It's been three weeks since a battalion of the South African National Defence Force (SANDF) was deployed to the crime-saturated Cape Flats. Last week, the Western Cape government reported that 25 murders had occurred on the first weekend after deployment, with 46 the next.
Are these figures important? In the absence of a clear goal and mechanism through which to measure progress, it is hard to say. Authorities should reveal their plan for the three-month deployment and regularly report on indicators by which success is being measured.
Cape Town is among the most violent cities in the world. In 2017/18 its murder rate was 69 per 100 000 residents compared to the national average of 36, and a global average of 6.1. Since November 2018, over 2 300 people have been murdered in the Western Cape province, most in specific parts of the Cape Flats. In Philippi for example, 14 people were murdered in just three incidents over the weekend of 5 June 2019.
Amid such bloodshed, the notion that streets saturated in armed men will bring order is intuitively appealing. But intuitions are often wrong. Action should be guided by a plan based on careful analysis of reliable and relevant data, and measured against a set of predetermined indicators and time frames.
It is often suggested that the bulk of Cape Town's violence is gang-related, but is this true? Police statistics suggest otherwise. In 2017/18 the South African Police Service couldn't determine a motive for 38% of murder cases reviewed in the Western Cape. Murders attributed to arguments, domestic violence and retaliation/revenge accounted for 23% of cases while 22% were identified as gang-related.
Gang-related motives may be easier to identify than others through crime intelligence, the modus operandi (such as types of firearms used, gun battles, hits) and victim tattoos, for example. However, when looking at the overall drivers of murder it becomes apparent that most violence in the province remains interpersonal rather than gang-related.
Similarly, the city's highest murder counts (totals) and rates (per 100 000 residents) are recorded in areas not historically associated with organised gangs, such as Nyanga (308/135:100 000), Philippi East (205/323:100 000) and Khayelitsha (192/110:100 000), rather than 'gang stations' such as Manenberg (61/63:100 000), Bishop Lavis (98/82:100 000) or Elsies River (54/69:100 000).
Also telling is that most murders occur over weekends and evenings, when people are likely to drink and socialise, and most probably result from sharp and blunt force trauma.
Considering the numerous factors driving the high rates of both gang and interpersonal violence in affected communities - e.g. child neglect, domestic conflict, poverty and unemployment, substance abuse, organised crime, weak state services - it isn't clear why the SANDF has been deployed. Is it to stop friends and lovers from killing each other, truant youth from robbing commuters on their way home from work, or organised gang members from shooting each other in the streets? Each requires a different strategy.
It is generally accepted that crime and violence cannot be stopped by police (or military) deployment alone. But police remain central to ending endemic violence. As the embodiment of the state's claim to the monopoly on force, their role is to ensure social stability so that crime generators can be addressed. But the two must occur together and be united by an intentional, clear plan.
What can the police do while the SANDF supports them?
There is evidence that hotspot policing is effective, especially when coupled with problem-solving initiatives. But hotspots are not high-crime police precincts. They are the small parts of such precincts in which harm is most common. For each hotspot, interventions should be adapted, including police tactics, to suppress violence while other agencies focus on addressing local enablers.
Focused deterrence is another effective tool. This involves identifying likely offenders (e.g. those at risk of perpetrating gang- or firearm-related violence), and offering them support services (e.g. substance-dependency treatment and job training), together with a message that violence won't be tolerated. A range of agencies must collaborate to provide these services.
Similarly, intelligence-led policing can have an impact. If intelligence was being effectively used from when the SANDF was deployed, one would have expected raids culminating in high-profile arrests, seizures of weapon stockpiles, and the flushing out of corrupt police. While raids, and stops and searches have occurred, their purpose isn't clear.
We also don't know what impact they have had on feelings of safety or perceptions of legitimacy in affected communities. How safe residents feel, the degree to which they trust police and soldiers, and whether they perceive them as legitimate authorities, should be key measures of this operation's success.
These, together with weekly numbers of people appearing in hospitals and clinics with violence-related injuries, would tell us a lot more about the operation's impact than reported murder alone. And yet even these are insufficient to measure 'success'.
Assuming a best-case scenario in which all violence ceases during the SANDF deployment, we'd still want to know how life in the affected communities has changed. And whether mechanisms to address structural drivers of violence have been planned or put in place. Are fewer learners skipping school, more pregnant mothers attending clinic check-ups, and more children attending after-school programmes?
To evaluate the success of the army's deployment, we need to know why they were deployed. In an ideal scenario, how would this end? Based on this vision, reliable indicators should be selected, tracked and regularly reported. Where data suggests the plan isn't working, it should be adapted.
The deployment of the SANDF in Cape Town provides the national, provincial and city governments with a rare opportunity to alter the violence entrenched in parts of the Cape Flats. To do so, however, they need to be clear about what they want to achieve and how they hope to do so.
Stuart Mbanyele, Consultant and Andrew Faull, Senior Researcher, Justice and Violence Prevention, ISS Pretoria