What to look for in South Africa's troubling crime statistics
Since the 19th century, a growing number of countries have attempted to systematically collect and analyse information on all the crimes known to their police.
At least 139 have made some such data public at least once in the last decade, but there is major variation in quality and in the frequency and ease of public access.
Crime is not at all evenly distributed in any country or city, so smaller scale data is essential. The US releases a semiannual preliminary report and an annual report. Many US cities publish quarterly, monthly, or even weekly crime statistics.
In Africa, crime or criminal justice statistics remain scarce, fairly ad hoc, and often of poor quality. And possible changes in the public’s reporting behaviour and the police’s recording behaviour can make it difficult to determine whether a change in the crime statistics is “real” or rather a “statistical illusion”.
In some countries, less than one-third of the crime that people experience is ever reported to the police. Especially for crimes that are substantially under-reported (such as sexual crimes), an increase in the recorded rate may mean more people are reporting their experiences to the police rather than that more crimes are happening.
South Africa faces similar challenges. The South African Police Service (SAPS) publishes national, provincial and station crime statistics and has done so for several years. Experts and activists have called for more regular and detailed crime statistics. Many lauded the announcement in June that the crime statistics would soon be released on a quarterly basis, but there is as yet no word on when or how this will be implemented.
Every year, the release of the SAPS crime statistics is met in equal parts with praise, criticism and scepticism. The African National Congress tends to interpret any reductions as indicators of success in policing and social development, the opposition does the reverse, and the police themselves tend to play down the extent to which they can be held responsible for the crime rates at all.
How the country makes sense of the crime statistics matters because it provides insight into how the society is shaped. More crucially, it should inform public and personal decisions about how best to make progress against crime.
What to look for in the numbers
The annual crime statistics usually show some rates going up even as others go down. Each of the almost 30 different crime types may be of particular interest to different people and interest groups, but we think that sustained trends in five areas provide an important snapshot of where we stand:
Murder rate: will the last few years' upswing in the national murder rate continue? Or will it stabilise or decline?
We have shown there was a long decline in the national murder rate between 1994 and 2011, followed by an uptick in the last three years although it is still at about half what it was 20 years ago.
As we indicate in our book, murder is often considered an indicator of levels of violence and crime more generally, so its trends are highly noteworthy. A continuation of the recent upward trend would suggest the significance of new dynamics driving violence, perhaps including factors such as greater political volatility and growing inequality.
Murder rate in Cape Town: will this again rise disproportionately to the rest of the country? Since 2011 Cape Town has taken a growing lead among the metros, to the point where its murder rate of more than 60 per 100,000 people is now about twice that of Johannesburg and among the highest in the world.
Further increases would be a very worrying sign, pointing to the increasing entrenchment of gangsterism and the availability of firearms, possibly related to the large-scale illegal sale of firearms by a corrupt SAPS officer. Tackling this will require effective joint action from national and local government.
Aggravated robbery trends: House and business robberies are among the very few crime types that have seen significant increases over the last decade. They are also major contributors to public fear, so their trends are important in understanding how people experience and perceive crime and policing.
Some experts also consider aggravated robbery to be relatively subject to police prevention, especially through crime intelligence. Continued increases could therefore suggest a failure by the police. Conversely, a decline would be an important achievement.
Burglary: These crimes have seen significant declines over the last ten years (supported by victim survey data), so a stabilisation or increase would be a major reversal of the long-term trend and a cause for concern. Although it does not involve physical violence, burglary is highly invasive. It is one of the most-feared crimes, so its trend is an important contributor to public perceptions of safety.
Drug-related crime and illegal possession of firearms: These crimes are heavily dependent on police action for detection. They are one of the few indicators of the extent of police activity. While such data are not definitive, increases for example in firearms seizures can be a measure of police attempts to reduce violent crime.
The data on recorded crime should always be interpreted in the context of information from other sources, primarily the crime victimisation survey completed annually by Statistics South Africa.
Finally, it must be stressed that crime statistics are public property. Every citizen has a right to access them easily, promptly and in a format that makes them useful to our lives and decisions. We also have a responsibility to read them critically but honestly.
Anine Kriegler receives funding from the National Research Foundation and the David and Elaine Potter Foundation.
Mark Shaw receives funding from the National Research Foundation.