[Home] [Databases] [WorldLII] [Search] [Feedback] South African Law Commission

You are here:  SAFLII >> Databases >> South African Law Commission >> Discussion Paper >> 97 >> CHAPTER FIVE

[Database Search] [Name Search] [Previous] [Next] [Download] [Help]

• Case Study Analysis Of Police Dockets
• 5.1 Introduction
• 5.2 Methodology
• 5.2.1 Scope Of The Study
• 5.2.2 Method Of Data Collection
• 5.2.3 Problems With The Data
• 5.3 Findings60, 61
• 5.3.1 Number Of Cases Analysed
• 5.3.2 Categorisation Of Cases Constituting The Study Sample
• 5.3.3 Description Of The Crime Incidents Studied
• 5.3.4 Gender Profile Of Victims
• 5.3.5 Racial Profiles Of Victims
• 5.3.6 Age Of Victims
• 5.3.6.1 Victims Younger Than 18 Years
• 5.3.6.2 Victims Between 18-24 Years
• 5.3.6.3 Victims Between 25-34 Years
• 5.3.6.4 Victims Between 35-44 Years
• 5.3.6.5 Victims Between 45-54 Years
• 5.3.6.6 Victims Between 55-64 Years
• 5.3.6.7 Victims Older Than 65 Years
• 5.3.6.8 Age Unknown/not Recorded.
• 5.3.7 Employment Status Of Victims
• 5.3.7.1 African Victims
• 5.3.7.2 White Victims
• 5.3.7.3 ‘Coloured’ And Indian Victims
• 5.3.8 Place Where Crime Took Place
• 5.3.9 Relationship Between Victim And Perpetrator
• 5.3.10 Gender Of Perpetrator
• 5.3.11 Victims’ Role In Their Own Injury
• 5.3.12 Weapons Used By Perpetrator/s
• 5.3.13 Type Of Injuries Sustained By Victims
• 5.3.14 Hospitalisation Of Injured Victims
• 5.3.15 Loss And Damage To Property
• 5.3.16 Police Progress On Investigation Of Cases
• 5.4 Analysis And Discussion
• 5.4.1 Defining Compensable Victims
• 5.4.1.1 Intentional Or Deliberate Violent Crimes
• 5.4.1.2 Injury Sustained
• 5.4.1.3 Prosecution And Punishment Of Offender
• 5.4.1.4 Citizen Versus Foreign Nationals
• 5.4.1.5 Financial Hardship Of Victims
• 5.4.1.6 Conduct Before/during/after The Incident
• 5.4.1.7 Victim’s Co-Operation With The Criminal Justice Process
• 5.4.1.8 Victims Of Domestic Violence
• 5.4.1.9 Defining The Seriousness Of The Injury
• 5.5 Conclusion

CHAPTER FIVE

Case Study Analysis of Police Dockets

This chapter contains the results of an analysis of selected police dockets at Randburg and Mamelodi police stations in Gauteng, South Africa. The analysis provides information about certain types of violent crimes and their impact on victims. This information is useful for making assumptions when costing a VCS (see Chapter Six) and for shaping possible policy scenarios. The docket analysis undertaken in this chapter also focuses on the usefulness of police information in adjudicating possible claims for victim compensation.

5.1 Introduction

Police dockets are the case files containing all relevant information about a recorded criminal case. Police dockets generally include basic facts and demographic information about the incident, statements by victims and witnesses, details of the activities undertaken by the police officers dealing with the case, and progress of the case through the criminal justice system.

Docket analysis can provide some useful information about the nature of violent crime (and the responses of the criminal justice system to it), although, as will be shown below, the quality of the information contained in the dockets fundamentally influences the usefulness of the analysis. The purpose of the docket analysis undertaken for this report, however, was to provide detailed information about certain types of violent crimes; and to assess the usefulness of police information in adjudicating possible claims for victim compensation. This information was required to assist in quantifying the financial impact of a potential victim compensation scheme in terms of possible policy scenarios.

5.2 Methodology

5.2.1 Scope of the study

Police dockets were analysed at two SAPS stations in Gauteng province, namely Randburg and Mamelodi.

Mamelodi is a former African township on the eastern side of Pretoria. It encompasses densely populated urban township areas, as well as some peri-urban areas. Randburg is a suburb in the north of Greater Johannesburg. It contains a new central business district (CBD), a number of large retail and entertainment precincts, informal settlements and formerly White suburban residential neighbourhoods.

The rationale for selecting these two stations was that:

• both cover large station areas (jurisdictions) and would therefore provide us with a substantial number of cases to analyse;

• they cover a range of different types of locales, i.e. suburban residential neighbourhoods, small CBDs, informal settlements, business and light industrial precincts, township residential neighbourhoods, and peri-urban areas;

• urban areas have the highest rates of violent victimisation – approximately 4% of households in urban formal areas, and 3% of households in urban informal areas experienced at least one violent crime in 1997, as opposed to 2% of households in non-urban and traditional areas (Statistics South Africa Victims of Crime Survey Pretoria 1998), which makes it prudent, then, when estimating volume and cost of criminal injury and compensation, to base such estimates on data gathered in urban areas;

• these two stations were easily accessible to researchers within the budget of the project.

While the rate of victimisation of residents of Gauteng is roughly similar to the average rate of victimisation in the country as a whole, (Statistics South Africa, 1998, p38); the sample of crime trends in Mamelodi and Randburg should not be taken as representative of South Africa as a whole, because they are largely urban areas and reflect only trends in the metropolitan heartland of the country. However, the analysis is suggestive of trends elsewhere. Our assessment of police treatment of the cases, and the quality of dockets, however, is probably fairly representative of national standards.

The following four crime types were studied:

• Murder

• Attempted Murder

• Assault GBH (with intent to do grievous bodily harm): this generally refers to serious assault, involving knives or firearms

• Aggravated Robbery: this generally refers to robberies involving knives or firearms, such as armed robberies and hijackings of motor vehicles.

These four crime types represent the most serious violent crimes to which the researchers could obtain access (there are legal problems and police concerns about allowing researchers access to rape and indecent assault dockets, which is why these cases were not examined). The selection of these crime categories for analysis was based on the assumption that these represent the most likely type of cases in which victim compensation would be sought, as they are generally considered the most serious crimes.

Closed police dockets concerning crimes that had been reported in the three months April-June 1998 were studied^[54]. The selection of this period was based on three main reasons.

First, police docket management practice in 1998 was similar to present docket management practice, and is, therefore, a reliable basis for analysis and projections concerning police dockets and their usefulness in a victim compensation scheme.

Second, if we had selected a more recent period, there would probably have been access to very few ‘closed’ dockets; as it takes the police many months (sometimes years) to close an investigation, particularly in cases of serious violent crime.

Third, there are well-known patterns in the reporting of violent crime (Nedcor/Institute for Security Studies Crime Index Vol 3 No 4 Number 4 Pretoria: Institute for Security Studies 1999). For instance, significant increases are visible at month-end and over holiday periods. We did not want the data to be too skewed by this; and also we wanted to capture a sense of the impact of holiday periods on levels of victimisation. We chose a period, therefore, which includes the April holiday season as well as the beginning of winter, i.e., holiday periods and some ‘ordinary’ weeks. We found that most of the incidents studied took place over weekends, i.e., 26% on Saturdays and 19% on Sundays. Most incidents studied took place in the early part of the evening, largely between 19h00 and 21h00.

5.2.2 Method of data collection

A data-gathering form was designed to enable capturing of relevant information. This form contained 26 fields of information, which could possibly be captured (see Appendix Five).

Permission to access the dockets was granted by the Research Component of the SAPS at Head Office and the office of the Provincial Commissioner of the SAPS (Gauteng),^[55] who instructed the officials at the Randburg and Mamelodi stations to assist the researchers.

Data capture at the stations consisted of the researchers^[56] sitting in the stations’ docket stores and reading through the relevant dockets. At Mamelodi,^[57] the docket clerks assisted by drawing the specific cases (murder, attempted murder, assault GBH and aggravated robbery) and handing them to researchers.^[58] At Randburg,^[59] the docket clerk handed the full set of dockets for the months of April-June 1998 to the researchers, who then drew out the specific cases for analysis.

The data captured on the forms were cleaned and entered into a statistical programme for analysis.

5.2.3 Problems with the data

Various problems were experienced in the course of the data capture at the police stations. They included:

• illegible handwriting by the police officers who completed the various forms and statements in the docket;

• incomplete forms, sometimes with entire sections not completed;

• the lack of data on injury.

The last point above relates to the fact that the SAPS forms and statements often do not contain any information about the nature of injuries sustained by the victim of the crime. Also, only a small proportion of dockets contained a J88 form completed by a medical doctor attesting to injuries sustained.

It needs to be borne in mind that many violent crimes are not reported to the police and, therefore, that any sample of police dockets cannot be taken to represent trends in the overall crime pattern in South Africa. The National Victim Survey found that only 60% of hijackings and attempted hijackings were reported, and only 83% of murders (Statistics South Africa, 1998, at 53). It also found that most individuals were unlikely to report assault (38% of cases were reported, 62% not) or armed robbery cases (41% of cases were reported to the police, 59% not - Statistics South Africa, 1998, at 57). This means that our sample of reported cases is unlikely to represent the total picture of victimisation in these four types of crime. If compensation for such cases was available, levels of reporting might increase if victims saw reporting as a method of obtaining access to compensation.

5.3 Findings^{[60], [61]}

5.3.1 Number of cases analysed

• A total of 522 dockets were analysed, representing all the available closed dockets at the two sites for crimes recorded in the months of April, May and June 1998.

• The low number of murder cases is due primarily to the fact that murder dockets are not generally closed until some years after their opening, thus making fewer available for research purposes.

5.3.2 Categorisation of cases constituting the study sample

This reflects the description of the cases according to the SAPS’s system for categorising crime incidents, in line with the Criminal Procedure Act and the Crime Code. The SAPS’s crime code categories are not very useful, in that they do not disaggregate between different types of incidents. For example, ‘assault GBH’ may describe a domestic violence incident or a bar brawl or a racist attack. These are all very different types of crimes.

• Two thirds of dockets received from Randburg related to aggravated robbery, while the greatest proportion (59%) of reported incidents at Mamelodi, were assault GBH.
• Overall, assault GBH constituted almost half of the cases analysed.

5.3.3 Description of the crime incidents studied

This reflects the description of the cases according to categories constructed for purposes of the analysis required in this study, based on our experience and analysis of violence in South Africa in general. The categories are labelled and described below.

• ‘Domestic or family violence’ refers to any intentional violence between members of the same immediate family, which includes only parents, children, spouses and intimate partners.^[62]
• ‘Violence between people who know each other’ refers to all violence between people who know each other, extended family members, friends or acquaintances.
• ‘Hijacking’ refers to actual or attempted hijacking of a motor vehicle.
• ‘Robbery’ refers to all robberies other than hijackings, including armed robberies at homes or businesses, street robberies and bank robberies.
• ‘Attack by stranger’ refers to any attacks where there is no apparent knowledge of the attacker and where robbery is not the motive.
• ‘Other’ refers to any other situations not described in one of the above categories, or where information was not available in the docket.

• Of the dockets studied at Mamelodi, half were violent crimes where the victim knew the assailant. This is as opposed to one in five such cases in Randburg.
• Aggravated robbery was the most frequently found violent crime in Randburg, as opposed to assaults (by strangers or by known persons) in Mamelodi^[63]. Our findings are consistent with the National Victim Survey which found that Africans were more likely to be victims of violent crime than property crime, and, conversely, that Whites were more than twice as likely to be victims of property crime than of violent crime (Statistics South Africa, 1998, at 39).

5.3.4 Gender profile of victims

• At both Randburg and Mamelodi, men were the majority of the victims of the violent crimes studied, with this being the case for 3 in 4 cases studied in Randburg, and 2 in 3 for Mamelodi.

• In Randburg, both men (70%) and women (55%) were more commonly the victims of aggravated robbery than of any other of the violent crime categories studied.
• By contrast, in Mamelodi, both men (48%) and women (79%) were more commonly the victims of assault GBH than of any of the other violent crime categories studied.
• In both sites, women were more likely than men to be the victims of assault GBH as reported to the police and contained in our sample. This is probably due to increases in reporting levels of violence against women in the post-1994 period, as a result of new government policies, as most forms of domestic violence are classified by the SAPS as ‘assault’. This is borne out by the analysis in the following table:

• In the Mamelodi cases, 68% of victims of domestic/family violence were women, while 75% of victims of domestic violence in Randburg were women. At both sites, domestic or family violence was the most common type of victimisation reported by the female victims in our sample.
• In both sites, men were the majority victims of the other types of crime, i.e., violence among acquaintances, hijacking, robbery and attacks by strangers.

5.3.5 Racial Profiles of Victims

• In both Mamelodi and Randburg, Africans were the majority of victims. African victims accounted for nearly 80% of all cases in our sample.
• Of all the victims of crime in our Randburg sample, more than half were African, and 43% White. The figures were similar for both men and women^[64]. Unsurprisingly, given the demographics of Mamelodi, nearly all the victims in that sample, both men and women, were African.

• African victims in our sample were more likely to be the victims of assault GBH than of any other type of crime, while White victims were more likely to be victims of aggravated robbery than any other crime.

• Assault GBH constituted 46% of all the cases studied, and the majority of all assaults were committed on African victims.

5.3.6 Age of victims

• People between the ages of 25-34 were the most common victims (38%) in our sample at both sites, while those over the age of 55 and under 18 were least likely to be victims of the violent crimes studied. This is in line with the findings of the National Victim Survey, which found that ‘of all individuals who had experienced at least one violent crime in 1997, almost a third (31.5%) were aged 16-35 (Nedcor / ISS, 1999, at 26 – 27). For both violent and property crime, the level of victimisation consistently declines with age.
• These findings correlate with the Pretoria^[65] Victim Survey (A Louw Crime in Pretoria: Results of a City Victim Survey Pretoria: Institute for Security Studies 1998, at 36) which found that just under half of the victims of violence (46%) in the city were aged between 26-40 years. In Johannesburg^[66], the equivalent data (ISS, 1997, at 15) suggest that most crime happens to men between the ages of 25-60 years
• The next most common age range of victims in our sample is between 35-44 years (26%).

A cross-tabulation of types of crimes with age and race of victims found the following:

5.3.6.1 Victims younger than 18 years

• The largest proportion (77%) of victims younger than 18 years old were the victims of assault GBH.
• This age group accounts for 3% of all victims studied.
• Of the two recorded cases where children under 18 were victims of an attempted murder, one was White, the other African.
• Of the 10 recorded cases where the child was the victim of assault GBH, all the victims were African. In the solitary case of aggravated robbery within this age category, the victim was also African.

5.3.6.2 Victims between 18-24 years

• The largest proportion (58%) of victims within this age category, were the victims of assault GBH.
• This age group accounts for 15% of all victims studied.
• 32% of aggravated robbery victims in this age group were White.

5.3.6.3 Victims between 25-34 years

• The largest proportion (48%) of victims within this age category were the victims of assault GBH.
• This age group accounts for 38% of all victims studied.
• 34% of people in this group were victims of aggravated robbery.
• 71% of the victims were African, and 26% White.
• Of the 32 cases of attempted murder within this age group, 90% of victims were African, and 7% were White.

5.3.6.4 Victims between 35-44 years

• Half the victims within this age category (49%) were the victims of aggravated robbery.
• This age group accounts for 26% of all victims studied..
• 61% of aggravated robbery victims studied were African, and 36% were White.

5.3.6.5 Victims between 45-54 years

• The largest proportion (44%) of victims within this age category were the victims of aggravated robbery.
• This age group accounts for 12% of all victims studied.
• 52% of all aggravated robbery cases within this age group occurred among White victims, with 44% of victims of this crime being Black.
• 75% of all attempted murder victims within this age group were African.

5.3.6.6 Victims between 55-64 years

• The largest proportion (63%) of victims within this age category were the victims of aggravated robbery.
• This age group accounts for 4% of all victims studied.
• 75% of aggravated robbery victims studied were White, and 25% Black.
• Of the 5 cases of assault GBH reported within this age group, all victims were African.

5.3.6.7 Victims older than 65 years

• Only two cases fell within this age group.

5.3.6.8 Age unknown/not recorded.

• In 3% of the cases, the age of the victim was either not known, or not recorded. Four of these were victims of attempted murder, 6 of assault GBH, and 4 of aggravated robbery.

5.3.7 Employment status of victims

The employment status of victims is relevant to the assessment of potential loss of earnings for compensation. The exact details of employment are not captured in the police docket, so researchers were required to make deductions from the available information in the docket^[67]. The categories used are listed in the left-hand column below.

• Unemployed people and blue-collar workers were the most likely to be victims of crime in the Mamelodi sample, while the most common victims in the Randburg sample were blue-collar workers.
• These findings are borne out by the findings of the National Victim Survey, which found that those in the lowest income category (household income of under R3000 per annum) were the most susceptible to violent crime in 1997 (Statistics South Africa, 1998, at 41).

When cross-tabulating the race of victims with their employment status, the following was found:

5.3.7.1 African victims

• Amongst African victims, 75% of those younger than 18 were students/scholars.
• Almost half the African victims aged between 18-24 were unemployed.
• Half the African victims aged between 25-34 were blue-collar workers, while this employment category contained half the African victims aged between 35-44, and 44% of the African victims aged between 45-54.
• The greatest percentage (60%) of the victims aged between 55-64 were unemployed.

5.3.7.2 White victims

• Half the White victims aged 18-24 were white-collar workers, while this category contained 45% of those aged between 25-34.
• 38% of White victims aged between 35-44 were white-collar workers and this was the employment category of 66% of the White victims aged between 45-54.
• 3 in 10 White victims aged between 55-64 were blue-collar workers.

5.3.7.3 ‘Coloured’ and Indian Victims

The cell sizes for these two groups were too small to be analysed.

5.3.8 Place where crime took place

• In both the Mamelodi and Randburg samples, victims were more commonly (39%) attacked in the street than anywhere else. This is probably related to the types of crimes that are reported to the police. As we know from the National Victim Survey, domestic and family violence is less likely to be reported to the police than most property crimes.
• Overall, a quarter of crimes studied took place in the victim’s own home.
• Incidents of crime took place for one in five people in Randburg at their place of business, while this was the case for only one in twenty in Mamelodi.

5.3.9 Relationship between victim and perpetrator

• In the Randburg cases, three in four (74%) victims did not know their assailant, while for Mamelodi, this applied to 40% of reported cases.
• In the Mamelodi sample, the majority of victims (59%) said their assailant was known to them as a friend, acquaintance or member of their extended or immediate family. This applied to 24% of cases in Randburg.
• In 45% of the cases studied overall, victims knew their attackers as friend, acquaintance or member of their immediate family.
• Attacks by unknown assailant/s took place in half the total cases studied.

• In Randburg, most victims, both men (78%) and women (60%), were victims of a crime committed by a stranger, while this applied to one in two men in Mamelodi, but only 18% of women
• Women in both Mamelodi and Randburg were more often than men the victims of crimes committed by their friends, acquaintances, immediate and extended family members.

5.3.10 Gender of perpetrator

• Men were, by far, the most common perpetrators of the violent crimes we studied. In Randburg, more than 90% of both male and female victims were victims of crime committed by men, while in Mamelodi, in 90% of all recorded dockets where a man was the victim of a crime, the perpetrator was also a man.
• In Mamelodi, 17% of the violent crimes against female victims were committed by women.
• One in ten of the crimes studied in Mamelodi were committed by women.

5.3.11 Victims’ role in their own injury

The police do not routinely collect information on the state of sobriety of victims of crime; so this aspect of a victim’s possible contribution to their own injury due to substance abuse is very hard to assess. Researchers were instructed to look for evidence in the statements in the docket as to whether drugs or alcohol had been consumed by the victim in the period immediately prior to the incident.

• In 64% of all cases studied, there was no evidence in the docket to suggest that the victim had used drugs or alcohol.
• In 33% of cases studied, it was not known whether the victim had used drugs or alcohol. There was no evidence to suggest that the victim had or had not taken any drugs or alcohol. This however, does not mean that there was no alcohol or drugs in the incident, only that it was undetermined whether any substances were used by the victim.
• This finding shows that police records (in their current form) would be an insufficient source of information on victim sobriety. More accurate records on this aspect may be available from hospitals and district surgeons, in cases where victims seek medical treatment.

• Incidents where the victim was provoked or involved in a fight were greater in Mamelodi (44%) than in Randburg (20%).
• A negligible percentage (1%) of victims were involved in a crime when they sustained their injuries.

5.3.12 Weapons used by perpetrator/s

• In Randburg, the primary weapon used against both men and women was a gun.
• The gun was also the main weapon used against men in Mamelodi, while for 44% of women, the actual weapon/s used was not recorded or established.

• Male perpetrators of violence in our sample were more likely to use a gun than any other type of weapon.
• Female perpetrators were more likely to employ knives and blunt or other objects.
• Cime victims in Randburg (and male victims in Mamelodi) were more likely to be subject to gun violence than to have any other weapon used against them.

• In the 236 cases where a gun was used, 23% were used during an attempted murder, and 73% during an aggravated robbery.
• In the 61 reported cases where a knife was used, 75% of all these cases related to a case of assault GBH.
• Blunt objects were employed in 20 recorded cases, largely assault cases.
• 23 cases were reported where multiple (more than one) weapons were used, the majority of them during aggravated robbery incidents.

5.3.13 Type of injuries sustained by victims

• In 46% of cases, the available information suggests that the victims sustained no injuries.
• Where injuries were sustained, the most common injuries sustained were cuts and bruises, and minor injuries^[68].
• Although guns were the most common weapon used in both Mamelodi and Randburg, bullet wounds constituted only 6% of all injuries. This is probably related, in part, to the fact that firearms were brandished or pointed in 48% of incidents, but not necessarily used to inflict injury.
• There was a marked difference in the injuries sustained by victims in Randburg and those in Mamelodi. While 71% of male victims studied in Randburg did not sustain any injuries, the same applied to only 42% of male victims studied in Mamelodi. Six in ten women victims (63%) were unscathed in the Randburg sample, while this applied to only 16% of female victims in Mamelodi sample.
• The greatest proportion of female vic