SA crimes against design #CrimeStats2020 special
Crime Stats day is always a busy one for journos in South Africa, as they rush to try and make sense of what’s been going on in our notoriously violent and crime prone society. This year it’s even more of a challenge since today’s count stops in March – so all lockdown-related drops in criminal behaviour are excluded. We’ll leave the analysis to the serious sites, but skimming through the presentation we couldn’t help but notice a few crimes ourselves… Against visual design that is.
We thought dodgy graphs that don’t respect the rules of the Y-Axis were banned by the Geneva Convention. Turns out they are an acceptable way to exaggerate a decline in key statistics. Like how to make a small percentage fall look really impressive.
There are other examples in the presentation. Now, you might think these are innocuous instances of poor design – except that almost without fail, the non-0 axis used in the presentation exaggerate declining trends which make things look like they are getting better. Numbers that are on the rise – and are therefore bad news – always start from the correct base. Why is that?
Co-incidence? He asked with an archly raised eyebrow.
I’m not even going to look at the sexual offences charts, since the only thing we really know about gender-based violence in South Africa is that we know nothing thanks to massive under-reporting of the problem. It’s the one crime stat where a rise remains potentially a good thing, as it suggest more women are comfortable coming forward and reporting attacks as much as anything else. (See ND passim)
My favourite crime against design this morning, though, is the context-free infographic. We know that the Police Minister has a real thing for prohibition under lockdown, so this year’s stats include a panel on liquor-related offenses.
Now, we absolutely agree that booze is inherently linked to evil behaviour. It’s absolutely a factor in awful, awful crimes from sexual assault to GBH to casual racism. And personally, I totally buy the argument that drink-fuelled incidents of stupidity and criminality are a problem for hospitals who need space for Covid-19 patients.
But put in context these numbers are astonishing… because they are so small.
The total number of reports of Assault GBH for the year is 166 720 – so less than 12% of cases involve alcohol. There were 21 325 murders, of which just 6% apparently involved drink. Frankly, my instant suspicion is that case dockets aren’t getting filled in correctly. In the UK, estimates vary but are pretty consistent that alcohol is a factor in around half of all cases of inter-personal violence. Are sober South Africans really that much more dangerous?
Whether it’s intentional or not, bad graphs don’t just tell the wrong story, they leave the foul phlegm of spin on our eyes and raise suspicion. I’m going to go and unwind with a cup of tea before reading the rest of the report, and hope that SAPS will do a better job next year.