CHAPTER 4

PUBLIC PERCEPTIONS ABOUT CRIME AND SAFETY

Information about public perceptions of crime and safety should be as important to policy makers as that about the actual crime rate. Perceptions influence behaviour, with the potential for both negative and positive outcomes for those trying to reduce crime. On the one hand, a belief that crime levels are spiralling and the police have little control over the situation may result in desperate citizens taking the law into their own hands. This undermines the criminal justice system and adds to the burden of the police. On the other hand, concerns about safety may incline people to adopt a more constructive approach to protecting themselves and their property—in the form of, say, burglar-proofing their homes or avoiding unsafe areas. Such actions can assist government’s efforts to reduce crime levels. It is therefore important to track public perceptions over time in order to maximise the positive impact on behaviour, and pre-empt negative outcomes.

However, understanding public perceptions about crime is not easy. Perception data gathered in quantitative surveys often seem contradictory and counter-intuitive, and so deter policy makers from using the results. For example, the public may rate the police as fairly effective and then, in the same survey, go on to criticise them for being corrupt. While some of the blame for this rests with the way the information is gathered and analysed, perceptions—whether informed or not—are real, and need to be understood rather than dismissed, no matter how complicated a task this may be.

Public perceptions of crime are influenced by both primary and secondary factors. Primary factors include first hand experiences of crime that an individual or their family or friends may have had. Secondary factors, which often have a wider impact, include media reports, other documentary information about the crime situation, and general word of mouth. Differentiated access to the media may thus be one factor in explaining different perceptions. Notwithstanding this, most South Africans have access to some form of media, whether television, radio or the press. It is hardly debatable that bad news sells, and that as a result, the media often portray South Africa as overrun by crime.

But the public do not simply absorb information in an uncritical or un-selective manner. Perceptions are also likely to be influenced by expectations and by what people are accustomed to. Those who are accustomed to high crime levels and poor service delivery may react with indifference when asked whether they fear crime or are satisfied with service from the police and courts. In contrast, those who have lived in relative safety will be outraged by a spate of burglaries in their neighbourhood.

Perceptions may equally be influenced by general sentiments about life and government, rather than by any specific incident. Service delivery surveys have shown for example, that the general public are far more critical of police and court service than those who have just had contact with police officials or prosecutors.10 The same studies showed that the general public tended to say the police are corrupt, while none who had recently used police services mentioned corruption as a problem. These results suggest that general perceptions about government performance, and corruption in this case, are often projected in answers about specific situations.

In this survey, a range of questions was asked about how people thought the crime level had changed, which crime types they believed were most prevalent, and which they feared the most. Questions were also included on feelings of safety when walking alone during the day and night. All these questions referred to the situation in the respondent’s area of residence. Views on who perpetrates crime and why, and on the impact of crime on behaviour, were also canvassed. The next chapter covers public perceptions about the performance of the police and courts, as well as community and individual responses to crime.

Views about changes in the crime level

Just over half (53%) of South Africans believed that crime in their area of residence had increased over the past three years. Although this represents a majority, it is surprising, given the high profile of the crime problem, that the margin is not wider. One quarter (25%) of South Africans believed the level of crime had stayed the same, and one fifth (21%) said crime in their area had decreased.

Those in Limpopo province were much more likely than respondents in any other province to say crime had decreased. People living in Mpumalanga, Eastern Cape and North West were most likely to feel that crime in their area had increased (Table 1, Appendix 2).

Besides provincial differences there were also differences evident between groups of different social, demographic and economic composition. For example, more than three quarters (78%) of Indians felt that crime had increased, while 61% of whites felt this way, compared to just over half (51%) of blacks and less than half of coloureds (48%).

These racial trends reflect those identified in earlier surveys, where Indian South Africans generally felt the most negative about safety and crime in South Africa. In a public opinion survey conducted by the Human Sciences Research Council (HSRC) just prior to the 1999 elections, 75% of Indians felt unsafe in terms of their daily personal safety, as opposed to 63% of whites, 44% of blacks and 43% of coloured South Africans.11 A more recent HSRC survey however found less difference between the race groups on the issue of changing crime levels. Conducted in July 2001, the study asked whether respondents thought violent crime and political violence had increased since 1994. Most (75%) felt that violent crime had increased since 1994, with 87% of whites, 84% of coloureds, 73% of Africans, and 73% of Indians saying this was the case.12

The 2003 National Victims of Crime survey showed that the length of time people have lived in the area is not significant in informing their opinions of how crime has changed within the area. The type of area in which South Africans live does however play a significant role in determining their perceptions about changes in crime levels. Those of all races living in both metropolitan (61%) and urban (54%) areas were significantly more likely to feel crime generally has increased than those living in traditional rural (48%) or farming (43%) areas. However, within each area type, the same racial trends as on a macro level are present, with Indians and whites in metropolitan and urban areas more likely to feel that crime has increased than black or coloured South Africans.

Views about violent and property crime levels

‘Crime’ is a broad description for a range of quite different acts, and the various types of crime may impact on individuals, households or communities in a variety of ways. Much of what gets exposure in the media is violent crime, and it is this type of crime that probably has the most devastating effect on people. Violent crime includes acts where force is used against an individual, such as murder, rape, assault, sexual assault, robbery and car hijacking. Crimes that involve the removal of belongings are classified as property crime and include housebreaking, theft of personal property, theft of vehicles, bicycles, crops and livestock. Survey respondents were asked whether they thought violent and property crime levels had changed in the past three years.

Despite the media’s emphasis on violent crime, more respondents thought that property crime had increased (55%) than those who said violent crime had gone up (47%). These views are easier to understand when disaggregated by area type. As with perceptions on crime generally, those living in metropolitan and urban areas were most likely to feel that violent crime in their area has increased over the past three years (Figure 12).

A much stronger correlation exists, however, between where respondents live and their perception of violent crime, than their perceptions of crime in general. Less than two out of five of those in traditional rural or farming areas felt that violent crime has increased. The differential, and indeed the correlation between location and perception, is less marked when considering property crime: while more than half of those in urban areas thought property crime had increased, only slightly less in rural areas and another two out of five in farming areas thought property crime had increased.

This implies that while more South Africans in urban areas, including the metros, feel that both violent and property crime is increasing, property crime is perceived to be on the rise in rural areas.

Generally, while the majority of those in all income categories tended to feel that crime has increased over the past three years, those with a per capita income in excess of R20,000 were more likely to say this. Specifically, 83% of those with an income of between R30,001 and R40,000 felt that crime had increased, as opposed to 62% of those earning between R1,001 and R1,500. This is explained in part by the fact that most of those earning higher incomes are resident in the metropolitan and urban areas, where most South Africans felt that crime has increased.

The perception that crime is increasing, as well as direct and indirect experience of victimisation, may influence South Africans’ views about specific crime types. Respondents in the survey were asked what one type of crime they thought occurred most in their area, whether they thought the level of this crime had changed over the previous three years, and then what crime frightened them most.

Crime types perceived to be most common

Housebreaking was the crime that a majority of South Africans said occurs most often in their area (Figure 13). This reflects the higher percentage of respondents who felt that property crime had increased than those who thought violent crime had increased. After housebreaking, robbery was perceived to be most common, followed by property theft.

While there is no statistical correlation between the crimes perceived as most common and the province in which people live, it is unsurprising—given that vehicle ownership is higher in Gauteng than other provinces—that of those saying vehicle theft was the most common crime in their area, most lived in Gauteng. Similarly, more vehicle owners tended to rate car hijacking as the most common crime than non-vehicle owners.

Respondents in all the provinces thought that housebreaking was the single most common crime in their area, with the exception of Northern Cape, where assault was believed to be the most prevalent crime. People in Northern Cape were also much more likely to think that rape was the crime that occurs most frequently. Gauteng residents were much more likely than those of any other province to say that robbery as well as car hijacking were most common (Table 2, Appendix 2).

Interestingly, while housebreaking was viewed as the most common crime in metropolitan, urban and traditional rural areas, assault was viewed as more common in farming areas.

The various per capita income levels do not appear to influence the perceptions of which crime occurs most, with housebreaking believed to be the most common crime across all income categories.

Given that the majority of those interviewed felt crime in their area had increased, it is unsurprising that two thirds (66%) of respondents felt that the one crime they thought occurred most in their area had increased over the past three years.

Most feared crime types

When asked what type of crime they were most afraid of in their area, murder was most commonly cited (25%), despite the fact that it was ranked only fourth on the list of those crimes perceived as most common. However, given the violence associated with murder and the fact that South Africa has one of the highest murder rates in the world, the results are not that surprising.

After murder, the crimes that people were most afraid of are housebreaking, sexual assault and rape, and robbery (Figure 14). That three of the four crimes that South Africans fear most are violent, indicates the significance of violence in understanding public fears about crime. It is also telling that three of the four crimes perceived as most common (Figure 13), are among the four crime types that South Africans are most afraid of.

When analysing fear of crime, the responses provided by respondents can be interpreted in a number of ways.13 First, people may worry about the impact of becoming a victim on their own life, or those of their household members. Second, the crime which respondents fear most may be the one they perceive as the most likely to happen to them. For example, 14% of vehicle owners reported that car hijacking was the crime they were most scared of, as opposed to only 4% of the total sample. However, the results suggest that generally it is violent crimes such as rape, assault or murder that people are most scared of, as these are likely to have the most traumatic impact emotionally and physically, and may result in death.

There was some variation when the results were disaggregated by race. While black and coloured South Africans tended to be most afraid of murder, Indians feared housebreaking most, and whites worried most about vehicle hijacking. In the case of whites’ perceptions, this can be partly explained by the fact that vehicle ownership is higher among white South Africans than any other racial group. The horrific media accounts of hijackings that end in rape and murder no doubt also play a role.

Gender was also a significant variable in understanding the crimes that respondents feared most. Women tended to fear rape and sexual assault most, while men were most scared of murder. This further illustrates the concerns about the impact of crime as discussed above.

Area type was also significant, with those in farming areas most afraid of murder. Murder also topped the list of most feared crimes for those living in metros and urban areas, although robbery and housebreaking were feared almost as much in these two areas (Table 3). Traditional rural areas were the only ones in which murder was not the most feared crime, with housebreaking being the more common worry.

Although murder was the one crime feared most overall in the country, it was only feared most in three of the nine provinces, namely North West, Eastern Cape and Western Cape. Rape was the most feared crime in Northern Cape and Free State, while in Gauteng it was robbery (followed closely by murder). In KwaZulu-Natal, Limpopo and Mpumalanga, housebreaking was the crime people worried about most (Table 3, Appendix 2).

Crime as a topic of conversation

Respondents were asked if they had talked about crime in any conversation with friends, family or colleagues over the previous two weeks. Fractionally under half (49%) of South Africans had discussed crime during this time period. The one crime type that was most often discussed was housebreaking.14 Given that housebreaking was believed to be the most prevalent crime and was the second most commonly feared crime, this is not unexpected. After housebreaking, the most talked-about crimes were of a violent nature, namely robbery, murder, and sexual assault or rape (Figure 15).

There was some variation in the trends according to province, area type, and race. Those in Gauteng (57%), KwaZulu-Natal (55%) and Limpopo (53%) were most likely to have discussed crime over the previous two weeks. Respondents living in Eastern Cape (40%) and Mpumalanga (29%) were least likely to have discussed crime (Table 4, Appendix 2). Those in metropolitan (58%) and urban areas (53%) were more likely to have discussed crime than people living in traditional rural (40%) and farming (39%) areas.

An analysis by race reveals the same trends in perceptions about crime identified earlier, with whites followed by Indians most likely to talk about crime, while coloureds and black South Africans were least likely to discuss the subject. This trend for Indians and whites to be more concerned about crime and safety than blacks and coloureds, was found in other results discussed later in this chapter, and in the next chapter on perceptions of the police and courts.

Feelings of safety

In keeping with the International Crime Victim Surveys (ICVS) as well as other victim surveys conducted by the ISS, respondents were asked how safe they feel walking alone in their areas of residence during the day and after dark.

In total 85% of South Africans said they feel safe walking alone in their area during the day, while only 23% felt safe walking alone at night (Figure 16). The percentage of South Africans feeling safe at night is higher than that recorded in other comparable site-based victim surveys in the country, including Cato Manor, Hillbrow/Inner Johannesburg, Cato Crest or Meadowlands.15 However, significantly less South Africans felt safe walking in their area at night than those surveyed in developing countries, including Africa and Latin America (see text box).

It is interesting to note that those South Africans who have lived in their area for ten years or more were most likely to feel very safe walking alone either during the day (62%) or during the night (12%).

A comparison between feelings of safety in 199816 and 2003 presents some remarkable differences. As Figure 17 below illustrates, during the day, the public felt generally as safe in 2003 as they did in 1998, if the ‘very safe’ and ‘fairly safe’ categories are added together. However, significantly more felt only fairly safe in 2003 as opposed to very safe in 1998.

This tendency towards feeling less safe becomes an unmistakable trend when the night-time results are considered. South Africans felt significantly less safe when walking alone after dark in 2003 than they did five years ago (Figure 18). More than double the number of people in 2003 than in 1998 felt very unsafe walking in their area after dark (58% in 2003 as opposed to 25% in 1998).

It has been argued that perceptions of crime and safety can be influenced by a range of factors, including the direct experience of victimisation. One might expect those who have been victims of crime in the recent past to feel less safe than those who have not been victims. While half (50%) of those who had been victims of crime in the past year still felt safe walking alone in their area during the day, it is somewhat lower than the almost two thirds (62%) who had not been victims of crime. A similar difference exists between victims and non-victims who felt safe after dark: while one quarter (25%) of non-victims felt safe walking alone in their area of residence after dark, less than one fifth (18%) of those who had been victimised in the past year felt safe.

The type of area in which South Africans live is a significant factor in determining how safe they feel, both during the day and at night. While those in traditional rural areas were most likely to feel safe walking alone during the day, those in farming areas were most likely to feel safe at night (Figure 19).

* 'Fairly safe' and 'very safe' responses have been combined to ceate the category 'safe' used here

There were also marked differences in feelings of safety between the race groups. Indian followed by white South Africans were least likely to feel safe during the day (Figure 20). The same pattern applied to results on night-time safety. This trend is similar to that about views on changes in the crime level discussed above, in which Indian and white people were more concerned about crime levels than coloureds and blacks.

During the day, people in Eastern Cape were more likely than those in any other province to feel safe walking alone, while those in Gauteng felt most unsafe (Table 5, Appendix 2). After dark, Gauteng residents also felt most unsafe walking alone at night, while those in Western Cape felt most safe (Table 6, Appendix 2).

Impact of crime

Crime, as well as the fear of crime, can affect individuals, communities and society in a range of ways and understanding the impact of crime would require a study focusing specifically on the subject. The aim of this survey was not to examine the impact of crime, and the few questions that were included were aimed at testing their usefulness in a quantitative survey of this kind.

In the first set of questions about impact, respondents were asked whether crime in their area prevented them from engaging in day-to-day activities ranging from recreational activities to those essential for survival such as walking to fetch water. They were also asked whether crime prevented them from taking part in these activities when alone, and when accompanied by one or more people they knew. Questions such as these that focus on specific activities rather than subjective measures of how safe people feel, can be the best quantitative indicators of how fear of crime is changing and whether crime prevention is working, if repeated in follow-on surveys over time.

The second set of questions covered whether respondents had, in their lifetime, witnessed a murder, and if so, how old they were at the time and whether they knew the victim. The aim of these questions was to estimate the extent of the phenomenon in South Africa, given the damaging psychological impact this can have, and the links between witnessing violence, particularly at a young age, and the risk of offending later in life.

Impact of crime on daily activities

Almost one third of respondents reported that the levels of crime in their area stopped them from letting their children play freely in the neighbourhood. Slightly less said that they did not walk, play or rest in open spaces or parks, and more than one quarter reported that as a result of crime they did not allow their children to walk to school (Figure 21).

One fifth did not walk to town or work because of crime, while slightly less did not walk to get firewood. Less than one tenth did not walk to fetch water due to the crime level.17 These last two, while relatively small, are a matter of concern given that both activities are essential in many rural areas for household survival. Black respondents were least likely to be prevented from walking to shops or to work or town (Table 4). This is unsurprising, given that most black South Africans who are poorer than other race groups, have fewer alternatives when it comes to the means of transport, and so are forced to walk regardless of their concerns about safety. In total, 15% of black respondents reported that the level of crime in the area prevented them walking to the shops when alone, compared to 17% of coloureds, 34% of whites, and 49% of Indians. The high percentage of Indians saying they would not walk to the shops or to work or town reflects the high levels of concern about crime and safety among this race group discussed earlier in the chapter.

In addition to race, the area that people live in also affected the extent to which daily activities are impacted by crime levels. Those living in the metro and urban areas were most likely to have their daily activities impacted by the crime level (Table 5).

Given the focus placed by government on entrepreneurial and Small Micro and Medium Enterprise (SMME) development as one means of creating employment, the fact that 26% of South Africans said that crime stopped them from starting or investing in a home business presents a challenge to this strategy (Table 6). It also emphasises the complex relationship between crime, the fear of crime, and various forms of development.

The implications of these findings are serious considering that most of those who said they would not use public transport if alone, walk to the shops alone, walk to work alone or allow their children to walk to school, do not own a car or even a bicycle. They thus do not have any alternative means of reaching their destination.

There was some variation in the impact of crime on behaviour at a provincial level. For example, the lowest percentage of people reporting that crime stopped them from using public transport when alone was in the Northern Cape (10%). By comparison, almost one third in Gauteng said this was the case. Gauteng also had the highest number of people reporting that crime stopped them from walking to the shops, walking, resting or playing in open spaces, allowing their children to play freely in the area, and allowing their children to walk to school (Table 7, Appendix 2).

Many of those who said crime in their area prevented them from undertaking the listed activities when alone, said the situation would change if accompanied by one or more people they knew. Almost three out of five (58%) of those who reported that they would not fetch firewood alone said that they would fetch firewood if accompanied by someone they knew. This applied to 57% of those who would not walk to the shops alone, as well as those who would not walk to town alone. Fifty-six percent of those who would not—when alone—use public transport, walk, play or rest in parks, and walk to fetch water said they would undertake these activities if accompanied by someone they knew.

Witnessing murder

More than one tenth (14%) of respondents had personally witnessed a murder. Unsurprisingly, given the greater concerns about violence and crime in general in urban areas, a significant relationship exists between the likelihood of witnessing murder and the type of area in which the respondent lives. Those in metropolitan (19%) and urban (15%) areas were more likely to have seen such an incident than those in traditional rural (10%) and farming areas (9%). The trends across the provinces were less significant (Table 8, Appendix).

Race is also a factor, with black South Africans significantly more likely to have seen a murder than whites and Indians: 15% of black South Africans said this was the case as opposed to 12% of coloureds, 9% of whites and 4% of Indians. Although these percentages seem low, they translate into nearly four million South Africans: 3,320,355 blacks, 321,535 coloureds, 303,176 whites and 34,564 Indians.

It has been noted that exposure to violence at a young age, together with exposure to crime, is one of the main risk factors for developing anti-social behaviour at a later stage in life.18 Given this, it is a matter of concern that the majority of those who witnessed a murder had seen the incident when they were between 16 and 25 years of age (53%). The next most likely age group was 35–45 years (28%). Males (17%) were also more likely to witness murder than females (10%).

More than half (52%) of those who had witnessed a murder knew the victim, with coloureds (67%), followed by Indians (56%) most likely to have known the victim. Just over half the black respondents (52%) and 31% of whites knew the victim of the murder they saw.

When the victim was known to the respondent, most commonly the victim was a neighbour (43%), a friend (21%) or another relative (excluding sibling, parent or child). Just 5% of respondents who knew the murder victim identified the victim as a sibling, and 2% said the victim was a parent.

Views about criminals

Those who perpetrate crime are often regarded as ‘outsiders’ or ‘others’—outcasts rather than ordinary members of society. However, the stereotypes about criminals belie the fact that those who commit crime live among us in our homes and neighbourhoods, and as the police readily point out, are often known to and protected by their community and family. If this is the case, the implications for crime prevention are significant. Efforts will need to be made to encourage people to be less accepting of criminal behaviour and to work with the police and courts to prosecute offenders. This requires not only carefully targeted awareness campaigns, but also the kind of treatment and protection for witnesses that encourages and maintains participation in lengthy trials.

In an attempt to explore this phenomenon, as well as public attitudes to criminals in general, respondents in the survey were asked whether they personally know someone who makes a living from crime in their area. They were also asked a range of questions about who they think commits crime in their area, and what their motives might be.

The results show that South Africans’ knowledge about who commits crime in their community is indeed high. More than one quarter (29%) of South Africans said they personally know someone who makes their living from crime in their area. More people living in the Eastern Cape said this was the case than in any other province, followed by those in Limpopo and Free State (Figure 22). It is interesting that the two provinces where people were most likely to know someone who makes a living from crime are also the two poorest provinces in South Africa.

Despite this provincial profile, there was little difference between the metropolitan (30%), urban (29%) and traditional rural areas (30%) in terms of people who know someone who makes their living from crime. However, less people in farming areas (20%) said this was the case.

Trends according to race were more significant than those by area type. Coloured (36%) and black (32%) South Africans were most likely to know someone who makes a living from crime, while less than one fifth (17%) of Indians and just 7% of whites said this was the case.

Respondents were asked two questions regarding who they thought was responsible for most of the crime in the area where they live. The first looked at the origin, or birthplace of perpetrators, and the second covered their residency. The responses clearly indicate that the vast majority of South Africans believe that people born in South Africa, rather than foreigners, are responsible for most crime (Table 8), and that it is people who live in the same area who commit most of the crime (Table 7). These perceptions were common for both violent and property offences.

Some of these views differed significantly, however, when analysed by race and by area type. While everyone agreed that foreigners were not responsible for most crime, a majority of whites thought that both violent and property crimes were more likely to be committed by South Africans from outside their area than by locals: 73% of whites said violent crime was likely to be committed by South African ‘outsiders’, compared to 37% of Indians, 33% of coloureds and only 25% of blacks.

Respondents living in the metropolitan areas of the country were much more inclined to think that ‘outsiders’ to their area—whether South African or of foreign birth—were the likely perpetrators of both violent and property crime. Nearly half (47%) of metro residents said people from other areas in South Africa were responsible for violent crime, compared to 36% in urban, 35% in farming and only 24% in traditional rural areas. Metro respondents (8%) were also more likely to blame foreigners for violent crime, compared to 4% in urban, and 2% in both traditional rural and farming areas. Similar trends applied to views on who commits property crime.

The results show that the public places the blame for most crime with South Africans as opposed to “people who were born outside South Africa”, and with the exception of white respondents and those in metro areas, with people local to the area where respondents live. The majority of South Africans therefore recognise that those responsible for crime are indeed local community members rather than ‘outsiders’—whether from another area in the country, or from abroad.

Respondents were also asked whether they think those responsible for committing violent and property crime are motivated by either “real need, greed or non-financial motives”. The latter would apply to all non-acquisitive violent crime.

The two most common reasons cited for both property and violent crimes were greed and non-financial motives. “Real need” was almost as common an explanation for property crime as the other reasons (Figure 23). Although it may be assumed that poorer respondents would be more likely to cite “real need” as the motivation for crime, no statistical relationship was found between respondents’ income or that of their households, and their likelihood to cite this as the primary motivating factor for crime.

Whether respondents lived in rural or urban areas did not alter these trends. There was some difference between the provinces, however, with those living in North West much more likely than any other province to cite real need as the motivation for property crime. More respondents in Limpopo than any other province said greed was the driving factor behind property crime. The trend for violent crime across the provinces was similar to that for property crime (Tables 9 and 10, Appendix 2).

In terms of race, views varied significantly. Indians were much more likely to cite greed than any other motivation, while within other race groups, non-financial motives were the most common reason provided (Table 9).

Implications of the survey results

Responding to declining feelings of safety

The results in chapter six will show that crime levels, as measured by the victim surveys, have dropped slightly since 1998. Despite this, the public feel much less safe now than they did five years ago. Unless a concerted effort is made by government to improve perceptions of crime and safety, the real achievements of the police and courts will make little difference to ordinary people.

A strategy is needed to deal with the problem. Lead by the SAPS national commissioner, it should have a strong public relations component. The message must go beyond saying crime in South Africa is not as serious as people think, or that the police crime statistics show that the situation is improving. Rather it could be based on the following elements:

• Government, and particularly the police, need to acknowledge the serious nature of crime in South Africa, and especially the high levels of violence. Statements to the contrary will simply fuel public concerns about their safety. Along with this recognition, the public needs to be convinced by the police that government knows what needs to be done to reduce crime, and is acting decisively.
  
  

• The public need access to accurate and up to date comparative information from as wide a range of sources as possible (including among others, police statistics and victim survey results such as these). An approach that attempts to reduce or downplay the importance of crime information will only increase public scepticism of police efforts to deal with crime.
  
  

• Well-publicised and convincing case studies about successes should be disseminated in a variety of ways. These could be both government or non-government initiated projects, and could include reductions in crime in notorious areas like Hillbrow, improvements in the experiences of victims and witnesses in court, the achievements of special units, or examples of successful partnerships between communities, police and municipalities for example, that have reduced crime.
  
  

• The strategy must include information on how the public and especially victims of crime can access the police, social workers, prosecutors and other relevant agencies, and make the criminal justice system work for them. Until the public believe that the police and courts place the interests of crime victims first, and until they know how the system works and what they can expect from it, they will continue to feel isolated and unsafe, regardless of what the statistics say.

Needless to say, the strategy would need to be based on a sound understanding of existing perceptions, and tailored accordingly. For example, the survey results indicate stark differences in the views of some race groups towards safety. The strategy would also need to be ongoing, dynamic and must evolve as perceptions change. This implies regular monitoring of public perceptions about crime and safety.

Apart from the important public component of the strategy, attention would also need to be given to working with other entities that influence public opinion such as the media, other government departments and civil society organisations involved in crime prevention, such as Business Against Crime.

Taking responsibility for crime

According to the results, the public do not believe that most criminals are motivated by need. They also do not believe that most crime is committed by ‘outsiders’ to their areas or by foreigners. And many personally know someone who makes a living from crime in their area. All this suggests that most South Africans believe crime occurs because of choices that are made within our homes and communities, rather than because of external circumstances like poverty or illegal immigration. This information can be used by those in positions of influence to encourage people to be less accepting of criminals, to take personal responsibility for reducing crime, work with the police and courts, and ultimately reverse the widespread acceptance of crime as a part of everyday life.

Focusing on metropolitan and urban areas

The results across several questions, ranging from whether crime has increased to how safe people feel, indicate that concerns about crime and the impact of crime are much more pressing in urban areas generally speaking, than in rural areas. The views of those in rural areas should not be ignored or negated, however. But given limited resources and the need to prioritise interventions, efforts to reduce the fear of crime should focus on urban areas.

Deepening our understanding of perceptions

In order to respond effectively to declining feelings of safety, these perceptions need to be properly understood. While the survey provides useful insights into the broad trends, some specific aspects need clarification. The most obvious example is the consistent trend among the race groups for whites and (especially) Indians to be much more worried about crime and safety than coloureds and blacks. This trend has been identified by other surveys, and although there is now no doubt that the trend exists, little has been done to explain it. Further qualitative research is needed to properly understand these views and to hopefully provide policy makers with practical ways of responding to them.