The media’s culture of confirmation bias

When it comes to South Africa, US media publishes articles that may have been written already before an event even happened.

Central Johannesburg (Simon Inns via Flickr).

Within hours of Adam Lanza shooting dead 20 Sandy Hook Elementary first-graders and seven adults, including his own mother, and taking his own life, media channels everywhere were festooned with analysis and opinion on what drove the 20-year-old to do it. Some blamed America’s gun culture and the free-availability of these weapons of mass massacre, while others attributed fault to mental illness, particularly (and incorrectly) autism. The violent video games Lanza played made him do it, some said. It’s the white male middle-class entitlement, I tell you, said others. Or maybe it was that Lanza was an ostracized victim of bullying.

The explanations were compelling, logical-sounding and some had statistics and reports to back up the claims. Most of the perpetrators of mass shootings in the United States are white middle-class males. Undiagnosed mental illness is a quiet killer. Video games do contribute to normalizing violence (but no more than the news, Hollywood action flicks or real life, in some areas).

However the problem is that there was scant information to say for sure whether, or if at all, any of these explanations fit the Sandy Hook shooting. In fact, many factual errors were aired and published before accurate information came to light, so much of this early analysis was based on preconceived ideas into which the shooting was made to fit. That this early analysis was convincing and had references to back the claims was because it had been all but already written and the writers were on the prowl for circumstances to mold to fit the analysis. It was confirmation bias at its worst, or best, depending on whether you like new information fettled to affirm your beliefs.

Which brings me to TIME magazine’s cover story this week, “Pistorius and South Africa’s culture of violence”. It is the finest example amid a bevy of  fine examples of confirmation bias in the analyses so far of why South African paralympian Oscar Pistorius shot dead his girlfriend, Reeva Steenkamp, in the early hours of Valentine’s Day.

“Why?”, the article wails. “Why is gun violence so prevalent in South Africa? Why is violence against women so common? Why was this homicide? Why did Oscar kill Reeva?”

It then proceeds to answer why it was homicide (as opposed to premeditated murder, the actual charge Pistorius faces) by artlessly blending history, research, news reports and anecdote to conclude that Pistorius shooting Steenkamp dead was a logical conclusion of South Africa’s segregated past, its persisting racial inequalities and the failures of the post-apartheid government to provide effective, reliable policing to the middle class. It does this based on one of those pre-existing, gummed-together narratives about South Africa that, if you excise enough persnickety contradictory information and gloss over the finer details, can be used to explain just about any act of violence committed by rich and middle-class South Africans.

The narrative goes something like this: South Africa is steeped in a racially unequal and divided history and present. This makes the haves, especially the rich white ones, like Pistorius, bloody scared of the black male have-nots coming to pillage and rape their women and children, which is why the haves are armed to the teeth, have a private security force and mistrust the criminal justice system run by the country’s first post-colonial black-led government. It makes them so scared and irrational, in fact, that they might mistakenly shoot dead their loved ones three times through a locked bathroom door for fear of the poor, black bogeyman.

If you beat a police officer to death with your bare hands, for example—as rugby player Bees Roux did when a black officer, Johannes Mogale, tried to pull him over on suspicion of drunk driving—it was because you thought you were being hijacked. If you are a well-heeled tourist, like Shiren Dewani, looking to murder your wife while on holiday in South Africa, your story becomes more passable if you hire black gunmen from the township to stage the deed as robbery gone wrong.

“For all its defenses, it [the government] failed to keep violence at bay. By Pistorius’ account, his fear of an intruder, the fear that keeps the people of South Africa apart still, caused the man so many saw as a unifying figure to shoot his girlfriend dead,” the article’s author, TIME magazine Africa bureau chief Alex Perry, writes.

But there are a number of problems with this. The first is that, as with much of the analysis of the Sandy Hook shooting, there is little credible information to conclude definitively on Pistorius’ state of mind. All we have, all Perry relies on to hold his central thesis together, is an affidavit Pistorius’ lawyer, Barry Roux, read to the court during the bail hearing. That affidavit is the killer’s untested version of events that night. There was no cross-examination and it is yet to be seen whether Pistorius’ account will gel with the forensic evidence, once it that been finalized. Nowhere does the article as much as hint to the fact that it is founded on an untested version of events given by the man who stands to lose the most at this point in the saga.

There are also suggestions that Pistorius’ version may have factual defects, such as why his girlfriend locked herself in the bathroom, why she did not answer when, as he claims, he warned the supposed intruder to leave his house immediately, and when exactly he put on his prosthetic legs. On the last one, the police say it was before he shot Steenkamp. He says it was after. The forensic evidence confirming the angle of the shots will resolve this, but there is enough doubt that we ought to regard with suspicion a 3000-word article that takes Pistorius’ supposed fear of crime as not only the cause of Steenkamp’s death but as the confluence of South Africa’s post-apartheid dissolution.

In blind pursuit of its central thesis, the article also suffers from massive overreach and glosses over the very disparities it highlights.

“To understand Pistorius and Steenkamp, to understand South Africa,” Perry writes, “it helps to know the place where the couple chose to spend their holiday. Cape Town.”

So yes, of course, the socioeconomic circumstances of a city 1200 miles away give context to the mindset of a man who spent most of his life nowhere near Cape Town. Never mind that those circumstances aren’t necessarily extrapolateable to Pretoria, where Pistorius lived and attended school, or Johannesburg, where he was born.

In its submission to the commission of inquiry into allegations of ineffective policing in Khayelitsha, a township near Cape Town’s newly upgraded multi-million dollar international airport, the Civil Society Prison Reform initiative pointed out that violent crime in the city, and in South Africa in general, was not uniformly experienced. While the national crime rate has gone down in recent years, the submission said, “the crime rate in Khayelitsha (and many similarly situated areas) … has not only bucked the national trend and saw increased crime, particularly violent crime, but is also much higher than many other suburbs of Cape Town.”

So even if, somehow, Cape Town’s yawning inequalities increase potential for violent crime, that danger is concentrated to a few, mostly poor areas where policing, the criminal justice system, and general measures to deter crime, like street lights, have failed. The epicenters of violent crime in South Africa are far flung from the sandy-white beaches of Clifton or the high walls and electric fences of Pistorius’ Silver Lakes townhouse estate in Pretoria, the capital.

Finally, also glossed over in the article is that blended in the crime statistics referenced is a significant amount of acquaintance violence, which is committed by perpetrators known to the victims and not some unknown black bogeyman. In fact, most cases of murder, assault and sexual assault in South Africa are committed by family members or people known to each other, according to the Centre for the Study of Violence and Reconciliation. Thus the black bogeyman element the country’s apparent culture of violence is mostly a work of popular middle-class fiction and is worthy of critical deconstruction, not propagation as Perry does in his article.

In the end, it may be that Pistorius, feeling vulnerable while hobbling around on his stumps and dogged by fears of crime, shot first and asked questions later, as South Africa’s erstwhile police commissioner Bheki Cele reportedly implored police to do in the face of increasingly violent criminals. It may even be that South Africa does indeed have a culture of violence, but there are important details and complexities to the situation and there is presently far too little credible information to make the leap from icon-shoots-cover-girl to a nation falling apart under the weight of crime and inequality. Those who’ve made the leap were likely mid-jump anyhow and Pistorius killing Steenkamp provided the perfect springboard to leap to ill-conceived conclusions.

Further Reading

Not exactly at arm’s length

Despite South Africa’s ban on arms exports to Israel and its condemnation of Israel’s actions in Palestine, local arms companies continue to send weapons to Israel’s allies and its major arms suppliers.

Ruto’s Kenya

Since June’s anti-finance bill protests, dozens of people remain unaccounted for—a stark reminder of the Kenyan state’s long history of abductions and assassinations.

Between Harlem and home

African postcolonial cinema serves as a mirror, revealing the limits of escape—whether through migration or personal defiance—and exposing the tensions between dreams and reality.

The real Rwanda

The world is slowly opening its eyes to how Paul Kagame’s regime abuses human rights, suppresses dissent, and exploits neighboring countries.

In the shadow of Mondlane

After a historic election and on the eve of celebrating fifty years of independence, Mozambicans need to ask whether the values, symbols, and institutions created to give shape to “national unity” are still legitimate today.

À sombra de Mondlane

Depois de uma eleição histórica e em vésperas de celebrar os 50 anos de independência, os moçambicanos precisam de perguntar se os valores, símbolos e instituições criados para dar forma à “unidade nacional” ainda são legítimos hoje.