A Citizen's Guide to Crime Trends in South Africa

Contents

Preface and acknowledgements

Introduction

Crime is costly

Crime statistics are powerful but beguiling

Crime is political

Notes

1 What are crime statistics?

Sources of knowledge about crime prevalence

The hidden figure and what to make of it

Crime statistics and police performance

The other crime statistics

Notes

2 The South African crime statistics context

The quality of the SAPS murder statistics

The population problem

Unfoundedness

Notes

3 Headline international comparison

Where South Africa fits in

Insight from the company we keep

The trouble with other crimes

Notes

4 Breaking down the numbers

Provinces

Cities

Precincts

Smaller concentration

Notes

5 A national history of murder, 1911–2015

After transition

Notes

6 The changing character of violence, 1994–2015

A shift in the nature of murder?

Trends in assault

The context of assault

Trends in robbery

Subcategories of aggravated robbery

The instrumentalisation of violence

Notes

Conclusion: what makes crime go down?

Notes

Appendix: the numbers

Official police recorded murder figures

Estimated murder rates per 100 000 in police jurisdiction

Other crime rates per 100 000

About the Book

About the Authors

Preface and acknowledgements

This book is the outcome of an attempt to respond to media queries on the annual release of crime statistics in South Africa. In 2015, the Centre of Criminology at the University of Cape Town produced a short descriptive guide to assist journalists and others who wanted to understand the problems with the SAPS crime statistics release in September and to place it in longer perspective. We are grateful to Barry Bateman, with whom we discussed it, and to Chris de Kock, who partnered on that initial product. However, we wanted something that provided not an overview of the data but started to show how the numbers themselves might help determine what must be done to reduce violent crime. This book is the result.

We are extremely grateful to our colleagues and students at the Centre of Criminology and the Safety and Violence Initiative for ideas, debates and discussions about different parts of the argument. A seminar in late 2015 provided some indication of the importance of the long-term analysis of murder data. Social media comments on an article we published in the Daily Maverick on homicide trends also proved to be a spur to more in-depth thinking about what the data might say about policy options. We must also thank Jeremy Boraine of Jonathan Ball Publishers for his enthusiasm for publishing a work that we believe is in the public interest.

This work is based on the research supported by the South African Research Chairs Initiative of the Department of Science and Technology and the National Research Foundation of South Africa (Grant No 47303). Any opinion, finding and conclusion or recommendation expressed in this material is that of the authors and the NRF does not accept any liability in this regard. Anine is also supported in her research by the David and Elaine Potter Foundation.

Mark is grateful for his family’s usual forbearance. Thanks, finally, from Anine to Thomas van Heerden for doing everything that wasn’t this.

As we were completing this book, in March 2016, Cape Town was rocked by the brutal murders of two teenage girls. Sixteen-year-old Franziska Blöchliger and nineteen-year-old Sinoxolo Mafevuka were killed about a week apart. One was found hours after being separated from her family during a run in Tokai Forest and the other was found at sunrise in a communal toilet in Khayelitsha. More than the similarity, what struck many was the contrast in how the cases were received by the police and the public. What was reignited was a debate – arguably the chronic South African debate – about the distribution and meaning of violence within pervasive systems of privilege, exclusion and fear. Addressing the problem of crime will inevitably require addressing those issues. It will also require that we all take the responsibility to inform ourselves about the nature of what we’re talking about. We are conscious that our work around crime data is about real human beings and their experiences. It is for that reason that this book is dedicated to Franziska and Sinoxolo.

Introduction

South Africans care a lot about crime. We think and worry about it, plan and insure against it, develop and share theories about it, report on and read about it, and talk about it … a lot. Crime is by no means a middle class and/or white preserve, but cuts across race and class, with black and poor people in fact disproportionately affected by crime (especially violent crime) ¹ and by fear of it. ² We are not the only country concerned with crime, but our high crime rates combine with other national anxieties – around race, social change, cohesion, history, and so on – to form a cocktail of issues that is potent and uniquely our own. We also find ourselves in the unusual and unenviable position of having both high crime levels and relatively good statistical reach and capacity. There are many who criticise what has been called our obsession with crime, and the ways in which it shapes our priorities, relationships, spaces and collective identity.

Without downplaying such cautions about the primacy of beliefs about crime in our national conception of self – quite the contrary – this book takes the view that our concern with crime is neither avoidable nor undesirable. The conversations about crime that happen in the press, on social media, in houses of every size throughout the country, in neighbourhood watch meetings, boardrooms, break rooms, bars, streets, queues and every other place we come together are an inevitable result of the place crime occupies in our social and political lives. More important, such conversations are an important arena of connection and contestation. Crime, its measurement and its control are not technocratic issues best left to those with specialised education and experience, and at ease with quantitative methods, people perhaps imagined as above the sometimes petty, sometimes downright ugly emotional fray of grappling with crime in this country. It simply doesn’t (or shouldn’t) take degrees in statistics, criminology or history to get a decent grasp on just what we should make of the figures and graphs that the South African Police Service (SAPS) currently produces on an annual basis for public consumption.

At the same time, we all share a responsibility to ground ourselves in reality. There is and always will be a lot about crime that remains open to interpretation and discussion, and scope for our personal and political worldviews to play out. But there are a few things that we should all be reasonably able to agree on, so that we don’t talk past each other or waste time trying to understand or solve problems that aren’t the ones that need to be understood or solved. Failing to engage with these issues impoverishes the quality of our discourse on every level, and makes it more difficult to make any progress. Besides the obvious – that crime enacts harm on those it touches – there are at least four major reasons why we should all continue to engage each other and ourselves in thinking about the magnitude and nature of crime, and why we should attempt to do so with some basic understanding of what we’re actually talking about: crime is expensive; knowledge in the form of figures can be especially beguiling; crime is unavoidably political; and finally, as this book aims to show, crime can tell us things about our society that we might otherwise miss or misunderstand.

Crime is costly

Driven in part by a growing focus on maximising fiscal ‘bang for buck’, the last three decades have seen the proliferation of attempts to measure and compare the costs of crime and crime control. The direct costs are relatively easy to compute. The 2016/2017 national budget allocated R87.5 billion for police services and R41.7 billion for law courts and prisons, for a combined R129.2 billion. ³ That’s about 10 cents in every rand of government expenditure allocated in a tight budget, at a time when global economic pressure is high and local political consensus around appropriate state spending is low. In addition, turnover in the private security industry is said to be in the region of about R60 billion a year. ⁴ So the direct annual costs of attempting to prevent and respond to crime in this country are around R189 billion.

Those costs buy the services of, and facilities used by, 194 000 SAPS employees, ⁵ 487 000 private security officers, ⁶ about 40 000 Correctional Services staff ⁷ and about 22 000 Justice department staff. ⁸ Thus the jobs of about three-quarters of a million people in this country are, at least to some extent, influenced by what the crime statistics tell us and how we interpret and respond to them. This is because the statistics are one of the most basic tools for understanding and quantifying our crime problem – how much of it there is, what shapes it takes in different places, and therefore where and how we should allocate resources to do something about it.

But the costs of crime go far beyond these direct financial allocations. There are the costs of the property damaged and lost, costs related to insurance, to securing homes and property, to dealing with the medical results of crime both in the short term (for example, treating injuries following assault and providing antiretrovirals for rape survivors) and the long term (for example, treating crime-related disability and trauma), and so on. A study in 2000 estimated that the direct medical costs and loss of income alone cost a survivor of rape R1 605 and a survivor of attempted murder R3 928. ⁹ Other research estimated that each homicide victim in the Western Cape in 1998 involved productivity and opportunity costs of about R88 000. ¹⁰ One attempt at including direct financial losses, as well as medical, emotional, institutional and private security costs, estimated that the aggregated cost of crime in South Africa amounted to 7.8 per cent of GDP in 2007. ¹¹ Cable theft alone has been estimated to cost the country R5 billion a year. ¹²

Crime diverts human, financial and time resources in the state, business and individual spheres away from more productive investments in development and growth. ¹³ The cumulative lost growth due to crime is hard to estimate, not least because of the challenge of capturing the lost benefits of spending on other things, that is, the opportunity costs. One rand spent on policing is one rand not spent on education or public transport; a productive worker who emigrates out of fear of crime is one who ceases to contribute to the local economy; a month spent on compassionate leave following a serious crime is a month not spent working; equally, a month in prison for a healthy young person represents a lifetime of reduced prospects of earning a salary to help their family; and on and on. Crime also disproportionately affects those to whom it is most devastating: the poor. Those who have very little are the least able to protect themselves from being robbed of it, and lack the insurance or resources to recover from the shock, thereby worsening cycles of poverty and entrenching inequality.

Methods of estimating the economic impact of crime have improved in nuance and sophistication over the years, for example extending to use self-reported offending rather than just official records, and attempting to account for the cumulative costs that criminal justice action can exert on already marginalised communities. These methods will no doubt continue to improve. But quantitative approaches will always be beset by debates about the appropriate weighting of costs of various kinds and against different groups in society, about whether some crucial factors can be meaningfully translated into monetary terms, and about how to take account of inevitably differing conceptions of fairness and equity. These questions are by no means only of local concern. Quantifying, comparing and setting targets so as to mitigate the impact of crime on people’s livelihoods and lives is no simple task. One of the United Nations’ Sustainable Development Goals, adopted in September 2015, is the promotion of just, peaceful and inclusive socities, and debate is ongoing around the issue of how to measure security as a component of human development. Whether neatly quantified or not, crime and how we respond to it manifestly have considerable impact on the livelihoods not just of victims, perpetrators and law enforcement personnel, but also of the entire society. Crime critically obstructs urgent efforts to improve national wellbeing.

Crime statistics are powerful but beguiling

Figures of all kinds have weight in how we think about our world. When we wish to know about and describe places, we often turn to the collective numbers associated with them, such as their population, aggregate wealth, unemployment rate and crime rate. But this is a fairly new way of approaching things. It was only really in the late eighteenth and early nineteenth centuries in Europe and America that there came to be sufficient capacity for, and perceived value in, neatly categorising and counting various things about the individuals in a society. ¹⁴ From about the 1830s, ever-larger volumes of numerical information came to be produced and used in the governance of people and places, helping to consolidate these into conceptual and practical bureaucratic units. ¹⁵ Quantifying rates of deviancy – of illness, madness and criminality – was central to this process from the first, with statistical description increasingly taking the place of causal arguments involving human nature or the supernatural. ¹⁶ In other words, where once crime may have been primarily understood as the outcome of the universal and individual struggle between good and evil, it came to be thought that there was a certain proportion of society that was dangerous, and that this proportion allowed for and required identification, measurement, ideally prediction and certainly management. ¹⁷

This is a simple but important historical point: crime statistics exist because they are thought to be useful. Collecting, compiling and interpreting them uses limited resources of money and time. The expense makes sense only if it is believed that having the resulting information can help give rise to better understanding and better decisions. As the United Nations manual on crime statistics puts it, statistics can be a partial but invaluable tool to ‘help Governments to assess and monitor the conditions, circumstances and trends of well-being and the social impact of public expenditures and policies’. ¹⁸ Statistics influence decisions about which areas should be allocated more police, what crimes should be targeted for reduction, which police station managers are doing well or badly and whether the police, courts and communities in general are doing the right things to combat and respond to crime. The crime statistics influence decisions about how those who enforce order should conceive of and conduct themselves in relation to the rest of society – for example, whether they should call themselves generals or commissioners – and when they should be allowed or encouraged to use deadly force. ¹⁹

The annual peak in popular interest around the release of the South African crime statistics, at least in the press and on social media, suggests that many ordinary people also consider them worthy of some form of attention. Although some people are highly suspicious of the validity of the statistics as they are presented by the SAPS, or by the poverty of the numbers in fully capturing the messy and painful personal realities of crime, few seem to question whether the task is meaningful or worthwhile at all. The outcry around the moratorium on the release of crime statistics in 2000 may well have been in large part expressive just of frustration and suspicion, but it was bolstered by a sense that we were worse off without that knowledge. There appears to be a general popular sense that, whether or not the numbers we have tell the story we want to hear or are of good enough quality, there is some point in having them. On the basis of our understanding of crime, which may well be partly informed by our understanding of crime statistics (although likely less so than by our own experiences, what we hear from peers and the media, and our other beliefs about ourselves and society), we may change our route home, our mode of transportation, our children’s curfew, our insurance profile, our purchases, our hiring practices, our neighbourhood, our vote,