South Africa is already effectively into a period of election campaigning, which will culminate in election day on 27 April 1994. Crucial to the control or minimising of political violence during this period will be the proper and effective use of the total available security machinery of the country and the energetic promotion of peace initiatives and political tolerance through the already well-established Peace Accord structures and international observer teams. The security machinery needs to be under the multi-party control of the Transitional Executive Council (TEC) to ensure that its efforts are totally directed towards promoting and supporting free and peaceful political expression and participation. At the same time its considerable resources must be brought to bear instantly and resolutely when potentially violent situations arise, either from' covert or overt sources. The resources exist but their improper use in the past has been part of the problem instead of the solution.

The various peace-promoting structures with their capacity to foster and encourage a peace culture and tolerance and to intervene in tense situations also have a vital role to play. They need to be fully supported and expanded, both politically and financially, in order to maximise their effectiveness.

At the rate of political killings of the last 3 years, 2600 people will die during the 10 month period from July 1993 to 27 April 1994, election day. Everything will depend upon the twin efforts of the peace-keeping and the peace-promoting structures, described above, as to whether that frightening figure can be substantially reduced or whether it will even be exceeded.

The election
The importance of the election itself in bringing an end to the violence, cannot be overestimated. Provided it is conducted in a free and fair manner, it will reveal the true support base for each and every participant and put an end to speculation about where the support lies. Political non-entities will be exposed and relegated to the rubbish heap of history; minor players will be cut down to size. Political posturing and grandstanding on the basis of imagined support will no longer have any meaning or impact and trying to win support by violence, threats, intimidation and dirty tricks will no longer have any relevance.

The Namibian experience is instructive. After a dirty election campaign into which the South African government had poured 100 million rand in an effort to affect the outcome by destabilising South West African People's Organisation (SWAPO) and promoting Democratic Turnhalle Alliance (DTA), the electorate's decision became known and a constitution was expeditiously adopted. Almost overnight political violence disappeared in spite of dire predictions to the contrary by those who warned that the ethnic diversity of Namibia would bedevil all efforts to bring about peace. But the end of the political struggle for power signalled by the election result brought with it the end to political violence. The purpose of such violence no longer applied.

In South Africa can we expect the same to happen? Certainly the violence related to the political jockeying for position must end once the election results are known. But will there then be dissidents who will not accept the results, will refuse to recognise them and have the power to act upon their refusal?

Acceptance of the results
If one examines the possible existence of groups powerful enough to fill the role of spoiler after an election and with enough muscle to mount a sustainable campaign of violence to thwart the result, then not one emerges at the level of UNITA in Angola, which is the example constantly referred to when pondering the acceptance of election results. The key questions in this issue are support base and sustainability; on both counts the 3 possible candidates fall short.

The first possible spoiler is the white political right wing including the members and supporters of the Conservative Party (CP) and extra-parliamentary groupings to the right of the CP. It could also include some disaffected members and supporters of the National Party. At the outside we are looking at 5% of the total population and probably less than 1% of total population if one is considering only those who are prepared to bear arms and engage in a violent confrontation. From where material support for this hard core would be forthcoming is difficult to imagine especially in a world where there will be strong international pressure for the acceptance of the outcome of democratic elections and for an end to apartheid.

The second possible spoiler is the white military right wing including elements in the security establishment, which are unwilling to be part of the transformation of police, defence and intelligence into structures ready to serve a democratically elected government. It could also include members of the security portfolios within the Nationalist government. There is also an obvious overlap with the political right wing. However, material support and sustainability would be totally in the hands of the elected government and dependent upon a healthy economy. A military coup is totally out of the question having regard to the high proportion of blacks in both the police and defence forces. There must also be a significant number of whites in both forces who accept that their professional careers and aspirations will survive and even be advanced in a non-racial democracy.

The third possible spoiler is the black right wing in the form of governments and administrations of those homelands that struggle bitterly to cling to their inherited power. They can be expected to draw support from the white right wing, each for their own reasons, in spite of being the strangest of bedfellows. However, a big question mark hangs over the support they can expect from their own populations where the desire for democratic freedom is extremely compelling. Another compelling factor against sustainability of a go-it-alone attempt is the question of economic viability. During 1992, about 15% of the national budget (or nearly 14 billion rand) went as straight non-repayable grants to prop up the 10 homelands of which the greatest single beneficiary was KwaZulu at R3.148 billion (Business Day, 1 September 1992). In addition it was reported in Business Day on 15 September 1992, that loans in the region of 7 to 8 billion rand made by South African commercial banks and the Development Bank of Southern Africa to the 4 TBVC 'states', are guaranteed by the South African government. This generous support from parent to offspring will disappear on election day, if not before. Replacing it from other sources would be an impossible task, and collapse inevitable.

In summary, the HRC has strong hopes that a free and fair election will dramatically reduce political violence and that any attempts to reject the result will be non-sustainable. However, long-term stability and peace are going to depend on how a democratically elected government performs in delivering the fruits of democracy to all the people of South Africa. Therein lies our best guarantee that violence and destabilisation will never return to haunt us again

<< Previous chapter Next chapter >>