A Citizen's Guide to Crime Trends in South Africa

Anine Kreigler, Mark Shaw

• Publisher: Jonathan Ball
• Year: 2016
• Print ISBN: 9781868427222
• eBook ISBN: 9781868427239
• Topic: Social Sciences
• Subtopic: Criminology
• Index: Social Sciences

South Africans care a lot about crime. We think and worry about it, plan and insure against it, develop and share theories about it, read about it, and talk about it… a lot.

But how much do we really know? Crime statistics do not belong to the government, academics, specialists, or the press. They are ours: we experience and report crimes and have a right to access and understand their official record. It should not take any particular expertise to get a grasp on what we should make of the figures and graphs that the South African Police Service produces every year.

A Citizen's Guide to Crime Trends in South Africa provides a basis on which to understand the statistics in a manner that is accessible to everyone. Each chapter challenges a set of oft-repeated assumptions about how bad crime is, where it occurs, and who its victims are. It also demonstrates how and why crime statistics need to be matched with other forms of research, including criminal justice data, in order to produce a fuller account of what we are faced with.

1

What are crime statistics?

Every year around September, the South African Police Service (SAPS) produces a number of spreadsheets indicating the number of recorded incidents for each of about 25 major crime types in the preceding year, which runs from 1 April to 31 March. It does so for each station, and provides totals for each province and for the country as a whole. It indicates the change in percentage terms from the previous year and some other reference years, and before 2015 also gave the provincial and national totals as rates per 100 000 people in the appropriate population. Soon after, it also releases an annual report that reflects these figures as graphs, and puts them in the context of annual targets, operational categories, crime reduction strategies and some further information about some of the crimes, for example the size of the businesses targeted in most non-residential robberies or the weapons most commonly used in murders.

This information gets a lot of attention from various quarters and is treated with varying degrees of scepticism. News broadcasts list the highlights and show file footage of police vans flashing their lights, radio hosts opine and ask laypeople to call in with their reactions, experts are asked for comment, infographics are shared, letters are written to the papers, and the police are almost invariably criticised for failing to do enough to control crime. Throughout, it is also often noted that these figures may not be ‘reliable’ or ‘valid’. The reasons for this are seldom clearly spelled out, but often seem largely to reflect a deep mistrust of the police at every level, from the sergeant at the desk all the way up to the National Commissioner and Minister of Police, and sometimes of the value of any quantitative research and of the experts’ location in the ivory tower. For reasons of history and politics, it may be that South Africa is a nation unusually wary of the claims made by those in power. Whatever its cause, this scepticism is an excellent starting point for examining any factual claims about our collective social reality, especially given how influential these claims can be. Most conversations about the crime statistics, from the most casual on social media to the most careful in an academic context, will at some stage have to engage with the question of what can reasonably be inferred from them.

A key difficulty with trying to create and make sense of knowledge about crime is that it is by its nature an illicit activity, a thing that someone usually wishes to keep hidden in its entirety or in its details (i.e. who did it). Another problem is that ‘crime’ is a very broad category of behaviours, including everything from premeditated murder through cable theft and shoplifting, to the consensual buying and selling of items or services that have been deemed inappropriate for legal trade. A third problem is that, especially for the more serious crime types, the participants are people who are considered particularly dangerous and untrustworthy, so can’t very easily just be asked to report on their own behaviour.1 As a result, research on crime will always present more challenges than research into, say, purchasing habits, or anything else that happens relatively openly, is relatively homogeneous and is done by people considered relatively respectable and reliable.

There is also a nearly endless range of things that we may want to know about crime. Its prevalence (how much crime there is in a given place at a given time) may seem like a fairly easy one to get a handle on, but unfortunately this turns out not to be the case. Although we have a number of sources to draw on for this knowledge, of varying degrees of formality and generalisability, each is crucially limited.

Sources of knowledge about crime prevalence

In trying to answer questions about crime, we all have at least four types of sources on which to draw: our own direct experiences; the experiences of others as they filter through to us by word of mouth, the media, local newsletters, and so on; knowledge based on purposeful research, such as surveys or interviews; and official information.2 Each of these has its advantages and may be adequate for answering some of the questions we may have about crime, but each is partial, and suffers from different limitations when it comes to the question of how much crime there is.

Knowledge based on our direct personal experiences of crime provides us with extremely rich, contextualised data. Because we gather it ourselves, we can be reasonably sure that it has not been distorted by others’ interests. The personal perspective based on being a victim of, witness to or perpetrator of crime is an unrivalled source of detail and emotion in building up the picture of crime. On the other hand, there is no reason to believe that one individual’s experience says much about the experiences of the over 50 million other people in the country, so it doesn’t help very much with the question of crime prevalence. Of the many things one might reasonably learn from experiencing a house robbery, the frequency of house robberies in general is not one of them. Our own memories, especially when they are traumatic, are a common and tempting but very poor basis for generalisation.

Next there is knowledge that filters through to us from others’ direct experiences. Again, this knowledge may be rich in detail and context – we may get a relatively full story in the victim, witness or perpetrator’s own words, or as described by someone else, such as a shared neighbour or a journalist. Because the pool from which we are drawing is slightly larger, this knowledge is probably slightly more generalisable than our own experience alone. If we hear about ten house robberies in our suburb this year and only heard about three last year, we may make a tentative claim about the relative frequencies of house robbery in this year and last. However, again we are working with an incomplete picture even of just our area, never mind our city or the country as a whole. We may be missing many people’s experiences for a range of reasons, or only hearing about the experiences that make for good stories, or they may have been intentionally or unintentionally distorted in the telling, or we may well just not be as good at remembering things about last year as we think we are. The process of filtering through others can either chaotically or systematically bring an unrepresentative proportion of the incidents to our attention, making this source of information limited in helping to draw good conclusions about how much crime there is.

A third source of knowledge about the prevalence of crime is one more often used by specialists than others with a less academic interest in crime, as it can be expensive to produce and sometimes difficult to interpret. There are many thousands of highly skilled people worldwide whose daily job it is to do quality research on crime, including (but certainly not limited to) on how much of it there is, or whether that is a question that can even be meaningfully answered.3 They go about this in many different ways, but two of the most common methods of getting a sense of how much crime there is (plus usually a range of related matters) are studies in which people are asked about their experiences of witnessing or being affected by crime, known as victimisation surveys, and ones in which people are asked about their experiences on the other side of the law – known as delinquency surveys. Perhaps the first formal crime survey ever was initiated in 811 by Emperor Charlemagne, and sent to magistrates and nobles to try to get a sense of the reasons for an apparent increase in property crimes in the empire.4

Attempts to measure crime and compare populations on the basis of standardised victimisation surveys have the advantage that they can work around some of the flaws and differences in criminal justice system reach and statistical capacity. They can be carefully planned so as to be as systematic and unbiased as possible. The people to be surveyed can be selected in different ways, including so as to make for maximum generalisability. Overall, a key advantage of a survey is that each person can (at least in theory) be asked the same questions in the same ways, which should make it possible to add together and compare their answers. If a suitable selection of people in one place and time reports having experienced house robbery with a certain frequency, and a suitable selection in another place and time report having experienced it at a different frequency, that is a fairly solid basis on which to draw conclusions about those relative frequencies.

There are a number of constraints even on this kind of research. First, when it comes to questions about large populations, conclusions will have to be drawn from the answers of a sample of people rather than every single one. Only the national census attempts to reach every household in the country with its survey questions, and conducting a census involves such hug...

Table of contents

• Title Page
• Preface and acknowledgements
• Introduction
• 1 What are crime statistics?
• 2 The South African crime statistics context
• 3 Headline international comparison
• 4 Breaking down the numbers
• 5 A national history of murder, 1911–2015
• 6 The changing character of violence, 1994–2015
• Conclusion: what makes crime go down?
• Appendix: the numbers
• About the Book
• About the Authors
• Imprint page