A recent analysis by Africa Check, an organisation that promotes accuracy in media reporting, has called into question the accuracy of commonly cited farm murder statistics for South Africa. By looking at the available data, they concluded that it is near impossible to accurately calculate farm murder rates. This calls into question the popular narrative that South Africa’s white farming minority is under siege.
Farm murders in South Africa are a highly politicised issue. Outside of the general concern for homicide in the country, the farm murder rate has been used by many as evidence that South Africa’s white minority is facing victimisation, oppression, and possibly genocide.
But even outside of spaces that are arguably motivated by an ideological agenda, there are others who have bought into the idea that South Africa’s farm murder rate is especially high. As Africa Check noted, during a parliamentary debate, the Freedom Front Plus and African Christian Democratic Party quoted farm murder rates of 97 and 133 per 100 000 people. This rate would make farm murders around three times the national average of 34 per 100 000 people murdered per year in the 2015/2016 period.
These murder rates, according to MPs, were quoted from the Institute of Security Studies’ Johan Burger, who claimed he never intended for them to be more than a vague indicator. Yet, even in this, it seems the manner in which Burger calculated the rate was inaccurate.
A farm murder rate is calculated by deciding on a definition of what constitutes a farm murder, dividing it by the population of those involved in farming, and then multiplying it by 100 000 to give an indication of how many people are affected per 100 000 individuals. A murder rate is important because the absolute number of murders by itself doesn’t tell us anything about how pervasive murder is for a certain group – those numbers would be expected, on average, to be greater for larger population sizes.
Africa Check’s article details exactly what they believe Burger did, but the gist of it is that the figure he used for number of farmers murdered a year (32) was not consistent with the population of farmers he divided by (32 375) in order to arrive at the murder rate. The number of murders included those on both commercial and non-commercial land, and smallholdings, whereas the population of farmers he used only included commercial farmers and those who were registered for VAT.
Given that the population size Burger was dividing by was smaller, this would inflate the murder rate. To put this into perspective, Statistics South Africa estimated that 2.3 million households were involved in agriculture based on its 2016 Community Survey. The agency’s chief director of structural industry surveys, Itani Magwaba, told Africa Check that this included subsistence, smallholding and commercial agriculture.
If this number is used alongside the highest current estimate of murders on farms (64 in 2016 according to AfriForum) this number results in a murder rate of just over 3.2 per 100 000 people. But again, this number is not necessarily representative as farm murders are not all captured, and “involved in agriculture” is too general a description.
Africa Check’s senior researcher, Kate Wilkinson, told The Daily Vox that if we want to work out a murder rate of farmers, which is different to farm murder rate (the former is the rate of farmers murdered while the latter is the rate of people murdered on farms), then at the moment we can only really use the Transvaal Agricultural Union of South Africa statistics, because they’re the only people who are collecting the number of farmers murdered. We’d then have to get an estimate of the numbers of farmers in the country, whether they be commercial, smallholding, subsistence – but right now that number doesn’t exist, said Wilkinson.
“Until we can be sure what murders are being measured and what definition is being used to estimate the total population size, we can’t calculate this figure,” she said.
Africa Check’s analysis of farm murders is perhaps unsatisfactory in answering questions of whether white South Africans face persecution in South Africa. But maybe that’s the point. Conclusions should be reached only after reviewing evidence; if it’s done the other way around we are no longer seeking “truth”, but rather looking to confirm a predetermined belief – a bias. With this in mind, there is perhaps only one conclusion we can draw from our inaccurate farm murder rates – that those who quote them may have a preference for politics over verity.