CHAPTER 7

OVERVIEW OF SELECTED CRIME TYPES

This chapter looks in detail at the crime types most likely to have the greatest impact on the South African public, as determined by the seriousness of the crime and its prevalence in the survey. Thus while murder, for example, is a serious crime, it is also one that, according to the official statistics, occurs only about 22,000 times each year, compared to some 500,000 recorded assaults, 300,000 burglaries, and 225,000 robberies.33 In keeping with this relative infrequency, the number of murders documented in the survey was too small to allow for detailed analysis.

The following crime types are discussed below: corruption, housebreaking (residential burglary), assault and sexual assault, robbery and hijacking, and stock theft. Because corruption was treated differently to the other crimes covered in the survey, the discussion below follows a different format to that used for the other offences. In terms of this chapter’s format, key points and implications are included for each crime type, rather than at the start and end of the chapter.

Corruption

Public perceptions about corruption

South Africans were most likely to think that corruption in the public sector has increased between 2000 and 2003: 40% said the number of requests for bribes has increased over this time period, while 30% thought levels have stayed the same, and only 17% believed corruption has decreased.

These views differed significantly among race groups, with Indians much more likely to think that corruption has increased: 74% held this opinion compared to 58% of whites, 37% of blacks and 30% of coloureds. This trend is consistent with the pattern highlighted throughout the survey, in which Indian, followed by white South Africans were much more concerned about crime and safety, and far more critical of the performance of the police and courts than blacks and coloureds.

Opinions about corruption varied marginally between people living in urban are rural areas: those in the metropolitan parts of the country were slightly more inclined to believe that corruption had increased (46%), than those in urban (42%), farming (37%) and traditional rural (33%) areas. A similar trend was found among the provinces (Table 26, Appendix 2.)

These views are no doubt informed by actual experience as well as perception. Media reports play an important role in shaping public opinion, and the media has, correctly, given corruption scandals extensive coverage. However, given that most of the survey respondents were poor, the perception that corruption—particularly pertaining to delivery of basic services—has increased, should be a matter of concern for those committed to ensuring clean government and the public service principal of ‘batho pele’ or ‘people first’.

Extent of corruption

It should be noted from the outset that many incidents of corruption or attempted corruption were probably not reported to the survey. Some victims may be aware of their perceived ‘complicity’ as the bribe payer (albeit on demand, possibly accompanied with the threat of withholding a service) and would worry that by answering the questions they might implicate themselves. Others may not be aware that being asked for a bribe in return for a service is a crime, and may instead see this as a ‘normal’ transaction fee required to ensure the delivery of services. Similarly, requests for ‘favours’ and ‘gifts’ may be overlooked as a form of corruption. These acts typically involve the purchase of a cold drink, alcohol or a meal in return for speeding up a service or the efficient delivery of a service. Some members of the public may regard this as simply an accepted form of gratitude.

Considering that corruption is one of the crime types that is likely to be under-recorded by victim surveys, the fact that it was the second most prevalent crime in the country is a matter of concern. Respondents were asked if, “over the past year, any government official asked or indicated that they would be receptive to either money, a favour or a present in return for a service that they were legally required to perform.” If all three variants of corruption—money, favours and gifts—are added together, corruption was the second most frequent crime type recorded by the survey with 5.6% of South Africans reporting experiences of corruption. When only cash bribes are counted, corruption ranks third. In total 4.6% of the sample reported corruption involving money, while 0.6% described an incident involving a favour, and 0.4% a present (such as purchasing a cold drink or meal for an official).

A comparison between the 1998 and 2003 national victim surveys suggests that the rate of corruption has almost tripled from 2% to 5.6%. Three factors could explain this dramatic increase:

• The increase over the past five years reflects a rise in petty corruption particularly at the site of service delivery, i.e. local and provincial government. Corruption at this level reflects a legacy of bad governance inherited from the past which if unchecked could develop into endemic corruption. The Eastern Cape is a good example of this, where the provincial leadership witnessed an unprecedented intervention by national government in late 2002 in an attempt to stop rampant corruption and maladministration.
  
  

• Media attention on cases of alleged grand corruption such as the arms deal has raised public awareness about what constitutes an act of corruption. This, combined with a greater awareness among citizens of their rights to fair administrative action, may have resulted in respondents reporting corrupt behaviour in 2003 that might have been regarded as ‘normal’ practice in 1998.
  
  

• Differences in methodology between the surveys could have influenced the results. The 1998 survey asked respondents if, “…any government official, for instance a customs official, police officer or inspector asked you or wanted you to pay a bribe for his/her service?” In contrast the 2003 survey asked respondents if “…any government or public official asked you or indicated to you that they would be receptive to the following: money, a favour, a present in return for a service (that the official is legally required to perform)?” In 2003, the description of a bribe was more comprehensive, including not only monetary forms of corruption but also favours and presents—of which there were many cases reported to the survey. In order to track corruption trends over time, future surveys should disaggregate corruption into these three forms to ensure comparability and a more accurate representation of the problem.

Nature of corruption

Bribes demanded

Corruption was most evident in encounters with traffic officials, followed by the police, and then during interactions with officials over employment opportunities (Figure 44). This supports the perception that corruption is a major problem both in local government traffic departments and the SAPS, and highlights the discretionary power of some of these officials. For example, the public largely interact with traffic officials on the road where the actions of corrupt officials are difficult to monitor. The frequency of demands for bribes by members of the SAPS is worrying given their role in fighting crime and corruption. These results underscore calls for a designated body or unit to monitor and investigate corruption within the ranks of the SAPS.38

The high number of requests for bribes in the course of inquiries about employment in the public service reflects the massive levels of unemployment in the country. When demand outstrips supply to such an extent, unscrupulous officials are aware that they can use this situation to their own advantage.

Corruption was next most common among officials responsible for paying pensions or social welfare grants. These are a major—and often the only—source of income for many impoverished households. The Minister of Social Development, Zola Skweyiya, has pointed out that approximately R15 billion earmarked for pensions, social grants and other forms of poverty alleviation has been ‘lost’ to corruption between 1994 and 2004.39 Reports of corruption were as common during applications for identity documents, a responsibility of the Department of Home Affairs, as for pensions and grants.

The public was next most likely to encounter bribery during applications for driver’s licenses, which again involves traffic departments, and by implication local government. Considering that traffic officials were implicated in the most common type of bribery recorded by the survey—encounters with traffic officials over fines—as well as during the issuing of driver’s licenses, the overall poor performance of local government is a major cause for concern.

Bribes paid

Those respondents who said an official requested a bribe, were asked whether or not they paid it. The most commonly paid bribe was for traffic fines, with an astounding 100% of respondents indicating that they had indeed paid the bribe (Figure 45).

An important explanatory factor is that bribes are often demanded in situations where road users have committed an offence such as speeding, overloading, or driving unlicensed or unroadworthy vehicles. Bribery in these instances may be used to ensure that the offender escapes a stiffer penalty (i.e. a R100 bribe is requested when the alternative is to pay a legitimate fine of double that amount). Unscrupulous officials may also however prey on road users who have not committed an offence. Anecdotal evidence suggests that nationals of neighbouring countries may be requested to pay a bribe or face the possibility of not passing through a border post ‘on time’ if they are not able to pay the much stiffer fine. Equally some South Africans may be willing to pay these bribes to avoid the ‘hassle’ factor of having to prove their innocence, particularly if they are passing through a province in which they are not resident.

The results clearly indicate a propensity among those surveyed to regard bribe payment to traffic officials as an innocuous exercise. The effect of such behaviour is however not only a loss of state revenue. When corrupt officials allow motorists who speed, or who are driving vehicles that are not roadworthy to proceed with their journey, the consequences for other road users are potentially disastrous. A breakdown in public trust of the integrity of traffic officials is also likely to result in an increase in lawlessness among road users.

After traffic fines, other services for which bribes were often paid were utilities (water or electricity) and telephone installation. These could well be illegal connections or illegal reconnections after disconnection, highlighting the discrepancy between the availability of such services and the ability of many people to pay for them. The fact that many who have been disconnected, resort to ‘illegal’ connections to ensure access to basic services underscores the fact that bribery is possibly seen as a means to facilitate access to public utilities. The privatisation of these services is unlikely to lessen this practice, given that a real need exists among poverty stricken households to have sustained basic services (water, electricity, telephony) and not merely a ‘connection’ which users are unable to afford given competing livelihood needs.

It is noteworthy that although policing was the second most likely sphere in which South Africans were asked for a bribe (Figure 44), none of the respondents admitted to paying the bribe. Although this may reflect integrity on the part of the respondents, it is also possible that they were reluctant to admit to bribing a police official for fear of repercussions. Bribery of an official in the criminal justice sector may also be viewed as more serious than that of a traffic official.

Private sector corruption

Only two questions were put to respondents regarding private sector corruption. This is because bribes are usually not requested when members of the public have a choice in procuring goods and services. Although corruption is a massive problem in the private sector (referred to as white collar crime) and is often used to gain unfair advantage in securing government contracts, most of these crimes involve the ‘elite’ rather than ordinary citizens. Quantifying this problem in a survey of a representative sample of South Africans would therefore be difficult.

Nevertheless, 4.3% of respondents said that either they, or a family member, had ever been asked for a bribe in return for speeding up a job application in the private sector, while 5% reported that they, or a family member, had ever been requested for a bribe in return for getting a job in a private company. While these figures seem high, it is important to note that these experiences were not restricted to the past year, as was the case in the questions on public sector corruption discussed above. The results again suggest the high demand for employment in South Africa and the accompanying opportunities this creates for individuals who are willing to partake in corrupt transactions. Although there is no way of knowing how many of these respondents actually obtained the jobs in question after paying the bribe, the results dispel the myth that in corporate South Africa it is always the ‘best person for the job’.

Reporting corruption to the authorities

All respondents, rather than just those who had experienced corruption, were asked whether they had ever tried to report a corrupt official. Very few (2%) said that they had. When asked who they reported the incident to, the most likely answer was “another official” (42%), presumably from the same department and who is possibly a superior to the official who asked for the bribe. Almost as many respondents said they reported to the local police station (40%), while significantly less used a telephone hotline (17%) such as a whistleblower hotline.

The primary reason given by the 98% of respondents who did not report bribery was that it would not have changed anything (Figure 46). Despite good whistleblower provisions (South Africa is one of only seven countries with legislation protecting whistleblowers) as many as 27% said they are afraid of reprisals. This is a major deterrent to reporting corruption as these whistleblowers are essentially victimised twice—both by the act of corruption as well as the potential threat of reprisal in the event of reporting it. A significant proportion of respondents said they did not know who to report the incident to. These figures reflect the relatively low-key approach towards promoting public awareness of the need for individuals to fight corruption and of how to report it.

Of the three main reasons for not reporting corruption, the view that it would not change anything, and the lack of knowledge about where and how to report (which together represent two thirds of responses) could be addressed in the short term by a sustained awareness campaign by the public service. This should be geared towards informing citizens of why they need to counter corruption, thus popularising a whistleblowing culture, as well as ensuring that people know what channels to use to report (i.e. through national hotlines). In the long term, the public needs to believe that their actions will result in speedy investigations and when appropriate, prosecution and conviction. One way of doing this is to inform citizens of convictions achieved as a result of information provided by whistleblowers.

The more complex issue is dealing with the fear of reprisal. This can only be tackled by ensuring that citizens feel adequately protected by the provisions of the Protected Disclosures Act (the ‘whistleblower’ Act) and that the relevant authority assists in ensuring that they remain free of intimidation in the event of disclosure. The SA Law Commission is reviewing ways to improve this key piece of anti-corruption legislation which should hopefully see whistleblowers, particularly those in the workplace, making better use of this instrument. Whistleblowers, as ordinary citizens, are key to effectively challenging both grand and petty corruption.

Implications of the survey results

• Although endemic corruption is not a problem in South Africa, the high level of petty corruption in certain sectors is a cause for concern, not least because if unchecked, it can become endemic. The results indicate that some traffic departments are probably vulnerable to this, and local governments need to tackle petty corruption jointly with national and provincial Departments of Transport.
  
  

• The continued high level of unemployment means that demand for jobs will far outstrip supply. Other that the urgent need to address the scarcity of employment opportunities, public and private sector employers must be seen to be doing enough to keep corruption and nepotism in check.
  
  

• The results suggest that South Africa ’s poor are especially vulnerable to petty corruption. This is worrying because their inability to access basic services due to the demand for bribes will further deepen socio-economic cleavages and contribute to their alienation from the democratic process.
  
  

• A dedicated public education campaign is needed to start to address the lack of awareness about what constitutes corruption and how to report it. Consensus among the elite will not be enough to stop corrupt practices. Ultimately broad public participation is required to promote a culture of whistleblowing in the public and private sectors, but also to ensure that corruption is prevented in future.

Housebreaking

The definition used in the survey corresponds well to the crime of residential burglary as recorded in the official crime statistics, but excludes commercial burglary. Because of common confusion around the definition, this survey, like others, captured a small number of ‘home robberies’ under the heading of housebreaking, as will be discussed further below.

Public perceptions of housebreaking

According to both the 1998 and 2003 surveys, housebreaking is the most commonly experienced crime in South Africa, with 7.5% of the respondents in 2003 saying their household had been burgled in the last year, compared to 7.2% in 1998. This would suggest that on a suburban block of 14 houses, chances are that one would be burgled every year. Not surprisingly then, the survey indicates that housebreaking is also the most commonly discussed crime, with 24% of respondents saying they talked about housebreaking with friends, family or colleagues over the previous two weeks. It was also perceived as the “most common crime in my area” by 38% of respondents—a higher percentage than that accorded to any other crime type.

Perhaps more unexpected ly, nearly a quarter of those interviewed said it was the crime they feared the most—only murder was mentioned more often (see chapter four). This suggests that the South African public has an accurate picture of the real risk of becoming a victim of housebreaking, and that the threat is feared. Indian respondents in particular were concerned about housebreaking, with 31% of those interviewed identifying it as their most feared crime. This is important because neither government’s National Crime Prevention Strategy nor the SAPS’ National Crime Combating Strategy have prioritised housebreaking, despite prioritising just about everything else. Typically, it is regarded as a mere property crime, without a real appreciation of the impact that having a home invaded can have on feelings of safety, or the threat that loss of property can have on the livelihoods of poor families.

It is a shame that more attention is not given to this crime, because the stereotypical suburban housebreaking can be prevented by traditional policing and security techniques (see ‘implications’ section below). From the outset, target hardening devices—including everything from lighting to alarms to precautions taken when on holiday—are an effective deterrent to all but the most committed or well-connected burglars. Unlike robbery, housebreaking takes place in a fixed location, and so burglars are constrained in their choices. Many burglaries are planned for a period of time, and unexpected police or security presence at the chosen location can foil the attempt. Entry, gathering of property, and exit typically take time if they are to be done stealthily, during which the burglar is vulnerable to surprise patrols. Opportunistic burglaries, which often involve forcible and noisy entries, can call unwanted attention from residents or their neighbours, including calls to the police. Finally, well-marked or otherwise identifiable property can be traced back to the burglars when located in area searches or by other techniques.

Many South African burglaries do not adhere to this suburban model, however. Burglaries in shack areas, for example, cannot easily be prevented by target hardening or patrols, and, when residents are at home at the time of the offence, given the small area of these dwellings, may involve an explicit or implied element of violence.40

Burglaries in farming or traditional areas are similarly not as amenable to target hardening or patrols as those in more densely settled areas. Even in urban high-rise buildings, there is little need for stealth when the threat of retribution and lack of community cohesion allow some to operate with impunity. Housebreaking is a crime that underscores our vulnerability, even in our own homes. This time, property may be the target, but next time, something more precious might be taken. After all, burglars necessarily know where we live.

Extent and risk of housebreaking

With 11,205,705 households in South Africa, a 7.5% annual victimisation rate for housebreaking with 57% of victims reporting to the police should result in 479,044 records of residential burglary in the police’s database. In fact, in 2002/3 the police recorded 393,159 incidents, or 18% less than expected based on the victim survey results.

This could be linked to under-recording by the police, because the fact that victims reported does not necessarily mean the incident was recorded by police (see text box), and no questions were asked in the survey about whether a case number was received by the respondent (which would indicate that the incident was actually recorded).

In terms of who is most at risk of housebreaking, white (14%) and Indian (14%) households were more likely to say they were burglarised than were blacks (7%) and coloureds (4%). Housing type was not associated with victimisation, but respondents from metro (8%) and urban areas (8%) were more likely to report victimisation than those from farming (6%) and traditional rural areas (7%).

Nature of housebreaking

Most of the burglaries discussed by respondents occurred at night (70%), but were as likely to occur when the residents were at home (50%) as when they were absent (50%). Just under half of the cases (46%) involved entry through a window, while in 36% of cases, a door was simply smashed open. A small number of these ‘burglaries’ were actually home robberies (see the section on robbery below), as six incidents resulted in injury to residents. Electronic equipment was taken in 50% of the incidents in which property was taken, and cash was taken in another 14%.

Reporting to the police and case outcome

Only 57% of housebreaking victims said they reported the crime to the police. The single largest reason for not reporting was because “it wasn’t necessary or important enough” (32%), which is generally the most common reason given for not reporting all crime types (see chapter five). This lack of ‘necessity’ may be tied to the fact that only 17% of victims said they were insured against housebreaking. Of those who were insured, 98% reported, compared to 47% of uninsured victims. More disturbing are the 19% of victims who said they did not report because they did not trust the police, and the 14% who said they “used other means to resolve the situation”. An additional 15% said they felt they did not have sufficient evidence to justify reporting. Only 2% of victims—all of whom were living in traditional rural areas—did not report because they felt threatened by the perpetrators.

Of those who did report, most were unhappy with the response they received from the police (54%), and 63% were unhappy with the response from the authorities overall (includes police and courts). There are many possible reasons for this high level of dissatisfaction, including unrealistic expectations of what can be done in a housebreaking case. Influenced by media images, victims may expect the police to dust for fingerprints, for example, when this is not always possible.

Respondents were aware that arrests had been made in 11% of cases reported, which is close to the 13% referral to court rate found in the 2000 police statistics41 and compares favourably with the 14% detection rate reported in 2002/3 in England and Wales.42 Keep in mind that these incidents could have occurred any time in the previous year, including very recently, so it is possible that more arrests will be made in the future. In 12% of the cases, the stolen goods were recovered. Perhaps the public should be informed that chances of a positive outcome are a little better than one in ten.

Aftermath of housebreaking

Housebreaking victims were equally concerned with getting life back to normal (29%) and recovering the stolen property (27%), with only 9% being most concerned with the apprehension or punishment of the perpetrators. A comparison across crime types shows that getting life back to normal and recovering property are the most common concerns for many victims (see Table 13, page 142).

Just under half (49%) of housebreaking victims took measures to protect themselves against repeat victimisation, the majority of which (57%) involved ‘target hardening’ security hardware installed in the home. In three quarters of these cases (66%), these measures did indeed make the victim feel more secure.

Implications of the survey results

• Given valid public concerns about housebreaking and high levels of fear, government and police policies need to target this crime type. Neither the National Crime Prevention Strategy nor the National Crime Combating Strategy prioritised housebreaking.
  
  

• This is one crime type for which the public should be encouraged to take greater responsibility for personal safety, since target hardening could be an effective deterrent; enhanced security on windows and reinforced doors would have been helpful in 82% of cases reviewed.
  
  

• Since police performance in making arrests compares favourably with international standards, high levels of public dissatisfaction suggest that the public needs to be educated about the reasonable prospects of success in housebreaking cases: your chances of seeing an arrest or recovering your property after a housebreaking are slightly better than one in ten.

Assault and sexual assault

The crimes of assault and sexual assault are extremely difficult to capture in a household survey—and are hard for the police to handle—for many of the same reasons. Some degree of physical beating is felt by many to be acceptable in some relationships, such as parent-child, elder-youth, and even husband-wife relationships. Similarly, in some communities consensual brawling in various contexts is considered normal, especially between male peers. Certain forms of verbal provocation or other slights are felt to justify physical retaliation by segments of the public. While the law has recently been updated to allow for spousal rape, this concept still escapes many South Africans. As a result, there is a disharmony between community practice and the letter of the law. Exactly the same physical attack could be deemed a criminal assault in one community and not in another, and opinions on the matter can vary across time, even for the same victim.

In addition, assault does not represent just one crime type with one motivation. It encompasses a range of circumstances, from domestic violence to bar room brawls, to attempted robberies, to hate crimes. Many, if not most, of these crimes involve people who know one another.

Assaults between intimates are unlikely to make it to the attention of the police, especially if injuries are minor and reconciliation timely. If the perpetrator is in a position of power or support over the victim, reporting may not be in the material best interests of the victim. For much the same reasons, the crimes are unlikely to be mentioned to a fieldworker conducting a door-to-door survey, especially if the location of the interview is the home, perhaps within earshot of neighbours or even the perpetrator himself. There is also an element of embarrassment in admitting victimisation, especially for some sexual assault survivors, who may even blame themselves for the incident.

As far as law enforcement is concerned, making arrests for assault is relatively easy, since the perpetrator is usually known to the victim. But prosecuting assault is difficult because, perhaps more than any other crime, assault is considered by the public to be an offence against an individual, not against the state. This view is supported by the fact that it can be the basis of a civil suit in addition to criminal charges. As a result, many complainants, or their families, feel they reserve the right to determine whether criminal action should proceed. This means that the perpetrator can influence whether he winds up in jail by negotiating with the victim, offering either reconciliation or compensation. This is not generally considered witness tampering, even by the police.

While the police could, in theory, mount a prosecution even without a cooperating victim, the practicalities of such an action and the case loads confronted by the criminal justice system would argue against it. This results in a high number of reported assault cases being withdrawn at the request of the complainant: as high as 61% of common assault cases and nearly 20% of rapes in 2000, for example.43 Even if authorities are able to keep the victim interested long enough to make an arrest and get the case to court, many cases are withdrawn at a later stage: in 2000, more than half of all assault cases that made it to court were withdrawn, and more than three times as many rape cases were withdrawn in court than saw a conviction.44 Since many of these cases are destined to go nowhere, the enthusiasm of the authorities for investigating assault may be less than for other crimes. Police and prosecutors may feel they are meddling in personal matters, and, in many cases, the public agrees with them.

Public perceptions of assault

The survey showed that assault is still viewed in an ambiguous light by the South African public. It is not believed to be as highly common as other types of crime: only 6% of respondents thought it was the one type of crime that occurred most in their area of residence, ranking assault in seventh place of the crimes believed to be most common. Similarly only 5% said assault was the one crime they feared most, ranking fifth after other offences. It is not surprising then that only 4% of respondents said assault was the most commonly discussed crime in the past two weeks (see chapter four).

Extent and risk of assault

The survey also suggests that public attitudes towards assault may be growing more conservative, or could be hardening. Only 2.2% of South Africans claimed to have been assaulted in the previous year, down from 4.2% in 1998. It is unlikely that this dramatic reduction can be attributed to a decline in real incidence, because the police figures show an increase during this period of time. While the rate of reporting claimed by respondents actually increased between the two surveys, from 38% in 1998 to a remarkable 55% in 2003, this cannot account for the increase in police figures if real incidence actually halved, as the surveys suggest.

Rather, it appears that the 2003 survey captured only the most serious assaults, as though lesser attacks no longer leave much of an impression on victims. Among the assaults detailed in the 2003 survey, 57% involved weapons and 73% resulted in injuries, 78% of which required medical attention, and 43% of which involved hospitalisation. These are no common assaults, but would probably have been captured by the police as assaults with intent to cause grievous bodily harm, or ‘assault GBH’.

If, in fact, 2.2% of the 30 million South Africans eligible for the survey experienced assault, and 55% of these people reported the crime to the police, the survey would project about 363,000 reported crimes in 2002/3. In fact, the SAPS recorded 548,847 assaults, including both common and GBH. But if we compare only the serious assaults captured by the survey to the figures for assault GBH alone, the numbers are startlingly close: about 265,000 assaults projected by the survey, compared to 266,321 recorded by the police. So, as is the case for robbery (see below), it would appear that the survey captured only the most serious assaults.

With regard to sexual assault and rape, the door-to-door survey technique has repeated ly shown itself to be ineffective in South Africa. The number of survey respondents willing to discuss their rape with a fieldworker at their door was very small: 17 people nationwide. This should surprise no one. While a woman may be willing to recount this most personal and traumatic of experiences to the police in the hope of stopping the rapist, she would have little incentive to undergo this trauma for the benefit of an anonymous person who shows up at her door. At this point in South African history, the topic is simply too sensitive to research in a general household survey.

Returning to assault, respondents from metro (3%) and urban (3%) areas were more likely to say they had been assaulted in the last year than those from farming (2%) and traditional rural areas (1%), but this may be linked to conservative notions about what constitutes an assault in rural areas. Coloured respondents were most likely to say they had been assaulted (3%), compared to other ethnic groups (2% each)—an observation that is supported by the fact that assault is most common in the two provinces (Western Cape and Northern Cape) where coloured people are in the majority. Of course, this may be due to greater openness to discussing the matter in this community, both to the police and surveyors, rather that greater incidence.

Nature of assault

The survey shows most of the assaults to be of an interpersonal nature. In only 11% of the cases was the assailant completely unknown, with 20% of respondents being confident that the perpetrator was from the community, though personally unknown to them. Another 30% were community members known to the victim, 17% were friends or acquaintances, and in 12% of the cases, the assailant was a spouse or lover (Figure 47).

The number of typical domestic violence cases captured by the survey was small, no doubt for all the reasons discussed above that limit reporting of such incidents. It is also possible that, because most assaults reported to the survey were very violent attacks, those domestic violence cases that involved less serious beatings were not reported to the fieldworkers. In only 14 cases (12%) was the perpetrator identified as the spouse or lover of the respondent; in 11 of these cases the victim was female, in 10 of these cases was the location of the assault identified as the home.

In 79% of spousal assaults, no weapon other than physical strength was used, but in 43% of these attacks medical attention was required. Since it is unlikely that 43% of all beatings in the home require medical attention, it seems that many respondents talked only about the most serious incidents to the surveyors. In 65% of spousal assault cases, the victim reported having been victimised by their attacker before. Perhaps surprisingly, alcohol played a lesser role in these assaults than in other assaults: in only 36% of the cases did the respondent believe the attacker to be under the influence of alcohol and in only 14% of the cases did the victim admit to drinking.

Compare this to the overall picture, in which 44% of the victims felt the assailant was under the influence of drugs or alcohol, and in which a third conceded that they themselves were under the influence of drugs or alcohol. While the number of cases is small, it is worth noting that coloured respondents were more likely than other ethnic groups to associate alcohol with the attack, and were more likely to be assaulted in a bar or entertainment area.

Since most of the crimes were interpersonal, it is not surprising that 42% occurred in the home, with only 30% on the street, and 11% in a bar or other entertainment area. It is also not surprising that 27% said they had been assaulted by this person before, including 65% of the spousal violence victims. In terms of numbers of perpetrators, 53% said the assault involved only one assailant (including all but one of the spousal violence survivors), 25% two, and 22% more than two attackers. Thus, 47% of the assaults were group attacks.

The reasons for the assaults varied, with 20% of the respondents concluding that the attack was due to long term personal anger towards the victim, 17% assuming it to be an attempted robbery, 15% to be motivated by sudden personal anger towards the victim, 13% by money disputes, 12% by jealousy or other romantic motives, and 12% by anger towards the friends or family of the victim. While the numbers are small, whites were more likely to claim that attempted robbery was the motive than other ethnic groups. All but one of the ‘attempted robberies’ involved groups of two or more.

Reporting to the police and case outcome

In a remarkable 55% of the cases, the victim reported the crime to the police. But this figure may not be as impressive at it seems at first if some of this ‘reporting’ was involuntary—the product of hospitals that treat the victims being responsible for calling the authorities. In 38% of the cases, the incident went unreported because it was deemed not important enough, which is pretty standard fare. More disturbing are the 18% of cases in which other means were used to resolve the incident, and the 17% where the crime was not reported because the police were not accessible.

Most of the victims were satisfied with the police response (57%) although just under half (47%) of those who reported were satisfied with how the authorities handled their case. A remarkable 38% of the victims who reported were aware of an arrest being made, probably immediately after the offence, and in 68% of these arrest cases, the victim was required to attend court. In 46% of the arrest cases, a conviction had been handed down. Thus, in 11% of all assault cases captured in the survey, and in 17% of the reported cases, convictions had already been handed down, despite the fact that all of these incidents had occurred in the last year. This high success rate is probably due to the fact that the majority of the victims knew their perpetrators, and the severity of the attacks provided corroborating evidence. Thus, if the complainant is willing to see out the criminal process, the prospects of success are good.

Aftermath of the assault

Over half (54%) of the respondents were chiefly concerned with getting their lives back to normal after the attack, a higher share than for any other crime type (see Table 13, page 142). For many, this is likely to have included reconciliation with the assailant. A good fifth (21%) however, wanted most to avoid repeat victimisation, including those who got a restraining order issued. A total of 15% wanted most for the perpetrators to suffer or to be taken off the street.

Over half (54%) changed their behaviour as a result of this incident. The most common modifications included avoiding certain areas (29%), being more alert (25%), and giving up going out alone altogether (20%). More than three quarters (77%) felt safer as a result of this action.

Implications of the survey results

• Assault is not one crime type, but many, each of which may require its own specialised intervention.
  
  

• While 79% of domestic violence incidents in the survey involved no weapons, 43% required medical attention. Victims’ interests must be maintained in order to ensure these assaults do not someday become murders.
  
  

• In nearly half (44%) of the cases captured, the victim felt the perpetrator was under the influence of alcohol and in a third, the victim himself was under the influence. This means that regulating alcohol availability and use could play a significant role in reducing assault.
  
  

• The greatest barrier to successfully jailing assailants is victim participation in the process—co-operative victims result in high conviction rates.

Robbery

Robbery was defined in the survey, as in the law, as the taking of property by force or threat of force. This includes a wide range of offences such as armed and unarmed mugging, bank and cash-in-transit robbery, vehicular hijacking, purse snatching and smash-and-grabs, and robberies that occur in the home and at business premises. To capture this diversity, the SAPS keeps track of several sub-sets of robbery, in addition to distinguishing between aggravated (generally, armed) robbery and common robbery. In this survey, vehicular hijackings were recorded separately from robbery.

Public perceptions of robbery

Robbery is a crime of high concern among the South African public. According to the survey, it is the second most commonly discussed crime, believed to be the second “most common crime” in respondents’ area of residence, and the fourth most feared crime (see chapter four). Looking at ethnic differences, 17% of Indian respondents thought robbery was the most common crime in their area, with 18% mentioning hijacking as most common—only housebreaking was a more common choice. Robbery was also regarded as the second most common crime among coloured (18%) and black (15%) people. While only 12% of white respondents felt hijacking was the most common crime, it was the crime most feared by white respondents, with 22% saying they fear hijacking the most—just slightly more than those who fear rape most (21%).

Based on both the official crime statistics and the victim survey, these perceptions do not correlate with reality. Only 17 respondents reported actually having been hijacked in the last year (0.3%), a number too small to place much credence in further analysis. While robbery was more common, a comparison of survey and official data indicates that the chances of any South African being robbed in any given year are about two out of a hundred (see discussion below on extent of robbery).

Robberies can be planned or opportunistic, the former often tied to organised crime, and the latter being very difficult to prevent. Because robberies are usually committed by strangers, do not take long to commit, and often involve attention-grabbing weapons, identification of the perpetrators is difficult. In addition, the property taken may be impersonal, such as cash, and therefore impossible to trace. This compounds the difficulty of responding to robbery.

Partly due to the low chances of apprehending the perpetrators and recovering lost property, robbery is generally one of the most underreported crimes. The property taken is rarely insured, there is usually little chance of the offence recurring, and traumatised victims simply want to get on with their lives. Of course, not having the details in the official record further reduces the state’s ability to prevent future robberies and to get the perpetrators off the street.

These difficulties are reflected in the low conviction rates for robbery. In 2000, the last year in which conviction rates were made public, the ratio of convictions to reported cases was only 3%.45 But this challenge is not unique to South Africa: in the United States and in England and Wales comparable ratios were only 5% and 8% respectively.46

Despite these difficulties, the SAPS has designated robbery as a “more policeable” crime,47 based on the notion that it can be reduced by increased police visibility. But robbery of individuals can be committed anywhere, and this allows criminals to plan their actions around police patrolling patterns. Indeed, unless they saturate an area with members, the more ‘visible’ the police, the easier they are to avoid in this respect.

The Ministry for Safety and Security regards the increase in recorded incidents of robbery since 1994 as one of the chief challenges facing the SAPS. Since 1994, national recorded aggravated robberies have increased by 50%, while common robbery has increased by 211%.48 Several explanations have been given by the police, including the suggestion that many reports of robberies of cell phones are fraudulent claims, filed for the purposes of acquiring insurance money in order to purchase a newer model of phone.49 Allegedly, robbery is claimed instead of theft in order to rule out counterclaims of negligence from insurers.

One theory the SAPS has not explored is the possibility that the levels of reporting by the public, rather than the actual incidence of robbery, have increased. Since the prospects of positive outcome are small (as was discussed above) the main reason for reporting robbery is out of a sense of civic duty. This inclination has hopefully increased for the majority of South Africans since 1994. This possibility is discussed further below.

Extent and risk of robbery

Two percent of respondents said they had been robbed in the past year, down from 2.4% in the 1998 survey. In addition to this apparent decline in incidence, levels of reporting to the police are also down, from 41% in 1998 to 29% in 2003. On its face, this does not tally well with the increase in the number of robbery incidents recorded by the SAPS since 1998.

But a more careful review of the survey findings reveals that not all types of robbery were captured in the poll. Just as there are many reasons why people would fail to report crime to the police, there are many reasons why people do not report crime to surveyors. One is that the incident is not regarded as memorable or important enough to report to a pollster months after the fact. Another is that the public does not understand how the legal definitions fit specific circumstances. In South African law, for example, bag-snatching—even when no resistance is offered—is considered robbery, although it might not be considered forcible by the average survey respondent.

In this case, however, it was probably the definition used in the survey, and specifically the example that was attached to the definition (“pointing a knife at someone and demanding their wallet”), that resulted in mainly serious robberies in which the use of force was evidenced through a weapon, being reported by respondents.50 Indeed, 81% of the incidents captured by the survey were armed robberies, and thus fall into the category of what the police call aggravated robberies. Bearing in mind that exactly the same definition of robbery was used in both surveys, this is much greater than the share of armed robberies captured in the 1998 poll (61%). The 1998 ratio corresponds almost exactly to that found in the police’s recorded crime statistics for that year (59% of robberies were aggravated), which is what would be expected if both aggravated and common robberies have the same reporting rate. Compare this to the difference between the 2002/3 recorded aggravated robbery ratio (56%) and the present survey ratio of armed robberies (81%). This suggests that, for whatever reason, fewer common robberies were captured in the 2003 poll.

Since the pool of people eligible for the survey (those over the age of 16) represents about 30 million South Africans, the 2% of victims reporting their robbery case to the police 29% of the time should have produced 174,000 reports. Since 81% of this figure can be classified as ‘aggravated robberies’, predictions based on the survey would be that about 141,000 incidents were reported to police. In fact, the SAPS recorded 127,000 incidents, so the survey findings correspond fairly well to official recorded crime statistics. It would appear that common robberies may have been under-captured in the survey. Because the survey appears to have more accurately captured the incidence of armed robberies, much of the following discussion will focus on this crime.

In terms of risk, the chance of any given adult South African becoming a victim of serious robbery in any given year is about two out of a hundred, but your chances vary quite a bit based on who you are and where you live.

Exactly half of the armed robberies examined in the survey took place in Gauteng and 17% took place in KwaZulu-Natal, which corresponds closely with the 46% and 21% shares held by the provinces in the official aggravated robbery statistics.

Just under 4% of Gauteng residents surveyed said they had been the victims of robbery in the past 12 months. Even looking at the official rates, the risk of being an aggravated robbery victim is twice as great in Gauteng than the national average (Figure 48). This is probably linked to the fact that the province is 97% urbanised,51 and a large part of the province falls into one of the two major metro areas of Johannesburg or Pretoria. Nationally, 3.6% of metro residents surveyed reported being robbed in the past 12 months.

Black people were disproportionately likely to say they had been armed robbery victims, comprising 92% of the victims surveyed, while they make up only 79% of the population. Men were far more likely to claim to have been robbed than women, with more than three quarters of the armed robbery victims in the survey being men. Among black men, 3.4% said they had been robbed in the last year. Younger people were also more vulnerable, with 54% of the victims being between the ages of 19 and 29, compared to the 27% share this demographic has of the national population eligible for the survey.

Nature of robbery

The majority of robberies captured in the survey took place on the street in residential areas (57%). But the second most common locus was, alarmingly, the respondent’s home, with 15% of incidents recorded as occurring there. The phenomenon of home robberies has been highlighted in past community victim surveys52 and, from last year, the SAPS started recording home robberies as a separate subset of robbery. But it would appear that the police’s crime information system is not fully utilising this category during its first year of existence. The survey projections suggest nearly 90,000 of these incidents occurred nationwide, and that 72% of victims reported the crime to the police. The SAPS however only recorded just over 9,000 home robberies in 2002/3. While these crimes were probably captured under the heading of general robbery, this discrepancy might lead the police to underestimate the scale of the problem.

While nearly three quarters of home robbery victims reported the crime to the police, only 12% of street robbery victims did so. The high rate of reporting of home robberies is probably due to many factors that would cause the victim to be more confident of a positive outcome in reporting, such as the length of contact with the perpetrator (improving chances of identification), the ability to identify the property taken, and the likelihood of multiple witnesses. In addition to the fact that 81% of home robbery victims were not alone at the time of the robbery (compared to 40% of street robbery victims), other witnesses, such as neighbours, could provide further eyewitness testimony and would probably be more likely to do so than random individuals on the street. Fear of repeat victimisation may also be a factor prompting reporting, with the householder taking proactive steps to prevent a second ‘visit’. In addition, the crime scene in a home robbery is the home, which—with access to a telephone—is a more convenient place for the victim to report the crime.

Guns were the most common weapon used in armed robberies (73%), being more popular in the home (83%) than on the street (73%). On the street, knives were also popular (26%). Injuries were more common in armed robberies in the home (24%) than on the street (17%), which may be a result of opportunistic violence, since home victims were no more likely to resist than street victims (both just under 50%).

Street armed robbers were more likely to work in groups (86%) than those robbing in the home (45%), so street robbers may exercise better physical control over victims, reducing the need to inflict injury. In addition, street armed robbers were more likely to target lone victims that those in the home: 60% of street robbery victims were alone at the time of the incident, compare to 19% of home robbery victims.

Armed robbery victims in the home were more likely to know the perpetrator (30%) than those robbed in the streets (16%), which suggests the element of intimidation may have been more important in the home context. This personal knowledge is likely behind the higher rate of arrest for reported armed robberies in the home context (24%) than the residential street context (8%).

The respondents said they were injured in the robbery in 14% of the cases, compared to about 30% of cases in 1998. This is probably due to the much higher proportion of less serious robberies recorded in the 1998 survey. Injuries were more likely to be sustained when the perpetrator relied on physical strength (54%) rather than a knife (37%) or a gun (4%). Ironically, then, the more deadly the weapon, the less harmful the encounter. When injuries were sustained, they were quite serious: respondents reported needing medical attention in 67% of cases, and were admitted to hospital in 55% of cases.

The most common items taken on the streets were handbags/wallets and cell phones, while home robberies most often claimed electronic equipment and cellphones.

Reporting to the police and case outcome

Due to the small number of robbery cases recorded by the survey, as well as the few victims who reported the crime to the police, meaningful analysis of data on arrests and case outcome is not possible.

Aftermath of robbery

As was suggested above, few robbery victims held out much hope of recovering lost property—only 9% identified this as the most important thing to them after the crime. But, like assault victims, some were stirred by a desire to see the criminals suffer (12%). The most common response by far, however, was the simple desire to get life back to normal (37%) (see Table 13, page 142).

Eleven percent most wanted to avoid future harm, and 73% said they had changed their behaviour in order to accomplish this. Thirty-four percent said they would avoid certain areas, 23% said they would be more alert, 17% said they had taken unspecified measures to make their property more difficult to steal, and 17% had given up going out altogether. Three quarters (75%) said this change had made them feel safer.

Implications of the survey results

• The public needs to be encouraged to report robberies to the police. Given the high levels of public concern about robbery, the serious and violent nature of the crime, and the fact that it has been prioritised by the police, it will be difficult to make corrective action work unless there are accurate records of these crimes that can be monitored and analysed. The differences in the nature of street versus home robberies reflected in the survey data reinforces this point: unless details are available on the nature of particular crimes, interventions that target specific elements of a crime cannot be developed.
  
  

• Given the seriousness of robbery in South Africa, the SAPS needs to clarify the massive increases in their own records of 50% for aggravated robbery and 211% for common robbery since 1994. It will be difficult to take any crime problem seriously while the possibility exists that these increases are to some extent due to fraudulent insurance claims.
  
  

• The SAPS needs to pay careful attention to recording, at police station level, of robberies that occur in the home, since the survey data suggests that many of these incidents are currently under-recorded in official statistics. Although most of the nearly 90,000 home robbery victims polled said they reported to the police, SAPS records reflect just over 9,000 cases as home robberies in 2002/3. A discrepancy of this magnitude might lead the police to underestimate the scale of the problem.
  
  

• On the street, most robbers work in groups and target lone victims—moving in groups and avoiding groups of other people on the street should be a practical prevention strategy employed by the public.
  
  

• Guns are key to robbery, so measures taken to reduce the number of guns on the street should be taken alongside a public information campaign not to resist robbery, whatever the weapon employed.

Stock theft

Public perceptions about stock theft

Only 6% of the respondents felt that stock theft was the crime that occurred most frequently in their area, with 82% of these responses coming from traditional rural areas and 91% from black respondents. Only 2% identified it as the crime they were most afraid of. In keeping with these figures, only 5% identified stock theft as the crime type about which they most recently spoke.

Extent and risk of stock theft

Stock theft was one of the most commonly experienced crimes, affecting 2.5% of respondents surveyed during the previous year, including 8% of black, 5% of white, and less than 1% of coloured and Indian respondents. Most of these crimes occurred in the Eastern Cape (34%) and KwaZulu-Natal (25%).

Nature of stock theft

Stock was most often taken from a kraal (37%), immediately outside the home (30%) or in the fields (30%). Poultry was the most common livestock taken (38%), followed by cattle (22%) goats (18%) and sheep (15%). Black African households were most likely to lose poultry (40%) or cattle (24%), while whites most often lost sheep (42%) and poultry (24%).

Most of the victims were poor—68% reported earning less than R1,000 a month—so victims were robbed of a significant portion of their annual income with the loss of one of the larger animals.

Most of the thefts occurred in the winter or early spring, with 57% occurring between the months of June and September.

Reporting to the police and case outcome

The theft was reported to the police in only 36% of the cases, which tallies with the 38% of victims who said they knew the offender by sight or name. When asked how they knew the offender, the most common answer was that it was general community knowledge (38%). Those who did not report to the police mostly felt it was not important enough (30%) or that there was no chance of recovery even if a report was made (32%). The respondent felt he knew who stole his property in only 9% of the cases, and the stolen animal was eventually recovered in just 4% of the cases.

Those who did report were satisfied with the police response in just over half of the cases, but in only 4% was the respondent aware of an arrest being made. Nearly 90% of those who were aware of an arrest said they had attended court, but that a conviction resulted in less than half (42%) of the cases.

Aftermath of stock theft

Over half the respondents (53%) reported the crime to an organisation other than the police, with 46% of these reporting to a traditional authority, and 25% reporting it to some “other” organisation. The most important thing to most victims following the crime was getting life back to normal (46%), with 24% focusing on recovering the lost livestock. Only 11% were most concerned with what happened to the criminals.

Just under half of the victims (47%) took measures to avoid further incidents of stock theft, most of which (82%) involved target hardening of one sort or another. Nearly 90% felt more secure after this action.

Implications of the survey results

Stock theft is clearly a crime that is resolved primarily outside the criminal justice system. Low reporting rates may be related to accessibility of the police in rural areas, use of traditional authorities to resolve the matter, and the fact that positive outcomes for reporting are rare. Less than 10% of respondents felt they knew who stole their animal at the time of the survey, and recovery of their property had occurred in less than 5% of cases. This is a pity because the impact of this loss on poor households could be significant.

Reducing the impact of stock theft could take two paths. One would be to reduce victimisation through target hardening. Since most of the victims felt safer after taking additional precautions, the police could assist by recommending safety measures before victimisation occurs. The other option is to increase the rate of recovery after the fact. This could be advanced by making livestock easier to identify or track. The state could become involved in registering ownership and marking animals accordingly. Low cost ways of doing this could be developed, and marking stock may also have a deterrent effect.

Conclusion

A very small share of those who become victims of crime are most concerned with the apprehension or punishment of the perpetrators, with most being chiefly concerned about returning life to normal and, where relevant, recovering lost property (Table 13). Victims of the violent crimes of robbery and assault were slightly more vindictive. Very few victims of any stripe were civic minded enough to be most concerned with incapacitating the offenders.

From a purely selfish perspective, this makes good sense. With the exception of interpersonal crimes like assault, where the victimisation may be continual, taking a single criminal off streets is likely to have little impact on the safety of the victim. In terms of providing a service to the victims of crime, it is therefore imperative that the police and the criminal justice system inconvenience victims as little as possible in their investigations. As far as recovering property is concerned, the public should be informed about the true chances of recovering their property through the criminal justice system.

All this has implications for those wishing to increase citizen participation in reporting and prosecuting crime. Victims must be encouraged to see beyond their selfish interests and to participate in the criminal justice process out of a sense of civic duty. Of course, it is incumbent on the police and the rest of the system to make this process as painless as possible, by reducing the number of appearances required and otherwise streamlining the process. The police, prosecutors, and judges need to be aware that witnesses are performing a public service, and should be treated accordingly.