Home |  Contact UsSitemap



Crime in South Africa: Perceptions, Trends and Possibilities

By Andrew Faull (Independent Researcher, ISS Consultant, UCT associate)



For almost a decade the South African criminal justice system has had as its broad goal, ‘South Africans are and feel safe’.1 Similarly, the National Development Plan (NDP) describes South Africa in 2030 as a place where people ‘feel safe at home, at school and at work,’ where life is ‘free of fear’ and ‘women walk’ and ‘children play safely outside’. To some this may appear a naïve wish. Narratives of crime and vulnerability are central to how we South Africans make sense our lives and country. Indeed, most participants interviewed for the first phase of the Indlulamithi Scenarios anticipated that crime would escalate from 2018 to 2030, but this need not be true.

In this brief review I draw on the two most common methods for counting crime2 – police statistics and victim surveys, to explore crime trends and related perceptions and feelings in South Africa. I show that contrary to popular belief most categories of crime have declined significantly since the early nineties, although some categories of violent crime have spiked. I end by exploring what these data, combined with recent evidence on violence and crime prevention, tell us about crime in South Africa’s near future.


Perceptions & Feelings about Crime and Safety

South Africans are concerned about the nature and prevalence of crime in their communities, and what they see as government’s failure to curb it. This is evident in various survey data. For example, a review of the Human Science Research Council’s (HSRC) South African Social Attitudes Survey (SASSA) data for the years 2003 to 2015 (Figure 1) shows that on average only 21% of respondents were satisfied with government’s efforts to reduce crime over the twelve year period.3 The only area of government performance judged more harshly was job creation, with which an average of just 10% of respondents were satisfied. Crime is also perceived as the second most important problem facing the country after unemployment, that government should address as measured by the Afrobarometer.4



Figure 2 shows the eight categories of crime in South Africa believed to be the most common in 2016/17, and which were the most feared, as measured by Statistics South Africa’s victim survey. As in previous years, burglary is both perceived as the most common crime (62%) and is the most feared (51%). In that the crime involves no confrontation and thus no physical violence between perpetrator and victim, this is counterintuitive. One interpretation of this offered by Shaw and Kriegler, is that respondents may confuse burglary with home robbery, the latter involving confrontation with intruders in one’s home and the use or threat of violence.5 This interpretation is supported by the fact that two other categories of robbery – inside and outside home - are the perceived as second and third most common and feared crimes.



Longitudinal perception data reveals something more interesting. Figure 3 shows that in 2016/17 39% of respondents believed violent crime had increased in their area of residence. This is alarming. And yet it is a marked improvement on the 2003 and 2007 surveys where 53% and 57% of respondents believed violent crime had increased.6 So perceptions have improved since2007, but they have also deteriorated. In 2011 and 2012 more respondents believed violent crime had decreased (43% and 38% respectively) than believed it increased. And while attitudes have altered slightly in the surveys since 2013, perceptions are significantly more positive today than they were ten to fifteen years ago. Very similar perceptions are found when respondents are asked about property crime, as illustrated in Figure 4. In 2016/17, 44% of respondents believed property crime had increased, 26% believed it had decreased and 30% thought it had stayed the same.




How do these beliefs shape the way South African residents live their lives? Not as much as one might think. Figure 5 shows that in 2016/17, 32% of respondents reported not spending time in public spaces because of crime, 20% would not let their children play outside, and 15% would not walk to town.7 These are not insubstantial numbers or claims. To feel unable to walk outside is not to live a just life. And yet the numbers represent only a minority of people, even if only because many don’t have the option not to use public transport or walk to the shop.



More revealing are measures of whether people feel safe walking alone in their area of residence during the day and at night. Figure 6 shows that from 1998 to 2016/17, and with the slight exceptions of 2007, there has been little change in the percentage of people who feel safe during the day, averaging 85%. This is important. Despite most people in most surveys believing crime to be increasing in their area of residence, feelings of safety during the day remain constant.

Where there may have been significant change in this data is in those who feel safe walking alone at night, dropping from 56% in 1998 to just 23% in 2003 and 2007 (corresponding with the perception data in Figures 3 and 4), then rising to 37% in 2011 and slowly dropping year on year to a low of 29% in 2016/17. So while people don’t feel significantly less safe today during the day than they did twenty years ago they feel increasingly less safe at night.



To summarise, more South Africans today and in recent years believe crime is getting worse than believe it is getting better or staying the same. The crimes believed to be most common, and which are most feared are burglary and robbery. Only a fifth of the country is satisfied with government’s efforts to tackle crime, and up to 30% of people feel unable to pursue various activities for fear of crime. Most people feel safe in their area of residence during the day, but increasingly fewer feel the same at night. In the next section I review key police data and trends and suggest that, while imperfect, they are broadly supported by the victim surveys.


Police recorded crime: fact or fraud?

Measuring crime is not a neat science. Who reports and measures what, when, how and who decides? The answers lie in bureaucracy, politics, power and economy,8 none of which I explore here. Suffice to say that what is reported to police represents specific rather than general experiences. For example, victims of car theft are far more likely to report their loss to police (94% in 2016/17) than are victims of housebreaking/burglary (51% in 2016/17).9 This is not because they believe police are more capable of retrieving cars than televisions, but because cars are much more expensive, likely represent a significant asset, and are thus likely to be insured. Thus, data on police recorded crime cannot be accepted at face value, but it should not be ignored either.

The most reliable police reported crime statistic is murder. Most people (95% in 2015/16) report murder to police and if they don’t, bodies usually turn up and the state is notified. Murder is also defined fairly, though not entirely consistently globally – the unlawful and intentional act of killing – making the category reasonably comparable across countries. The global murder rate in 2012 was 6.2 per 100,000 population but 31 per 100,000 in South Africa.10 In 2016/17 it was 34 per 100,000, roughly five times the global average.

Murder is a good proxy for other categories of violent crime. Where murder rates are high, the rates of other violent crime are also likely to be high. What does the murder data tell us about crime in South Africa? As illustrated in Figure 7, South Africa’s recorded murder rate has halved from a high of 69 in 1994/5 to 34 in 2016/17, having reached a low of 30 in 2011/12.11 While the decline is stunning, the uptick in recent years is concerning .



Source: South African Cities Network, 2017; SAPS 2017

Table 1 shows that, as predicted by the murder data, similarly impressive declines are found in attempted murder (-49%), assault GBH (-52%) and common assault (-49%) as reported to police.12 However, the democratic era has brought with it a surge in various categories of robbery, which continued to rise in 2016/17. The robbery data hints at a growing willingness for offenders to confront and threaten with or use violence against those whose property they wish to seize. I will return to this in the next section.



Some readers may question the validity of this data. As we saw in the previous section, the drastic fall in crime suggested by these figures is not reflected in feelings of safety measured in victim surveys. Might victims simply be opting out of the criminal justice system, failing to report crime to police? Figures 8 and 9 show that in recent years satisfaction with police and the rates at which victims report some, but not all of the most regularly reported categories of crime, have dropped by a few percentage points.




Does this mean that the fear and doubt recorded by victim surveys is justified? Is the data that shows declines in crime more indicative of public apathy than anything else? Were it not for the spectacular drop in murder, one might think so. That and what the victim surveys reveal, which I now return to.


Actual experiences of crime (victimisation rates)

One of the greatest contributions victim surveys make is to provide fairly rigorous estimated victimisation rates independent of crimes reported to police. This is usually achieved by asking respondents whether they or a member of their household has been a victim of crime in the preceding year (or another period), followed by numerous follow up questions, including ‘Did you report it to the police?’.14

Importantly, the data from South Africa’s seven victim surveys corroborates many of the trends found in police data. That is, significantly fewer people report being victims of crime today than did in 1998.

Figure 9 shows that in 2016/17, roughly 7% of households were affected by crime, compared to 25% in 1998, a 71% reduction. Similarly, 3.5% of individuals reported being victims of crime in 2016/17 compared to 4.7% in 2013/14, a 25.5% decline in the last four years (the older victim surveys did not distinguish between individuals and household victimisation in the same way). Held up against the perceptions of crime and feelings of fear sketched in the opening section, these figures should be welcomingly sobering: very few people are victims of crime each year.




Where to from here?

Whether we believe it or not, most categories of crime are declining in South Africa. It is, however, far harder to discern why this is than to explain what caused it in the first place. In their excellent book on the subject, Kriegler & Shaw suggest that there is no mystery to South Africa’s pervasive and violent crime:


Within the last century, South Africa has gone from an overwhelming rural to a majority urban profile. Within living memory, millions of South Africans had their lives entirely uprooted by being forcibly removed from their homes and relocated to areas far from jobs, with poor infrastructure, chronic overcrowding, insecure tenure and devastated system of community control and socialisation…[M]any families were drive by the demands of the labour market into precarious, migrant, divided lives. [More recently] some areas were…embroiled in…a low-grade civil war. For the vast majority of the country, relations with the state have until recently been characterised not by accountability or trust, but instead by humiliation, indignity and violent confrontation.15


They add that South Africa has followed ‘the recipe’ for a violent society ‘to the letter’ and added to it the availability of firearms, a culture of binge drinking, consumerist values, a home for organised crime and drugs.16

While they add more nuance than I include here, Kriegler and Shaw sketch three possible scenarios for South Africa’s near future. The first sees a continued rise in murder through a failure to address the above drivers of violence, combined with declining state legitimacy. Their second scenario sees violent crime remain at present levels, with increasingly more murder linked to robbery. Finally, they suggest that improved and legitimate policing, lawfulness and trust in the state, resources and interventions targeting crime hotspots, and the upliftment of young people, particularly men, could see violent crime continue to decline.17 These are scenarios supported by a broad, if incomplete evidence base, all of which is recognised in South Africa’s policy architecture.


The Policy, Practice and Leadership Gap

At an instinctive, emotional level most South Africans may think the solution to the country’s crime lies in the criminal justice system. But, unsurprisingly, offer people a choice of areas in which they think government should spend money to reduce crime and far more pick ‘economic development’ (61%) than ‘law enforcement’ (20%) or ‘the judiciary/courts’ (13%).18 Similarly, given a moment to think about it, many would agree that environments in which peoples’ lives are characterised by dignity, purpose and healthy relationships are less likely to be unsafe than those where life is unpredictable and precarious..

Should South Africa make gains in shaping the society imagined in the NDP, it is very likely that general crime and violence will decline over the next twelve years. To this end, the country is not short on evidence-informed vision or policy.19 Instead, it is curtailed by its inability to mend a broken primary and secondary education system or stimulate the labour market to end chronic unemployment, particularly among young people. With these addressed, one would expect to see improvements in other areas which evidence suggests are key to violence prevention. These include:20 a) cultivating safe, stable and nurturing relationships between children and parents; developing life skills in children and teenagers; c) reducing the availability and harmful use of alcohol; d) reducing access to dangerous weapons; e) promoting gender equality; f) changing cultural and social norms that support violence; g) supporting and caring for victims; h) promoting reason and rationality over intuition, ‘culture’ and superstition21; i) promoting good state governance and the rule of law, including by clamping down on corruption and organised crime, utilising order-promoting technologies and promoting police legitimacy.22

The apparent impunity with which key political leaders have abused their power and siphoned off state funds in recent years, predicts state illegitimacy and incompetence, and thus more crime. It should not be underestimated how significant the change in African National Congress (ANC) leadership in December 2017 and early 2018 was, and the potential it represents for a more peaceful South Africa, come 2030. Safety begins with good governance. It is thus in the appointment and performance of the new cabinet that South Africa is likely to find the best indication that 2030 will be safer than 2018.



1 See: SAPS (2010) South African Police Service Strategic Plan 2010-2015, SAPS: Pretoria; SAPS (2017) South African Police Service Strategic Plan 2017/18, Pretoria: SAPS.
2 Newburn, T (2017) Criminology, 3rd Edition. Abingdon & New York: Routledge, pp. 45
3 Democracy, Governance and Service Delivery (DGSD) Programme (2016) Service Delivery Challenges in South Africa, South African Social Attitudes Survey (SASAS) COMPENDIUM OF RESULTS. Human Sciences Research Council, p. 8
4 Afrobarometer, (2016) Summary of Results: Afrobarometer Round 6 Survey in South Africa, 2015, Johannesburg, South Africa. Available at: http://afrobarometer.org/sites/default/files/publications/Summary of results/saf-r6-sor.pdf., p.43
5 Kriegler, A. & Shaw, M. (2015) A Citizen’s Guide to Crime Trends in South Africa, Jonathan Ball: Cape Town.
6 ISS 2003 survey monograph, ISS 2007 victim survey
7 Stats SA, 2017
8 Reiner, R. (2016) Crime. Cambridge:PolityPress; Newburn, T. (2017) Criminology, Third Edition. New York: Routledge
9 Victims of Crime Survey 2016/17
10 UNODC (2013) Global Study on Homicide: 2013. Available at: https://www.unodc.org/documents/gsh/pdfs/2014_GLOBAL_HOMICIDE_BOOK_web.pdf (accessed17 February 2018)
11 Urban Safety Reference Group (2017) The State of Urban Safety in South Africa, p.18
12 Urban Safety Reference Group (2017), p. 20
13 Note, this is ‘robbery with aggravating circumstances’, slide 44 of SAPS 2017 crime stats presentation
14 Newburn, T. (2017) p. 56
15 Kriegler & Shaw (2015), loc:1552
16 Ibid. loc:1566
17 Ibid. loc:3229
18 Statistics South Africa (2016) Victims of Crime Survey 2015/16, p. 24
19 For insight into the policy framework see: Kruger, T., Lancaster, L., Landman, K., Liebermann, S., Louw, A., Robertshaw, R. (2016) Making South Africa Safe: A Manual for Community-based Crime Prevention, Pretoria: CSIR; Urban Safety Reference Group (2017) The State of Urban Safety in South Africa: Report. South African Cities Network: Johannesburg.
20 World Health Organisation (2010) Violence Prevention: The Evidence. Available at: http://www.who.int/violence_injury_prevention/violence/4th_milestones_meeting/evidence_briefings_all.pdf (accessed 22 February 2018)
21 Pinker, S. (2011) ‘Decline of violence: taming the devil within us.’ Nature 478(7369), pp. 309-311;
22 Eisner, M. (2015) How to reduce homicide by 50% in the next 30 years, Igarapé Institute. Available at: http://homicide.igarape.org.br/ (accessed 18 February 2018);


Share content with FFD



Features Archive


new-sampnode-logo rockefeller-logo-footer-new

Foresight For Development - Funding for this uniquely African foresight site was generously provided by Rockefeller Foundation. Email Us | Creative Commons Deed | Terms of Conditions