Africa Check’s credibility crisis

14 December 2017  

To use inaccurate statistics in public can be catastrophic to a person or organisation’s credibility. Unfortunately, it is evident that everyone is not equally concerned about this. It is precisely for this reason that there is a real need in South Africa for an institution such as Africa Check, founded to verify the correctness of statements made by public figures.

Unfortunately, Africa Check has established itself as an organisation that actively promotes a particular political ideology and that is fixed on publishing sensational findings under overdramatised headings rather than engaging in sincere fact checking. Not only have I repeatedly been misquoted by Africa Check and declared to have been “wrong” or “dishonest” based on something I never said – I have also repeatedly seen it happen to others.

The first time was in 2013, shortly after the founding of Africa Check, when one of their fact checkers called me to clarify a certain statement that I made regarding white poverty. I was quoted on BBC to have said that more than a hundred thousand white people live in circumstances “like these”. The circumstances in question involved people living in Wendy houses, back rooms, derelict buildings and other kinds of housing. Africa Check reported my statement to be false, based on a survey which found that 8 000 white families lived in shacks, caravans and tents. This is the very definition of a straw man argument: concluding that someone is dishonest because something he never said isn’t true. What is even more interesting about that encounter was the way in which I was quoted by Africa Check. We discussed the matter telephonically in Afrikaans. When the story was published, my words were translated into terrible English, complete with “sic” indications. The misrepresentation of someone as being not proficient in English raises questions about that person’s intelligence and makes it easier to question facts presented.

I have repeatedly noticed that people affiliated with Africa Check position themselves as ideological activists, rather than objective fact checkers. I have also noted an apparent lack of training in logical reasoning. For example, when a notable individual linked to Africa Check (allegedly in charge of training) tweeted the following: “Because let’s be honest, it’s MORE FUCKING LIKELY ALL WHITE MEN WILL GO TO THE MOON THAN LOSE THE VOTE”, and I responded with “This type of tweeting is damaging to @AfricaCheck’s reputation as a non-partisan fact checker”, that same person responded to my tweet, claiming that I accused her of committing hate speech. #Strawman.

Earlier, Dr Frans Cronje, CEO of the South African Institute of Race Relations, stated the following with reference to Africa Check:

“(T)here is a strong propensity to declare as ‘wrong’ or ‘incorrect’ fact-based claims that could easily be argued to be right. This is dangerous as it will deter journalists, researchers, and politicians from citing data in their analyses.”

Africa Check lashed back, reporting him to have said something completely different and attacking him based on that which they have made up. The acting editor of Africa Check wrote,

“In a Politicsweb article, Frans Cronje of the IRR then criticised Africa Check for questioning the institute’s findings and suggested that scrutinising the evidence, good or bad, for claims made in public debate is ‘dangerous’.”

By responding as such to Cronje’s statement, Africa Check confirmed that there was merit in his concern. Read it again… What Cronje said was that it was “dangerous” to declare as “wrong” or “incorrect” fact-based claims that could easily be argued to be right. Africa Check then claimed that he said that scrutinising evidence for claims made in public debate was ‘dangerous’ and then criticised him based on what they falsely claimed him to have said. Seriously?

As was the case in my own experience, the portrayal of Cronje’s statement could only be one of two things: A malicious misrepresentation of his statement, or a negligent misrepresentation of his statement.

More recently, Africa Check published a sensational story in which it claimed that AfriForum was lying about farm murders. They conclude this after analysing what they claim to be AfriForum’s position on farm murder statistics. This analysis, however, only proves that the author had a flawed perception of AfriForum’s position on farm murder statistics and seemingly doesn’t understand the underlying nuances of the topic. Had Africa Check attended any one of the many conferences and events where I had personally explained AfriForum’s position on farm murder statistics, they probably would have taken a different approach. This would, however, have resulted in a less sensational, less clickbait-friendly analysis. I explained AfriForum’s position and the challenges on the statistics regarding farm attacks on several occasions, and also wrote extensively about it in a book that will be published in the next year.

AfriForum has repeatedly outlined the flaws in the available data on farm murders, but believes the data to be the most reliable attempt to calculate a ratio at which commercial farmers are being murdered. Africa Check however suggests that we at AfriForum are oblivious to the very same flaws in the data which we have pointed out ourselves, then criticises AfriForum’s conclusion, committing a series of calculation errors and logical fallacies in the process, only to conclude that AfriForum was dishonest from the get-go. All the while, Africa Check has made no attempt to provide a more accurate calculation for the ratio at which South African commercial farmers are murdered and prefers to stick to a conclusion based on the fact that the calculation is not perfect, inadvertently confirming AfriForum’s initial stance: that the data has certain flaws which should be taken note of, but that the calculation used by AfriForum is the best attempt at quantifying the ratio.

In an article dramatically entitled Racial scaremongering in South Africa makes light of women’s murders, Africa Check responded to claims made by Steve Hofmeyr regarding black-on-white crime. Hofmeyr had stated that “white women are likely to be murdered by unknown black males”. Africa Check’s team aggressively reacted on Twitter and responded with an analysis, declaring his claim to be false. Africa Check however failed in its own analysis, misinterpreting the sources that they have chosen to disprove Hofmeyr, when those sources prove that there is merit in Hofmeyr’s claim.

Put differently, what we have repeatedly seen Africa Check do is to respond to a person or organisation that argues, for example, that 1 + 1 = 2 by stating that this person is wrong or dishonest because Africa Check can prove that 2 + 2 is in fact 4 and not 2. Electrifying language such as “dodgy stats”, “dishonest” and “scaremongering” is then used, presumably to clickbait readers into visiting their website.

Having experienced that Africa Check misrepresents the views of those whose facts they check, that they compare apples with oranges, and that the organisation (or at least some of its reporters) are ideological activists who lack training in logical reasoning or the application of their training, it is evident that the organisation has a growing credibility crisis.

Ernst Roets

Ernst is Deputy CEO of AfriForum

Follow Ernst on Twitter at @ErnstRoets