SA Farm Attack Report [02]: Incidence and Nature of Farm Attacks

Report of the Committee of Enquiry into Farm Attacks, 31 July 2003

• 00: Table of Contents (PDF)
  
• 01: The Appointment of the Committee (PDF)
  
• 02: Incidence and Nature of Farm Attacks (PDF)
  
• 03: Examples of Farm Attacks (PDF)
  
• 04: Case Studies: Direct Attacks (PDF)
  
• 05: Case Studies: Land Invasions (PDF)
  
• 06: Victims of Farm Attacks (PDF)
  
• 07: Perpetrators of Farm Attacks (PDF)
  
• 08: Investigating Officers and Prosecutors (PDF)
  
• 09: Submissions to the Committee (PDF)
  
• 10: Literature Review (PDF)
  
• 11: The Farming Community (PDF)
  
• 12: Crime and Its Causes in South Africa (PDF)
  
• 13: The Criminal Justice System and Farm Protection (PDF)
  
• 14: Comparative Studies (PDF)
  
• 15: Legislation on Land and Land Reform (PDF)
  
• 16: Security on Farms and Smallholdings (PDF)
  
• 17: Trauma and its Treatment (PDF)
  
• 18: Conclusions and Recommendations (PDF)
  
• 19: Summary of the Report (PDF)
  
• 20: Oral and Written Submissions and Bibliography (PDF)

Report of the Committee of Enquiry into Farm Attacks

31 July 2003

THE INCIDENCE AND NATURE OF FARM ATTACKS


INTRODUCTION

Farm Murders in South Africa (Carte Blanche 1/2) Farm Murders in South Africa (Carte Blanche 2/2)

All experienced investigating officers interviewed by members of the Committee are of the opinion that farm attacks are not a recent phenomenon. At the same time the indications are that farm attacks have been increasing over especially the last decade. The statistical data on this phenomenon, however, are not readily available in the crime registers of the SAPS, now the computerised Crime Administration System (CAS). The reason is simply that there is no such crime as a farm attack. A farm attack is the manifestation of other crimes, usually robbery or housebreaking and robbery, sometimes accompanied by murder, rape, etc.

Statistics on farm attacks have been collected by mainly three bodies, viz. the South African Agricultural Union (SAAU, now Agri SA), the South African Police Service, through their Crime Information Analysis Centre (CIAC), and the National Operational Coordinating Committee (NOCOC), in which the South African National Defence Force is very involved. As will be seen below, none of them has statistics which are completely accurate.

The South African Agricultural Union started collecting statistics on farm attacks and murders in 1991. Because their statistics showed an upward trend in the incidence of farm attacks, the SAAU requested at a meeting of the NOCOC on 4 February 1997 that urgent steps be taken to deal with the problem.

One of the results of this request was that the CIAC started collecting data on the incidence of farm attacks so that the extent of the problem could be evaluated. This they did by means of the so-called incidence report, which the police had to complete at station level after each farm attack. This was really an incident report, giving particulars of the type of attack, name of the farm or plot, nearest town, time and date, investigating officer, particulars of victims and suspects, description of the incident, weapons used, items robbed, suspected involvement of farm workers, the presence of farm shops or stalls, arrests made, any apparent motive, security forces involved, etc.

In 1998 farm attacks also became a so-called priority crime. The National Operational Coordinating structure had priority committees to deal with various problem areas, such as gang violence, taxi violence, cash-in-transit robberies, bank robberies and political violence. Priority committees on rural protection (including security on farms and smallholdings) were added in 1998. The NOCOC therefore started keeping statistics on farm attacks for operational purposes. These statistics are collected primarily by means of the so-called daily situation report on all priority crimes, which is submitted by each policing area on an almost daily basis, and also by means of reports received from SANDF commandos involved in operations against farm attackers. These statistics are collated on the premises of the police head quarters in Pretoria by designated staff of the SANDF.

The deficiencies of the statistical databases on farm attacks assembled by the three bodies quickly became clear to the Committee. Neither the incidence reports, nor the daily situation reports, nor the reports by the commandos proved to be totally reliable. Consequently the CIAC and the NOCOC now have a meeting once a week to compare statistics. They try to gather information on farm attacks by whatever means available, including newspaper reports and reports from the various agricultural unions, and, if necessary, full reports are then requested from local police stations. After the creation of the Committee the discrepancies became especially apparent and problematic, and both the CIAC and the NOCOC now make a special effort to have databases which are as complete and accurate as possible. They correlated their figures for 2001, and since then their databases have been in agreement.

In its latest report on farm attacks, the Crime Information Analysis Centre analyses farm attacks to a much greater extent than in any previous report. The report contains a great deal of very useful information. Although the report had not been made public yet, it was made available to the Committee in January 2003. Much of the information in this chapter is derived from the CIAC report. The CIAC also supplied the Committee with other statistical information, and recalculated certain figures to meet with the requirements of the Committee.



THE INCIDENCE OF FARM ATTACKS

The Committee attempted to collate the statistics on farm attacks from the various
sources, but found that where the statistics overlapped they differed quite considerably. It is therefore difficult to construct a complete statistical picture.


The different statistics

The figures on farm attacks registered by the South African Agricultural Union (now Agri SA) between 1991 and 1997 are given in Table 1.

It is not clear how accurate these figures are. The SAAU got its information primarily, though not exclusively, from local agricultural societies, and the likelihood is that the emphasis might have fallen on commercial farmers, rather than on, for example, smallholders not affiliated to the agricultural societies. These statistics were obviously not intended to be totally complete, but rather sufficient to identify the problem and to demonstrate trends.

The statistics of the CIAC for attacks on farms and smallholdings are given in Table 2. (Although the figures for 2001 have not been made public yet, the Committee has
permission to use them. They may still change in very minor respects.)



Although the information is intended to be as accurate as possible, the CIAC had the following to say in May 1999: ‘The CIAC is still convinced that not all information available at ground level is reported to it or the Crime Intelligence Management Centre by all the parties involved at that level.’ Supt. J.C. Strauss from the CIAC confirmed to the Committee that even as late as 2002 it was still a problem, especially in some areas in KwaZulu-Natal, although the situation had improved.



The NOCOC made its whole database on farm attacks available to the Committee. Stretching from 1998 to 2001, it contains information on some 3544 farm attacks. (See Table 3.) Initially this information was gathered mainly through the commando system and later the daily situation reports. It should be noted that the statistics are intended primarily for operational purposes. The NOCOC processes the information as soon as possible after an attack, and it is not necessary for its purposes to be either absolutely accurate or to follow up the information. If a victim of a farm attack dies some time later, for example, the database may not necessarily be updated immediately to indicate the death. It is estimated that the NOCOC database is about 90% reliable.

Generally speaking, the statistics for 1998 and 1999 kept by the NOCOC are higher than those of the CIAC, while during 2000 and 2001 the position is reversed. The reason for this is easy to explain: the Committee was informed by Supt. J.C. Strauss that due to teething problems, it had taken quite a while to get the local station commissioners to report farm attacks regularly and accurately. In fact, he also gave this as the reason for the huge increase in the CIAC figures for farm attacks between 1997 and 1998. As the reporting improved, the CIAC figures rose, and it is to be expected that from 2000 they were higher than the NOCOC figures.



The combined statistics


The situation therefore is that the figures of neither the SAAU, the NOCOC or the CIAC are completely accurate. The Committee has nevertheless attempted to construct as complete a picture as possible of the incidence of farm attacks since 1991. (See Figure 1.)



The figures between 1991 and 1996 are those of the SAAU, because they are the only ones available for that period. The CIAC figures for 1997 are so unreliable that the SAAU statistics are used for those years as well. For the period 1998 to 2001 the CIAC figures are used, although the indications are that their figures for 1998 and 1999 may be less accurate than those of the NOCOC. However, the CIAC’s effort to improve the accuracy of the incidence reports, together with the cooperation of the NOCOC over the last few years, means that their figures have become more reliable. The total number of attacks registered in the eleven years between 1991 and 2001 came to 6564, but in reality was probably somewhat higher.

It is clear from the above tables that the number of recorded farms attacks has been on the increase at least since 1991. The Committee is of the opinion, however, that the increase is less dramatic than would appear from Figure 3. The reason is that it is doubtful whether the SAAU statistics covered all farm attacks. As indicated above, they would no doubt have concentrated on commercial farmers. It is also doubtful whether all attacks on other farm residents, which may now qualify as a farm attacks in terms of the definition, would have been included in their statistics. This means that the incidence of farm attacks, as defined at present, might well have been higher than is indicated by the available statistics, especially during the first half of the 1990s. Furthermore, the CIAC also acknowledges that there was an underreporting of incidents, especially during the first few years that it kept statistics on farm attacks. Therefore the increase in the number of incidents over the last few years may partially be due to the improvement in reporting.

As far as murders during farm attacks are concerned, the combined figures of the SAAU and CIAC indicate that 1254 people were killed between 1991 and 2001. The SAAU figures indicate that the rate per incident fluctuated between 17.3% and 23.3% between the years 1991 to 1996. The next year (1997) saw a jump in the murder rate to 30.2%, which accompanied a sharp drop in the number of incidents. This is completely inexplicable, and the Committee is satisfied that it was largely due to inaccurate and incomplete reporting.

The statistics kept by the CIAC indicate a consistent decrease in the murder rate from 18.5% in 1998 to 14.5% in 2001. (The CIAC figure for 1997 is too unreliable from which to draw conclusions.) This would seem to indicate that farm attacks are becoming less violent. However, it is reasonable to assume that underreporting of farm attacks in the past affected mainly the less serious cases and not cases involving murder. There is therefore the possibility that, because less serious farm attacks are now being recorded to a greater extent than before, the murder rate is being reduced at the same time. In real terms however, the murders remained constant at between 142 and 147 from 1998 to 2001. (It should be noted that in fact the number of murders during 2001 may be slightly less than 147 because there is still some dispute whether two of the murders recorded in the Eastern Cape were in fact committed during the course of farm attacks.)


The sudden increase in incidents between 1997 and 1998

Much has been made of the dramatic increase in the number of recorded farm attacks between 1997 and 1998, which saw a jump of 63.6%. Some saw that as evidence of the beginning of some sort of concerted and organised campaign against farmers. The
Committee is convinced that that was not the case at all. First of all, as will be seen from Figure 1 there was a steady upward trend, in an almost perfectly linear pattern, in the incidence of farm attacks from 1991 to 2001 inclusive, with the exception of 1996 and 1997. The sudden and sharp drop in the incidence of farm attacks in the SAAU statistics is quite inexplicable and, as will be seen below, the statistics also indicate a sudden increase in the murder rate in 1997, which is even more inexplicable. This may well indicate inaccurate reporting rather than reality. Of course, the official figure for 1997 kept by the CIAC is even lower than the SAAU figure, but the Committee is satisfied that that was due to the teething problems experienced with the new-introduced reporting system for farm attacks. Attacks on smallholdings during 1997, especially, were not being reported.

This conclusion is further strengthened by the remarks in a CIAC report: ‘(T)he broad patterns observed during the first six months of 1998 with regard to the distribution of crimes accompanying the attacks on farms and smallholdings, the modus operandi employed by criminals involved, the motives associated with attacks and the general profile of both victims and attackers did not deviate much from the analysis pertaining to attacks during 1997’. If there were some sinister force behind the apparent increase in farm attacks, one would have expected there to be other differences indicating, for example, a different motive for the attacks.

The Committee should not be understood to imply that farm attacks did not increase during the last decade. All the evidence points to an increase in the number of incidents and underreporting could only have had a marginal effect.



Incidents per province

In its 2001 report, the CIAC also gives the incidence of farm attacks per province. (See Table 4.)



It will be seen that the figures for Eastern Cape, Free State and KwaZulu-Natal decreased between 1998 and 2001. There is some doubt about the sharp decrease for KwaZulu-Natal, and the CIAC is of the opinion that there may have been some underreporting. The figures for Gauteng, Mpumalanga and North West have shown a sharp increase, while those of Limpopo, Northern Cape and Western Cape remained fairly constant over this period.

The CIAC also analysed the murders committed during farm attacks in each province. (See Table 5.) From this information it is possible to calculate the average murder rate per province. The figures vary too much from year to year, however, to draw valid conclusions, and a comparison of the annual figures for 1998 to 2001 is more useful. (See Table 6.)



Referring to Table 6, it is noteworthy that, proportionally, most murders occurred in KwaZulu-Natal, Northern Cape and Western Cape. It may be significant that those three provinces are also consistently amongst the provinces with the highest incidence of violent crimes (murder, rape and serious assault) in general in South Africa.




Farm attacks during 2002

The CIAC made their provisional figures for farm attacks during 2002 available to the Committee, on the understanding that it should be made clear that the figures have not been verified. (See Table 7.) The CIAC warned that no definite conclusions should be drawn from them.

It will be seen immediately that, while the total number of attacks decreased by 1.1%, from 1011 in 2001 to exactly 1000 in 2002, the number of murders decreased by 24.5%, from 147 to 112. Supt J.C. Strauss from the CIAC and Supt R. Pretorius of Crime Information Management Centre (CIMC) are both at a loss to explain this drastic reduction in the murder rate. It is possible that farmers now have a greater awareness of security, and that they tend to handle farm attacks better.



The Eastern Cape showed the most marked decrease, both in terms of the number of attacks and the rate of murder. This change cannot be explained easily, but it may be significant that the Committee found the Eastern Cape to be one of the provinces where the SAPS was most active in dealing with the problem of farm attacks. Not only did the provincial CIAC office ensure that it was keeping accurate statistics, but the police there, in conjunction with the Eastern Province Agricultural Union, were also very active in the promotion of safety on farms. The Free State, Northern Cape and Western Cape also showed a marked reduction in the murder rate, but their figures are relatively small and may not accurately reflect trends.


Incidence of attacks as a proportion of farm or smallholding population.

The prevalence of crime is usually calculated in terms of its incidence per 100 000 of the population during the period of one year. That would have been a useful tool to determine whether farm and smallholding residents are at a greater risk of being subjected to the crimes most associated with attacks on them than the population in general. Unfortunately, it has proved to be virtually impossible to do so, or even to calculate the incidence of attacks as proportion of the total number of farms and smallholdings. The reason for this is that the total number of farmers, farm residents or even farms is unknown. The same applies to smallholdings.

In an article Antoinette Louw, a researcher at the Institute for Security Studies, used the Britz-Seyisi report as a basis for drawing certain conclusions about the murder rate for farm residents (farmers, farm workers and their dependants). From the fact that only 31 murders occurred in 198 attacks on farms (excluding smallholdings) during the first five months of 1998, and assuming that there were 4.5 million farm residents in South Africa, she calculated that the murder rate on farms was only 0.6 people per 100 000 of the population (which translates in 2.4 per 100 000 for the whole year). This she compared to the general murder rate for the whole of the country, which was 13 per 100 000 for the first three months of 1998, with a projected total of 52 per 100 000 for the whole year.

Apart from some minor mistakes in her calculations, there are two serious flaws in her reasoning. Firstly, it is impossible to say how many persons are living on South African farms. Secondly, by definition all social fabric crimes are excluded from farm attacks. Those are crimes of murder, rape, serious assault and indecent assault, usually committed by someone known to the victim and often involving the misuse of alcohol. Social fabric crime is rife in rural areas, and there is no reason to assume that it is any different on farms.

On the other hand, Professor Willem Naudé and Larette van Rensburg of the Faculty of Economic and Management Sciences at Potchefstroom University, stated that in South Africa ‘a farmer is four times more likely to be murdered than a member of the general population’. It is not clear on what figure the two academics based their calculations. The general murder rate for 1999 was 55 per 100 000 of the population. There were 144 farm murders in 1999, and if one relies on the official figures of about 61 000 commercial farming units, it would give a murder rate on farms of 236 per 100 000, roughly four times the general murder rate. However, that does not take into account that not only farmers but also their families, and the workers and their families, fall victim to farm attacks.

Kwanalu (the Kwa-Zulu-Natal section of Agri SA) calculated that farmers in that province were targeted for attack at between two and four times the rate of the overall population. It would appear, however, that the figures used by Kwanalu for their calculations were seriously flawed.


Crimes committed during farm attacks as a percentage of crimes in general

As indicated above, the incidence of crime is usually measured as the number of incidents per 100 000 of the population. During 2001 some 21 180 murders were committed in South Africa. With a population of 48 820 000, according to the 2001 statistics, the incidence of murder is therefore 47 persons per 100 000 of the population. Unfortunately, because of the difficulty of establishing the number of farmers, farm workers and their dependants, the incidence of farm attacks cannot be determined on that basis. It is possible to determine what percentage of particular crimes in South Africa are committed during farm attacks. (See Table 9.)

Of the 21 180 murders committed in South Africa during 2001, 147 or 0.69% were committed during farm attacks. This correlates reasonably well with the percentage of .82% for attempted murder. According to the figures for rape only 0.13% of all recorded rapes in 2001 occurred during farm attacks. This is quite significant, as it confirms that rape is not one of the main motives of farm attacks, and is in fact relatively rare. Rape is a notoriously underreported crime, but that would apply to rapes in general and not only rapes during farm attacks.



There can be little doubt, however, that the percentage for armed robbery is quite incorrect, because it must be assumed that substantially more than 689 robberies were committed during farm attacks. As said above, the CAS system of the police may not always register robberies if a more serious offence, such as murder or rape, has been committed.



Attacks on smallholdings

Some writers have expressed the view that attacks on smallholdings have inflated the combined figure for attacks on farms and smallholdings. It is only since 2001 that the CIAC has made a distinction between attacks on farms and smallholdings for statistical purposes. The figures are given in Table 9.



The figures indicate that 62.3% of the premises targeted during 2001 were farms and 37.7% were smallholdings. However, the CIAC specifically draws attention to the fact that the figures for Gauteng tend to distort the picture, since the 93.5% for smallholding attacks in that province is by far the highest of all the provinces. In fact, of the 381 attacks on smallholdings, Gauteng accounts for 260 or 68.2%. In all the other provinces the greater majority of attacks took place on farms. In fact, the average percentage for attacks on smallholdings in the other eight provinces is only 16.5%.

It proved to be very difficult to get accurate information on attacks on smallholdings for the years prior to 2001. The Committee itself analysed the entire NOCOC database of over 3500 cases, which theoretically should indicate whether the property attacked was a farm or a smallholding. The results are shown in Table 10.



In terms of the NOCOC statistics the proportion of attacks on smallholdings increased from 29.3% in 1998 to 38.6% in 2001. Furthermore, in real terms the number of attacks on farms remained virtually constant over the same period, at 590 and 591 respectively, whereas attacks on smallholdings increased from 245 to 371, i.e. by 51.4%.

After hovering around 30% from 1998 to 2000, the sudden increase of attacks on smallholdings in 2001 is inexplicable, and the Committee is inclined to think that it was at least in part the result of a conscious effort on the part of the NOCOC to keep more accurate statistics of attacks on smallholdings, as was the case with the CIAC. For this reason the proportion of attacks on smallholdings in the CIAC statistics (37.7%) correlates very well with the NOCOC figure of 38.6%.

There is another indication that the earlier NOCOC figures for attacks on smallholdings may be too low: In 1998 Assistant Commissioner Britz and Director E. Seyisi undertook a study of attacks on farms and smallholdings which had taken place between January and May that year. They found that 34.8% of the 305 attacks took place on smallholdings, which is substantially more than the NOCOC percentage of 29.3%.

In all probability, therefore, the proportion of attacks on smallholdings may have been substantially higher than indicated, especially for the years prior to 2001. The Committee is satisfied, however, that attacks on smallholdings did increase over the period from 1998 to 2001, both in real terms and as a proportion of all attacks on farms and smallholdings.

As far as murders during attacks on smallholdings is concerned, Britz and Seyisi found in their study that of the 53 persons killed in attacks on farms and smallholdings during the first five months of 1998, 22 or 41.5% had been killed on smallholdings. Since only 35% of the attacks had taken place on smallholdings, the inference may be drawn that attacks on smallholdings are more likely to result in murder than attacks on farms. It must be pointed out, however, that the study only covered five months and therefore did not take into account possible seasonal fluctuations or how the ratio could have changed later in the year.

It will be noticed that Gauteng, the province with by far the greatest proportion of attacks on smallholdings (93.5%), and Limpopo (32.7%) both have a below average murder rate per incident. One can therefore say with a fair degree of certainty that the murder rate during attacks on smallholdings is no higher than during attacks on farms but, if anything, a little lower.



THE NATURE OF FARM ATTACKS

Farm attacks may take many forms and can be executed in various ways. (Again, unless otherwise indicated, the term ‘farm attack’ is used in respect of attacks on both farms and smallholdings.) It can happen on any day of the week and at any time of the day or night. It can take place inside the house or outside. Many different crimes can be committed during the course of a farm attack. All kinds of weapons can be used in the attack, and anything of value can be stolen. Some attacks leave the victims unharmed, in other cases they are killed or seriously injured. Some attackers have a clear motive, others not. These and other features are discussed below.


Monthly, daily and hourly distribution of farm attacks

It would seem that the occurrence of farm attacks are distributed fairly evenly throughout the year (see Table 11), possibly with a slight decrease during January and December. (The reason for this strange phenomenon is not known.) An analysis by the Committee of the NOCOC database for 998 to 2001 shows the same trend. This phenomenon is difficult to explain. It does not mean, however, that the vulnerability of potential victims is diminished significantly during those two months.



Similarly the distribution of farm attacks during the days of the week is fairly even, with possibly a slight increase on Fridays. The reason for this is probably that wages are usually paid on Fridays, leading prospective attackers to believe that there are large amounts of money on the farm. Also, traditionally many farmers go to town on Fridays to pick up the children and provisions, so that it is easier for attackers to enter the farm or farmstead unobserved. The figures for 2000 and 2001 are set out in Table 12.



The CIAC has also analysed the time of attacks during 2000 and 2001 to try and establish which times of the day farm attacks were most like to occur. Unfortunately the time slots used for each year differ, so they cannot be combined in the same table.) The times for 839 attacks during 2000 were registered. (See Table 13.)



One would have liked to know how many attacks took place during the hours of daylight and how many at night. For 2001 the times of some 940 attacks were registered and the time slots used by the CIAC are more useful, being periods of three hours each. It made the calculations in Table 14 possible.



It is obvious that farm attacks can be expected any time of the day. Surprisingly few attacks are carried out during the small hours between midnight and daybreak, while the most dangerous periods are, firstly, the evenings and, secondly, the mornings.


Place where farm attack occurs

The analysis done by the CIAC indicates that about half of all farm attacks take place inside the safety of the home. (See Table 15.)This may happen in the evening while the victims are still awake and the attackers suddenly or stealthily enter the house. Sometimes the victims are overpowered while they are in bed. Often the husband will get up to investigate a noise in the house and is then overpowered in the passage or elsewhere in the house. It is also quite alarming to see how many victims are overpowered in broad daylight by attackers who enter freely through open doors and windows. That not only includes the farmer or his wife, but very often domestic workers who are overpowered in the absence of the employers. Lastly, sometimes the attackers would enter the house while the occupants are away, and then ambush them on their return. Although this does not happen frequently, this type of farm attack is often very brutal, resulting in death or serious injury.



About a third (29.2%) of the victims are attacked outside but in the immediate vicinity of the house. Typically they would be overpowered and then forced into the house where the robbery would be completed. This may happen while the farmer is working outside, or even when he is returning home from somewhere else. Quite often the victim is inside the house, when he hears a noise outside and goes to investigate, only to be overpowered and forced back into the house. Sometimes the attackers wait for the farmer to come outside on his daily routine, such as switching off a water pump or putting a vehicle away.

Some attacks take place elsewhere on the farm. In KwaZulu-Natal, for example, it has happened on a number of occasions that the farmer was riding around in his truck and is then ambushed in the sugar or timber plantations. These attacks are frequently fatal. The reason is probably that the farmer cannot be stopped or immobilised without shooting and killing him. Also, the attacker needs time to make his get-away before the alarm is raised. Also included in this category are cases where a female farm resident is overpowered while walking on her own, and raped. This is in fact quite common.

In a small number of cases the farmer may be ambushed at a gate on the farm road. The objective of the attack may be to rob the farmer of his vehicle, firearm or cellphone. Typically the attackers would wait in ambush in the bushes or tall grass next to the road. Again, while this does not happen very often, these attacks are often fatal. Sometimes the farmer is followed from town after he has withdrawn money, e.g. to pay wages.

Quite a few attacks take place at the office. This usually happens on the bigger farms where large amounts of money is paid out for wages, especially on Fridays or Saturdays.

Finally, there are other forms of violence, such as fields or plantations being set alight. This happens quite frequently in KwaZulu-Natal, where large scale intimidation of farmers often take place. Other forms of intimidation include cases where farm workers are forced to give information to suspected farm attackers, e.g. concerning the movement of people on the farm or where the money is being held.


Specific offences committed during farm attacks

There is no such crime as a farm attack, but it can constitute one or more of several crimes. It usually takes the form of robbery or housebreaking with intent to rob and robbery, usually with aggravating circumstances. (Robbery with aggravating circumstances is the legal term for armed robbery.) Sometimes the robbery is accompanied by other crimes such as murder, rape or assault. Sometimes murder or rape occurs on its own.

The CIAC made an analysis of the various crimes committed during farm attacks in 2001. Expressed as a proportion of the total number of farm attacks carried out during 2001, the frequency of the most prevalent crimes committed in the course of the attacks, as registered by the SAPS, is expressed in Table 16.



Unfortunately, these figures are not reliable, for three reasons. Firstly, although nominally all crimes committed during a farm attack are supposed to be registered in the computerised Crime Administration System (CAS), this does not happen in practice, and often only the most serious crimes are registered. Where a victim is killed during a robbery attack, for example, only the crime of murder may be registered and not the armed robbery. The same applies to rape. It is therefore almost certain that the proportion of attacks involving robbery is much higher than indicated in the table. What can be said, however, is that the figure for murder is probably quite accurate. The rapes reported will also be registered fairly accurately.

Secondly, in law, assault (whether common or serious), and sometimes even attempted murder, forms part of the crime of robbery. This means that even if a person is assaulted very seriously during the course of a robbery attack, the assault will not be registered as a separate crime. For this reason it should not be assumed that assaults took place in only 20.6% of the cases - the percentage is likely to be much higher.

Thirdly, although there can be more than one victim during a robbery attack there can still only be one charge of robbery. One can therefore say that, in terms of crimes actually registered, robbery was committed in 68.2% of farm attacks. However, one cannot say that, in the same way, attempted murder was committed in 24.2% of the cases, assault in 20.6%, etc. The reason is that more than one charge of murder, assault, etc., may be laid for any single farm attack, depending on the number of victims. To calculate the frequency of those crimes, therefore, one has to take into account the number of victims.

According to the CIAC there were a total of some 1398 victims in the 1011 farm attacks during 2001. The number of crimes as a proportion of victims is given in Table 17.



It can therefore be said that in 2001 a victim of a farm attack had a 17.5% chance of having an attempt made on his or her life, and a 10.5% chance of actually being killed. However, the chances of being assaulted were much greater than 14.9%, since many, if not most, robberies are accompanied by physical assaults in any case, which will not be registered as a separate offence. The chances of a female victim being raped were also greater than 5.0%, because that percentage was calculated on male victims as well. There were in fact 571 female victims, which means that a female victim had a 12.3% chance of being raped during a farm attack in 2001.


Victims killed or seriously injured during farm attacks

Apart from victims killed, the CIAC has also determined the number of victims seriously injured in farm attacks in the various provinces during 2001. (See Table 18.) For purposes of comparison the statistics for murders are also provided.) From the figures it appears that a victim of a farm attack had a 10.5% chance of being killed and a 34.6% chance of being injured. It is interesting to note that the two provinces with the most victims - Gauteng and Mpumalanga - also tended to have the fewest killings, with the exception of Limpopo. The highest fatality rate was in the Northern Cape, but the sample was too small to draw valid conclusions from.



Gauteng had a murder rate of 8.5%, compared to the national average of 10.5%. The province had the smallest percentage of injuries of all (29.8%), whereas the national average was 34.6%. Bearing in mind that 93.5% of the attacks in Gauteng occurred on smallholdings, and that the province accounted for 68.2% of all attacks on smallholdings (see Table 9), these figures would seem to indicate that attacks on smallholdings generally are less violent than attacks on farms.


Victims raped

It will be seen from Table 16 that according to the CIAC 70 women were raped during farm attacks in 2001, making up 5% of all the victims or 12.3% of the female victims. The NOCOC database indicates that only 31 of the victims were raped during 2001. (See Table 19.) One therefore has to conclude that the NOCOC figures are not very accurate in this regard. The reason is most probably that many female victims may be hesitant to tell farmers, members of the commando or policemen who have come to their assistance that they have been raped, and may only reveal it at a later stage.




As a percentage of the total farm attacks registered by the NOCOC, rapes made up 1.9% in 1998, 3.9% in 1999, 4.8% in 2000 and 3.2% in 2001. The unreliability of the NOCOC figures for rape is further emphasised by these wild fluctuations, and the Committee is of the opinion that no valid conclusions can be drawn from them.

What is clear, however, is that a large proportion of rape victims are black. The Committee is not convinced that the increase in rape on black women has been as dramatic as the figures would suggest. It may to a large degree be the result of better reporting of such cases. A frequent scenario is of a black woman walking to the farm shop on the neighbouring farm, and on the way she is be overpowered by a stranger and raped. Because a stranger is involved, the rape would be registered as a farm attack rather than a social fabric crime. A few years ago, however, that might not have been regarded as a farm attack.


Weapons used by attackers

Of the 1011 farm attacks during 2001 the main weapon used by the attackers is known in 909 cases registered by the CIAC. (See Table 20.) The weapons most commonly used were firearms (63.8%). It should be noted that the table refers to the primary or main weapon only - there might have been other weapons as well, such as a combination of a firearm and a knife.



It is interesting that in 11.7% of the cases the attacker arrived unarmed, although he might have armed himself during the course of the attack. In those cases, also, it would be reasonable to assume that the intention was to steal rather than to rob. (In law, for the offence of robbery there must be some form or violence or threat of violence towards the victim.)


Property robbed in farm attacks

The CIAC found that money had been robbed in 31.2% of the farm attacks in 2001, firearms in 23% and vehicles in 16%. (See Table 21.) In total, the CIAC recorded that 466 firearms had been robbed in the 233 separate farm attacks in 2001, while 170 vehicles had been robbed in 162 attacks.



Obviously these items could have been stolen on their own or in combination with other items. In other instances something else was mostly stolen or, in some cases nothing at all was taken. Unfortunately the CIAC report does not have the number of cases where items other than money, firearms and vehicles were stolen. The NOCOC database does have some information to allow the Committee to make calculations and, although the NOCOC database is not entirely reliable, it is interesting to compare the figures. In the 962 cases for 2001 on the database money was stolen on 296 occasions (30.1%), firearms on 227 occasions (23.6%), vehicles on 155 occasions (16.1%), while other items, such as cellphones and televisions sets, were stolen during 427 attacks (44.4%).


Motives for farm attacks

The Committee analysed the 3544 cases on the NOCOC database in an attempt to determine the motives for the farm attacks. Obviously, more than one factor may motivate a farm attack or cause an attacker to select a specific farm. The Committee therefore tried to identify the cases where the apparent primary or main motive was either robbery, intimidation, political, racial or labour related. A clear motive was apparent in 2644 cases.



In the other cases the motive could not be established from the database alone because nothing was apparently stolen, or there was another motive such as rape. In Table 22 these other cases are ignored for the present purposes.

• It will be seen that in 2361 (89.3%) of the cases, the motive was clearly robbery. As explained above, this figure is likely to be substantially higher, since the NOCOC database does not accurately reflect the items stolen.
  
• There were 188 cases (7.1%) of intimidation, for example where crops or buildings were set alight or where people were shot at without reason.
  
• In 52 (2.0%) cases some political or racial motive was apparent. To determine political or racial motives the Committee relied largely upon utterances made on the crime scene, the remarks of the compilers of the statistics or on some other indication. One has to assume that such overt indications of a political or racial motive will not always be present and that the figure may be higher than the 2.0% indicated.
  
• Forty three cases (1.6%) were labour related, such as disputes over wages.

The cases where apparently nothing was stolen, are discussed in the section below.

One must assume, however, that there may be more than one motive for a particular farm attack. Some perpetrators, for example, who received amnesty for their attacks from the Truth and Reconciliation Commission, had also robbed a large amount of goods. On the other hand, however, one must also assume that there may be some attackers who attempt to justify the attack afterwards by trying to give it a political colour. Furthermore, some farm attackers try to throw the police off the scent by leaving political or racial clues.


Farm attacks where nothing is stolen

Farm Murders in South Africa - SkyNews Min. of Safety and Security: Charles Ngakula: 'If you don't like the crime, leave the country'

There seems to be a perception that it happens frequently that a farmer or his family are attacked but nothing is taken. That is often seen as proof that the motive for farm attacks is not robbery but something more sinister. Often these cases receive much publicity, reinforcing the belief that it is a common phenomenon. The Committee’s investigations show, however, that there are relatively few farm attacks where nothing is robbed, without there being a logical explanation for it.

The belief that nothing has been taken in a particular farm attack is often based on suppositions and incorrect information. A case in point is the Scheepers case, which received much publicity at the time and which was referred to the Committee by the Transvaal Agricultural Union for special investigation. Gerhard and Nellie Scheepers were brutally murdered on their farm in the Ermelo district on 2 December 1997. The TAU was under the impression that nothing had been stolen, whereas in fact cupboards in the house had been emptied and the safe opened and several handguns and other weapons taken. The attackers had also stolen Mrs Scheepers’s motorcar, which had in fact led to their being traced by the police.

According to the NOCOC database nothing was stolen, or it is unknown whether anything was stolen, in almost one quarter of all the farm attack cases for the period 1998 to 2001. The Committee found, however, that in the greater majority of those cases there was some logical explanation for this, e.g. the attacker was fought off, or help arrived before the robber could get away with the booty. The Committee nevertheless identified about 86 cases where there was no obvious reason on record why nothing had been taken. It was impossible to fully investigate all of those cases, but the Committee contacted 36 investigating officers of some of the cases. It was found in almost all those cases either that various items had in fact been stolen, or that there was some logical explanation for the fact that nothing had been taken.

These NOCOC cases include several also referred to the Committee by the TAU. Mrs Marais was ambushed at the gate on the farm Newlands, Warmbaths, on 16 March 1998 and shot dead. Apparently nothing was stolen, but it turned out later that her handbag with its contents had been taken. Mr and Mrs Ronaldson were attacked on the farm Sunrise Close, Pietermaritzburg, on 17 March 1998. The husband was killed and his wife injured. The NOCOC report states that nothing was stolen, whereas in fact several items were stolen. Mr Frauenstein was murdered on the farm Sunnygrove, Kidds Beach, 5 September 1998 when he surprised burglars in his house. The burglars ran away, possibly because there were farm workers on the scene. Nothing was in fact stolen, but the attackers had already packed a television set and other items for removal. Mr Bouwer, who lived alone on the farm Uitkyk, Kareedouw, was murdered on 10 September 1998. According to the NOCOC database nothing was stolen, but in fact the bedroom as well as a storeroom was ransacked. There are other examples as well.

The situation is therefore that there are far fewer cases where nothing has been stolen than is indicated on the NOCOC database. The reason for this inaccuracy is probably that the information is passed on to the NOCOC at a very early stage, when it may not be known yet what items have been stolen. This can happen easily, especially where the victim has been killed and it is only established later on that certain items in fact have been robbed. It is easy enough to establish that items such as motor vehicles are missing and, as seen above, the NOCOC figures in this regard are fairly accurate. The loss of other items, such as cellphones or money, may not be so apparent. The NOCOC information is often discussed at the monthly and weekly meetings to which organized agriculture also has access, and the Committee surmises that that may be the source of some of the misconceptions.

The Committee is therefore of the opinion that it is a fallacy that farm attacks are often carried out without anything being stolen or without there being a logical explanation for the fact that nothing is taken. This does not mean that there are no such cases at all. As will be seen below, there are attacks which can be said to be part of a process of intimidation, or politically or racially motivated, or which are labour related. The Committee is merely saying that it is a small minority of cases. There are also some cases where it is impossible to establish any motive at all.


Involvement of farm workers in the attacks

An analysis of the NOCOC database indicated that the involvement of farm workers in farm attacks is very limited. (See Table 23.) It will be seen that the percentages varied from 3.8% in 1998 to 2.0?% in 2001, with an average of 2.8%. Even though the NOCOC figures in general are not completely reliable, these figures can be regarded as reasonably accurate. The reason is that guilty farm workers usually disappear immediately after the attack, marking them as obvious suspects.



Even if it is accepted that this figure may be somewhat higher, it is clear that farm workers do not constitute a significant proportion of farm attackers. This really puts paid to the argument raised that many farm attacks are carried out by disgruntled workers.



THE VICTIMS OF FARM ATTACKS


Race of the victims

Because of the racial connotation being given to farm attacks by certain individuals and organisations, the CIAC started making breakdowns of figures along racial lines from 2001 onwards. (See Table 24.)



In 2001, white people made up 61.6% of the victims, black people 33.3%, Asians 4.4% and coloured 0.7%. Although the CIAC does not have accurate figures for previous years, it notes in its 2001 report that the black people are increasingly being victimised. Unfortunately they do not have any figures for the previous years.

The Committee also analysed the NOCOC database to determine the racial composition of the victims of farm attacks. Prior to 2002, the NOCOC database did not specifically state the race of the victims and the Committee had to rely on various indicators, including the name of the victim. In the South African context, however, this is an accurate way in which to determine the race of a person. In many instances the race of the victims could not be determined, but the race of some 3306 victims could be determined fairly accurately. This number is more than big enough to be a fairly reliable sample. (See Table 25.)



The figures would seem to indicate that the proportion of white victims is in fact becoming smaller, although there is a possibility that the apparent increase may partly be due to the fact that there is a greater coverage of attacks on other people.

The Committee also analysed the NOCOC database in an attempt to determine the race of victims of rape committed during the course of farm attacks. The result in respect of those cases where the race could be established is given in Table 26.



Because of the small sample, it is difficult to draw reliable conclusions. Furthermore, it is generally accepted that rape is one of the most under-reported crimes. It does seem, however, as if the proportion of recorded rapes amongst white females is decreasing, while that amongst the other racial groups is increasing. Again, this may partly be due to increased coverage of such attacks on black people.


Age and gender of victims

The CIAC analysed the age of some 733 victims of farm attacks during 2001 in great detail, with a break down for each province. (The ages of the other victims are not known.) It will serve no purpose to enclose all that information in this report, and the information in Table 27 will suffice:



In terms of these statistics, 28.4% of all victims are 60 years of age or older. The CIAC remarks, however, that the average age of the victims seems to be lower than in previous years.



During 2001 some 59.2% of the victims of farm attacks were male. (See Table 28.) This predominance was evident everywhere except in the Free State, where there were slightly more female victims than males. (The sample is probably too small to draw a reliable conclusion from.)


Status of the victims of farm attacks

During 2001 some 974 (69.7%) of the victims were either the owners, lessees or managers of the farms or smallholdings, and their dependants. Some 412 or 29.5% were
employees or their families, and 12 or 0.9% were visitors. (See Table 29.)