Crime Trends in South Africa 1985-1998

Schönteich, M. & Louw, A. (1999). Crime Trends in South Africa 1985-1998. Paper commissioned by the Centre for the Study of Violence and Reconciliation as part of a review of the National Crime Prevention Strategy carried out for the Department of Safety and Security, June.

 

Paper commissioned by the Centre for the Study of Violence and Reconciliation as part of a review of the National Crime Prevention Strategy (NCPS) carried out for the Department of Safety and Security, June 1999.

Martin Schönteich is a former Researcher at the Institute for Security Studies (ISS).

Antoinette Louw is head of the Crime and Justice Programme (CJP) at the Institute for Security Studies.

Crime Trends: 1985-1998

Summary

• Serious crime has been increasing consistently since the mid-1980s. The public perception of a 'crime wave' in recent years is not supported by the statistics.

• The above trend applies to both violent and property crimes.

• Real crime trends can best be discerned using data over a lengthy period of time. Analysis of statistics over the past 15 years suggests that the so-called 'stabilisation' in crime since transition in 1994 (based on only four years worth of data) may be a short term phenomenon only: crime levels may continue to increase as they have done for the past 10-15 years.

• Of the offences reported to the police, robbery is the only crime trend that differs significantly from the average: more specifically, hijacking and robbery without aggravating circumstances (such as mugging) have increased much faster than crime levels generally. It is possibly that this trend (involving public, violent and personal crimes) that has contributed to an increase in fear of crime in the past four years as well as the perception of a 'crime wave'.

• The proportion of violent crimes, as a proportion of all crime reported to the police, is not increasing. However violent crime is increasing at a faster rate than property crime.

Overall Crime Trends

Levels of serious crime have been increasing fairly consistently since the mid 1980s. There was a surge in the number of reported crimes in 1993/94. This surge was, however, primarily because of a statistical quirk. In 1994 the ten 'black homelands' with a combined population of 17.8 million were reincorporated into the rest of the country. Crimes committed in the homelands territories were captured by the South African Police Service's statistical net for the first time only then.

Violent Crime

Between 1994 and 1997 levels of reported violent crime,¹ measured as a ratio per 100 000 of the population, stabilised and even declined. The exception was rape and robbery without aggravating circumstances, which increased significantly over this period.

During 1998 the levels of several serious violent crimes increased, however:

• Between 1997 and 1998 the number of reported attempted murders and robberies with aggravating circumstances increased.
• The consistent decrease in the number of reported murders that had been underway since 1994 was halted during 1998.

The only violent crimes that decreased between 1997 and 1998 were rape and assault.

It is possible, therefore, that the decrease in levels of violent crime between 1994 and 1997 was a short term phenomenon only, and that crime levels will continue increasing as they have been for the last 10 to 15 years. (While levels of crime have been increasing since the mid 1980s, there have been years when the number of reported crime decreased. For example, between 1975 and 1995 the national murder rate dropped on seven separate occasions. Even so, murder increased by 119% over the whole period. Rape dropped five times but still increased by 150%. Robbery dropped eight times but increased by 171%.)

Analysis of violent crime trends on a monthly basis shows clear cyclical trends which describe the nature of most violent crime in South Africa: levels of murder, rape, and assault GBH, are at their highest over the Christmas and Easter holiday periods. Moreover, according to victim surveys conducted by the Institute for Security Studies (ISS) in 1997, violent crime – most notably murder and assault – are more likely to occur on Fridays and Saturdays than on any other day of the week. Explanations for these cycles of violent crime point to increased alcohol consumption during holiday periods.

Another important factor (although like alcohol, one that remains untested) is domestic and acquaintance violence. The national Victims of Crime Survey conducted by Statistics South Africa and the Department for Safety and Security in 1997 shows that 77% of the victims of assault, and 75% of sexual assault victims, knew their offenders. The figure for murder is almost as high.

Property Crime

The SAPS Crime Information Analysis Centre (CIAC) calculates crime levels for all crime types as the number of reported crimes as a ratio to 100 000 of the population. This is not always useful in measuring changing levels of crime – especially property related crime. To understand and respond to crimes aimed at property – whether homes, cars, or cattle – ratios should be calculated according to the number of units of property in use in the country. Failing that, the raw numbers are preferable.

Various categories of reported property related crimes,² measured as a ratio to the population, decreased between 1994 and 1997 (housebreaking, theft of motor vehicle, theft out of motor vehicle, and other forms of theft). However, between 1997 and 1998 the level of these crimes increased. Only stock-theft and shoplifting decreased consistently between 1994 and 1998.

Violent vs Property Crime

The 20 most common serious crimes reported to police increased by 7.8% between 1994 and 1998 (from 1.9 million in 1994 to 2.2 million in 1998). (The population increased by 9.8% over the same period.)

The number of reported violent crimes, as a proportion of all crimes reported, increased only marginally between 1994 and 1998. However, violent crimes are increasing at a slightly faster rate than property crimes: between 1994 and 1998 the number of reported violent crimes increased by 11.2%. The number of reported property related crimes increased by 9.2%.

Despite the fact that crime has not necessarily become more violent between 1994 and 1998, South African's felt increasingly concerned about their own safety during this period. According to the Human Sciences Research Council (HSRC), in 1994, 73% of South African's felt safe and 16% unsafe. At the end of 1998 some 45% felt safe, and 48% unsafe.

This decreasing feeling of safety might be explained by the fact that robbery and car hijacking have increased substantially over the last few years. The number of reported robberies without aggravating circumstances increased by 91.6% between 1994 and 1998. Reported carjackings increased by 126% between 1995 and 1998. (The average for all crime during this period was 7.8%.)

Of all violent crimes covered in the ISS city victim surveys, people were most at risk of robbery (defined largely as crimes such as mugging). In fact, robbery was the second most prevalent of all crime types in all the metropolitan areas surveyed. Car hijacking, although less likely to affect a broad cross section of the population, is extremely violent and receives high media coverage.

Trends in the NCPS' Priority Crimes

Violence Associated with Inter-Group Conflict
Taxi violence

The incidence of taxi related violence decreased between 1996 and 1997, but increased between 1997 and 1998. In 1996 there were 557 reported cases of taxi violence (in which 285 persons were killed). This declined to 339 incidents (218 killed) in 1997, and increased to 667 incidents (291 killed) during 1998.

KwaZulu-Natal

Political fatalities in KwaZulu-Natal have declined substantially since 1993 (1 489 fatalities) when political fatalities in the province were at their highest. The biggest drop occurred between 1994 and 1995 as the election related violence fell away, and the IFP and the ANC leadership pushed for the peaceful resolution of political conflicts in the province. There was a slight increase in the number of fatalities between 1997 (226 fatalities), and 1998 (250 fatalities), primarily because of the conflict around Richmond in the KwaZulu-Natal midlands. However, during 1998 the number of fatalities peaked at 56 in July, whereafter the number of monthly fatalities declined considerably to 17 each in both November and December.

Organised Crime

Organised crime tends to grow rapidly during periods of political transition, when levels of violence are high, leading to state resources being concentrated in certain areas only and gaps emerging in which organised criminal groups may operate. Inadequate border controls, a good transport infrastructure, markets for and sources of contraband such as endangered species products and cannabis, a good banking system, and a supply of trained recruits have contributed to the growth of organised crime in South Africa.

In the first half of 1997, the CIAC identified 192 organised crime syndicates comprising 1 903 primary suspects operating in South Africa. The CIAC also identified some 500 target groups with a combined figure of 1 184 primary suspects. (Target groups being organised crime groups which were under investigation but whose structure and activities had not been fully identified.) Given South Africa's poorly resourced crime intelligence structures, and the consequent difficulty the police have in identifying organised crime groups, the real figures are likely to be higher. The CIAC is busy compiling an updated organised crime threat analysis to be released during the second half of 1999. Early indications are that the prevalence of organised crime activity has increased since 1997.

Criminal gangs, whose members are often the foot soldiers for organised crime groups, are widespread throughout South Africa. A survey conducted by Market Research Africa in November 1997 revealed that 61% of urban adults (74% for black adults) were aware of gang activity in their neighbourhoods. The highest proportion of gang related violence occurred in the Eastern Cape, Western Cape, and KwaZulu-Natal where 18% of respondents had been victims of gang activity.

White Collar Crime

The number of crimes reported to the police's commercial branch increased from 51 117 in 1995 to 59 515 in 1998 (an increase of 16.4%). The increase between 1997 and 1998 was, however, only 1.5%. The value involved in the reported cases increased from 2.8 billion in 1995 to 4.6 billion in 1998 (an increase of 64.7%).

Most cases of commercial crime are reported in Gauteng and the Western Cape (42.8% and 14.7% respectively, during 1998).

The most common type of crime reported to the SAPS' commercial crime is fraud (84.1% of cases in 1998), and forgery and uttering (7.1% in 1998).

Corruption Within the Criminal Justice System

There are no overall figures available on corruption in the departments comprising the criminal justice system. Moreover, as with private sector corruption, it is likely that the majority of incidents of corruption taking place within the criminal justice system are not reported and do not register in the available statistics.

From the available information it is, however, possible to conclude that incidents of corruption in the criminal justice system are on the increase. (Care should, however, be taken not to read too much into these trends. The departments comprising the criminal justice system were largely closed to public scrutiny before 1994, and bodies such as the statutory Independent Complaints Directorate (ICD) did also not exist then. Such bodies and greater transparency on the side of the criminal justice system have probably contributed to a greater awareness and reporting of corruption.)

Between 1996 and 1998, criminal prosecutions of cases involving collusion or corruption of police members in relation to escapes from police holding cells more than doubled from 91 cases in 1996 to 195 in 1998. Moreover, the number of police officers suspended nationally as a result of involvement in escapes increased from 16 in 1996, to 108 in 1998 (an increase of 575%).

Vehicle Theft and Hijacking

The theft of motor vehicles declined consistently from 1994 to 1996, and then increased during 1997 and 1998. Some 104 303 vehicles were reported stolen during 1994, by 1996 this had decreased to 96 715 (a decline of 7.3%). Between 1996/97 there was an increase of 4%, between 1997/98 an increase of 6.8% (to 107 513 vehicles).

The incidence of car hijackings reported to the police has increased consistently since 1995 (the first year in which the CIAC kept separate figures for this type of crime). In 1995 some 6 683 carjackings were reported. This increased to 12 860 (1996), 13 011 (1997), and 15 111 (1998).

While carjackings increased by 126% from 1995 to 1998, the reported incidence of the hijacking of trucks increased by 241% over the same period. Namely, from 1 695 reported incidents in 1995, to 5 773 reported incidents in 1998. Of the twenty most common serious crimes reported to the police, the hijacking of trucks has experienced the highest increase since 1995. Reported carjackings experienced the second highest increase over the same period.

Crime Involving Firearms

The proportion of crimes involving firearms has increased at a rate greater than the increase in general crime over the last few years. However, not all crimes are experiencing an increased use of firearms. It increased for murder, but decreased for attempted murder. Moreover, there was a slight increase for the crimes of robbery and the illegal possession of a firearm.

• While the number of murders reported since 1994 has been declining, the number of murders committed with a firearm increased. In 1998 some 12 267 out of 24 875 (or 49.3%) murders were committed with a firearm. This is up from 41.5% in 1994 and 1995, 44.2% in 1996, and 45.6% in 1997.

• The proportion of reported attempted murders committed with a firearm decreased, however, from 86.7% in 1994, to 74.8% in 1998.

• The proportion of robberies committed with a firearm increased slightly between 1996 (75.8%), and 1998 (78.7%).

• The number of incidents of the illegal possession of a firearm or ammunition increased from 13 413 in 1996 to 14 554 in 1998 – an increase of 8.5%. Reported cases of the negligent handling of a firearm increased by 16.7% between 1996 and 1998.

• The incidence of reported theft of firearm increased from 7 285 in 1994 to 11 391 in 1998 – an increase of 56.4%. The negligent loss of a firearm increased by 31.8% between 1996 and 1998.

• The SAPS recovered 20 682 illegal firearms during 1998, up from 14 460 in 1994. However, the number of illegal rifles recovered by the police declined by 20.8% between 1994 and 1998. Over the same period the number of illegal pistols recovered increased by 113.1%.

Proposals to Improve Crime Information Gathering Capabilities

To develop an effective crime prevention strategy it is crucial that policy makers and practitioners have access to up to date and relevant statistics concerning crime.

Crime results from a combination of many factors. Whether or not a crime is committed usually depends on three factors: the characteristics of the offender, the victim, and the 'environment' where the crime is committed. The environment includes the physical location as well as the people or activities in that area that might deter or encourage the offender in his actions.

These information requirements apply equally to national crime prevention programmes as they do to community based projects. A key difference for the NCPS – and an important challenge – is finding this information at the national level. There are many focused studies, case studies, pilot projects underway which provide some of this data. However if the NCPS is to fulfil its stated functions (and leave community based projects to local government and partnerships) data will need to be sought at the national level.

Adequate information is required on:

The nature of crime (in order to understand the problems and identify the right solutions)

• the victim (race, age, sex, occupation, type and location of home or business, type of stolen property)
• the offender (age, race, sex, distance travelled to commit the offence, previous criminal history and probable motivation for committing the crime)
• the environment (where and when the crime occurred)
• how crimes are committed
• involvement of factors like alcohol and drugs
• when crime occurs

The amount of crime (in order to prioritise the most serious problems and target communities)

• when crime occurs
• crime rates (to compare geographic areas and select worst affected places)
• risk of crime (identify communities and individuals most at risk)
• impact of crime

In theory the SAPS collects most of this information through its case dockets. The kind of crime committed, the location of the crime, some offender and victim details, the modus operandi used, and the time the crime occurred, could all to be found in properly completed case dockets. The problem is that little of this information is collected. And when it is, the nature of the SAPS CAS system makes it difficult to access and analyse to produce the data as outlined in the bullet points above.

At the same time, not all the required information could be supplied by the police, no matter how effective their data collection and processing system. Information on victims, and in particular victim profiles, crime rates, risk of crime and impact of crime cannot be supplied by the police. Victim surveys and similar in-depth studies are more appropriate sources.

Three key priorities for the NCPS should be:

1. Improving the existing SAPS crime information systems

Rather than developing complicated new data collection systems, the police should improve what they already have. The committee of inquiry into the collection, processing and dissemination of crime statistics completed n March 1998 and commissioned by the minister for safety and security compiled detailed recommendations in this regard. The committee found that the CAS system was potentially a useful one. The main problem is poor management of the collection of information at all levels, and very little priority within the SAPS to crime information and its uses for prevention.

Changes are needed both to the computer systems and in respect of training of police officials concerned with information gathering and processing. Detectives should be trained to put as much available information into every docket they work on. Resources need to be dedicated to improving the accessibility of what is essentially a detailed and more than adequate database. By properly using what it already has, the police would be in a position to substantially improve its knowledge on offenders, victims, and the environment in which crimes occur.

2. Fill the gaps left by an improved police database

Information from police sources is always limited with respect to certain crime types: crimes committed against women and children, domestic violence, crimes involving juveniles (both as victims and offenders), commercial crime and corruption within the criminal justice system.

Of these, probably the most critical information gaps in the development of crime prevention programmes in South Africa is information about juveniles and information about offenders. Young people are most likely to be victimised and make up most of the offenders. Also, crime prevention projects in other countries focus substantial attention on targeting offenders: without accurate offender profiles, this has been impossible so far.

Filling these gaps will require focused studies on each of these sectors. These could be in the form of victim surveys or more qualitative studies.

3. Focus on estimating risk for particular communities and individuals

Victim surveys are the most useful tool for calculating risk (and impact). In order to optimise these functions, the national victim survey should (in terms of design and use of the data) in future be amended in the following ways:

• Focus less on comparability between surveys and more on specific needs.

• Plan for, rather than avoid, comparisons with police statistics: one of the most important aims of victim surveys is to evaluate whether changes in crime levels over time are real or a product of crime reporting patterns by the public. To do this, crime definitions in the survey should closely match police definitions.

• Focus on providing data that police statistics cannot – rather than over complicating the survey and making it too long to make up for police inadequacies.

• Focus on calculating how risk varies: the British Crime Survey (BCS) for example has risk indicators based on:

  • where people live (inner city, urban and rural; on a main road, side road and cul-de-sac, in council estate and non council estate areas)
  • their financial resources (household income)
  • the structure of their household (age of head of household, number and relation of household head to children, household head employment status, tenure, accommodation type)
  • their personal lifestyle (hours home unoccupied on an average weekday)

 

• Factors associated with risk in South Africa are likely to be similar. Sampling frames and questionnaires should be adapted to allow the collection and analysis of information which enables compiling risk profiles that can be acted upon.

Sources Used

SAPS Crime Information Analysis Centre (CIAC) Quarterly crime reports
South African Institute of Race Relations (SAIRR)
Insitute for Security Studies (ISS) victim surveys

Notes:

¹ Violent crime includes the following crimes as recorded by the SAPS: murder, attempted murder, rape & attempted rape, robbery with aggravating circumstances, other robbery, assault GBH and common assault.

² Property crime includes housebreaking business and residential, car theft, theft out of cars, shoplifting, stock theft and other theft.

© Centre for the Study of Violence and Reconciliation