CRIME PROFILE: VIOLENT CRIME

____________________________________________________________

Published in Crime in Pretoria
Results of a City Victim Survey

Definitions of violent crime, both in police statistics and victim surveys, generally refer to any crime in which the offender has some physical contact with the victim. These would include any robbery (including car hijacking), mugging, rape, assault and murder. For the purposes of this study, however, the crimes that will be considered in this ‘violent crime’ profile are assault, sexual assault and murder. The circumstances in which many of these crimes occur, are similar and the interventions to reduce their occurrence are likely to be related.

Car hijacking, which is traditionally considered to be ‘violent’, will be outlined in the ‘vehicle crime’ profile. Mugging/robbery, for similar reasons, will be profiled distinct from other crime types.

While sexual assault is included in this crime profile, detailed analyses of many of the circumstances around which these crimes occurred are prohibited by the small numbers of respondents in each category. As a result, levels of sexual assault will be discussed, but not the details of this crime category. (These are considered in more detail in the Addendum on sexual harassment and assault at the end of this report.) It has been recognised that crime surveys measure violent crimes less well than property crimes,²⁵ even though the violent crime estimates in the Pretoria city survey are far higher than those recorded by police statistics. Some violent crimes may not have been considered a crime, for example fights between young men in bars or shebeens. Domestic violence may either be overlooked as a crime, or may be too sensitive to disclose to interviewers in a survey. Indeed, violence in which the victim and offender are known to one another is least likely to be reported to a victim survey.

The following characteristics of violent crime will be considered:

the levels of violent crime and a profile of who the victims are;



who is most at risk;



when and where violent crime occurred;



the circumstances in which the crimes were committed, including numbers of victims and offenders and their relationship; and



the nature of the violence including weapons used and degree of injury.

The violent crimes were defined by the survey in the following way:

Assault includes any incident in which the victim was personally attacked or hurt. This may have occurred in the home or elsewhere in public places, and may have been committed by someone known to the victim or by a stranger(s). Weapons may or may not have been used.



Murder includes the deliberate killing of someone in the respondent’s household or immediate family residing in the same house as the respondent.



Sexual assault includes any action in which a woman is forced, either by threats or the use of force, to perform sexual acts against her will. These acts may include — but are not restricted to — sexual intercourse, and can occur in the victim’s home, someone else’s home or a public place. The perpetrator may or may not be known to the victim.

Between 1993 and April 1998, 499 violent incidents were reported to the Pretoria city survey. Of these, 246 were assaults (the majority of which happened in 1997), 219 were murders (most of which happened in 1995), and 34 were sexual assaults (Table 2).

The comparatively high number of crimes that occurred, according to victims, in the first four months of 1998 probably reflects the reality to some extent, but also the phenomenon of ‘telescoping’ — a tendency in victim surveys for respondents to recall crimes as happening more recently than was actually the case.

 

Table 2 suggests very high levels of murder in Pretoria. However, this figure (a trend consistent in other city surveys) is undoubtedly the result of overreporting. Respondents were asked whether a member of their household had been murdered over the past five years. Given the seriousness of murder (and thus people’s willingness to discuss it), and the extended nature of many urban South African households, as well as migrancy which links urban and rural families, respondents in all likelihood related murders that happened to people living beyond their immediate homes, and possibly even outside of the five years survey period.

Of the people who reported violent crimes to the survey, the vast majority were African (86 per cent). Just under half (45 per cent) were between the ages of 26 and 40 years and 59 per cent were males. The figures reflected in Table 3 show the numbers of violent crime victims as a proportion of race, age and gender categories for the different types of violent crime.

The victim profile in Table 3 indicates some degree of risk associated with violent crime. Table 4 shows the risk for all adults in Pretoria of violent crime relative to other crime types. Of all those interviewed in Pretoria, Africans, people aged between 26 and 40 years and men were most at risk of violent crime.

An analysis of violent crime victimisation within the race, gender, age and settlement type subsamples provides more insight into the relative risk for various groups:

Risk of violent crime within each race group: 22 per cent of all Africans in Pretoria were victims of violent crime between 1993 and April 1998. This can be compared with 9 per cent of Asians, 5 per cent of whites and 3 per cent of coloureds in the metropolis. This means that Africans’ chances of becoming victims of violent crime are four times higher than those of whites. Figure 16 shows the risk for Africans, coloureds, Asians and whites for different types of violent crime.

Figure 16: Risk of crime for different race groups 1993-April 1998

Risk of violent crime within each age group: Young people are at greater risk of violent crime in Pretoria than older people. Twenty per cent of all people between the ages of 18 and 25 years were the victims of violent crime between 1993 and April 1998, compared with 17 per cent of those in the 26 to 40 age category, 14 per cent of those between 41 and 60 years and 4 per cent over 60 years of age. While the risk of assault is evenly spread across age groups, the risk of murder is higher for the youngest age category surveyed in Pretoria.



Risk of violent crime for men and women: Men in Pretoria are only slightly more at risk of violent crime than women. Sixteen per cent of all males interviewed said they were victimised, compared with 14 per cent of women. These risks are consistent across different types of violent crime, namely assault and murder. Sexual assault is not included here since men were not interviewed on this crime.



Risk of violent crime for people living in different parts of the city: 15 per cent of those who live in Pretoria’s townships were victims of violent crime, compared with 7 per cent of people residing in informal areas, 5 per cent in the suburbs, and 1 per cent in the inner city.²⁶

Most violent crimes (57 per cent) were committed during the week. Although less (43 per cent) were said to happen on the weekend, a large proportion of those offences (43 per cent) that occurred during the week, occurred on Fridays. Two-thirds of violent crimes were committed after dark: 64 per cent happened between 18h00 and 6h00. Twenty three per cent occurred in the afternoon between 12h01-18h00, and violence was least likely to take place before midday (13 per cent).

Victims were asked to specify the exact location of the crime. Answers were then grouped into the categories of township, suburb, inner city and informal settlement. Of those assault and sexual assault victims²⁷ who could identify where the crime happened, 51 per cent said they were assaulted in a township, 30 per cent in a suburban area, 10,5 per cent in the inner city and 9 per cent in an informal settlement. This supports earlier analyses which show that Africans (who largely inhabit townships and informal settlements) bear the brunt of violent crime in Pretoria (and, indeed, in the country as a whole).

Respondents were also asked to specify other more generic locations. Most happened in public places (64 per cent). Thirty six per cent happened in private areas, including the victim’s home and someone else’s home (Figure 17). Many violent crimes occur in environments familiar to the victim, which in turn suggests that it is more likely to be acquaintance violence rather than stranger violence (discussed below).

This presents particular problems for those concerned with reducing violent crime levels in Pretoria. Offences committed in private places are both difficult to detect and respond to, and rely heavily on the victim’s co-operation for successful prosecution. Given the challenges in reacting to many of these crimes, particularly those occurring in the home, interventions should aim to prevent these offences from occurring in the first place. Those violent crimes that are committed in public places such as streets and places of entertainment, on the other hand, can often be reduced through measures aimed at controlling and regulating the issuing of liquor licenses, the carrying and availability of weapons in such places, operating hours, and restricting under age customers. Other factors which require more long term interventions have also been linked to violent crime, such as overcrowding, housing density and poverty.²⁸

Figure 17: Where violent crimes happened

Figure 18: Weapons used in violent crime

More than half of the violent incidents are a one-on-one event. People are most at risk of violence when they are alone: 59 per cent of all violent crimes were committed under these circumstances. This trend was more pronounced for sexual assault victims than for others, of whom 74 per cent were alone when attacked.

Most cases of violent crime (59 per cent) also involved only one offender. Equal proportions of attacks were committed by two people (20 per cent) and by more than two people (21 per cent). Incidents of assault were more likely to involve more than one offender than murder and sexual assault.

A significant proportion of violence can be regarded as ‘stranger’ violence, since in 46 per cent of cases the victim did not know the offender. However, most victims knew their attackers: 32 per cent by name and 21,5 per cent by sight only. Murder was more likely to involve acquaintances than assault: 60 per cent of the victims were said to have known their attackers either by sight or by name, compared with 47 per cent of assault victims.

Knowing the offender should assist the police in investigating and prosecuting these crimes. However, conviction levels for murder in South Africa are much lower than in other countries,²⁹ despite the fact that many victims know their attackers (a trend which is also evident from the other city victim surveys conducted by the ISS). According to the police, victims are more likely to withdraw charges when they know the offender (particularly in the case of domestic violence). They may also be more easily intimidated and thus less likely to proceed with the case, or to make a convincing witness.³⁰

Victims were asked about several factors which could help to explain the violent crimes they had experienced:

Alcohol and violent crime: People were largely unsure about the relationship between alcohol abuse and the offences they described. Equal proportions said the attacks were related to alcohol consumption (32 per cent) as did those who said they did not know (32 per cent). Assault victims were slightly more sure than victims of violence, in general: 37 per cent said the incident was related to alcohol abuse.



Drugs and violent crime: The link between violence and drugs is even weaker than that with alcohol. Only 9 per cent of victims thought there was a relationship; the largest portion were undecided (47 per cent) and 44 per cent said there was no connection.



Gangs and violent crime: Just over half (52 per cent) of all victims of violence thought there was no connection between gangs and the offences they reported to the survey. A large proportion, however, were unsure: 37 per cent said they did not know, and only one in ten said the attack was gang-related.

Nature of the violence

The largest proportion of violent crimes in Pretoria were committed without using a weapon: physical strength was used in 44 per cent of violent offences. A gun and a knife were used in 20 per cent and 18 per cent of attacks respectively (Figure 18). Although weapons were not used in nearly half of all violent incidents, this did not reduce the level of injuries sustained by victims of assault and sexual assault: the vast majority (83 per cent) said they were hurt in the attack.

Those who reported violent crimes such as assault and murder were much less likely to say they changed their behaviour as a result of the crime than victims of burglary, car theft or mugging and robbery. Fifty nine per cent of assault victims and 54 per cent of people who reported a murder, said they changed their behaviour. In contrast, over 70 per cent of the victims of burglary and robbery responded similarly. It is likely that many people are unsure how to prevent violent crimes of this nature — particularly those that occur in the home and between acquaintances or family members.

The majority of violence victims said they responded by avoiding particular situations: 61 per cent changed their behaviour in this way. Few responded by improving home security (11 per cent). This could suggest that the majority of these victims cannot afford such measures, but probably relates more to the nature of the violent crimes they are trying to avoid. Since much of the violence reported to the survey occurred in the home or surrounds, and between people who knew one another, precautions such as home security and being more careful and aware, are unlikely to have much effect. The avoidance behaviour, however, paid off: 61,5 per cent of those who took preventive measures felt safer as a result.