The original article, with full citations, is available in Afrikaans at ontlaer.
On 26 September 2026, some 3,000 white crosses stood on the grass at the Washington Monument. Lex Libertas, led by Ernst Roets, formerly involved at AfriForum, staged the display together with the New York Young Republican Club.
Farm murders have long been a sensitive issue in South Africa, and given the racial polarisation that is almost always present in the debate, it deserves to be taken seriously.
But precisely because it is serious, it also deserves to be examined with honest statistics.
A loaded place to plant
Unfortunately, the white crosses were not planted in a political vacuum. Just a few days earlier, on 15 September, Marco Rubio announced visa restrictions for South African officials said to be involved in “race-based discrimination” and “uncompensated land expropriation.” Donald Trump had already signed an executive order in February 2025 that halted American aid to South Africa and admitted 59 Afrikaners as refugees.
In May 2025, Trump confronted President Ramaphosa in the Oval Office with white crosses that he described as “graves” of more than a thousand white farmers. In fact, it was a memorial display for two murdered farmers. President Ramaphosa’s administration calls the allegations a misrepresentation, while Trump’s allies see them as a legitimate concern about minorities.
Just three days before the crosses were planted, on 18 September 2026, South Africa’s foreign minister, Ronald Lamola, himself spoke at the Congressional Black Caucus Foundation’s African Diaspora Forum in Washington. There he called Trump’s Afrikaner resettlement programme “Apartheid 2.0,” warned that American policy toward South Africa could harm black-white relations in the country itself, and denied that the government targets any racial group. The Democratic Alliance promptly challenged him to apply the same standard at home to the ANC’s own race-based policies, such as BEE and legislation on land.
The tension also cannot be separated from other friction between Washington and Pretoria. South Africa’s genuine genocide case against Israel at the International Court of Justice, pending since December 2023, has created major friction with Washington DC. According to Middle East Eye, Israel itself approached US Congress members in 2024 to pressure the South African government into dropping the case.
What is certain is that the pain of farm murders is now part of a larger political struggle, with roots that go back to the role players themselves, and that every figure should therefore be checked more carefully.
What the crosses show, and what they cannot prove
The list of names that Lex Libertas released in August contains more than 2,203 victims who were self-identified and verified. The public was invited to help correct omissions and errors. There are therefore roughly 800 more crosses than names on that list, about 36% more. However, I could not confirm how many names the final display contained.
The crosses also do not tell us how many of the victims were white farmers. The police definition of a farm murder covers everyone who lives, works on or visits farms and smallholdings. Among the dead there could therefore also be farm workers, managers and visitors of all races. That also appears to be the case when one looks at the photos.
Roets has long leaned on one figure. In 2017, AfriForum, where Roets was deputy chief executive at the time, published a murder rate of 156 per 100,000 commercial farmers. That would mean a farmer is 3.5 times more likely than the average South African, at 50 per 100,000, to be murdered.
In 2017, Africa Check asked how the figure was calculated, and Roets replied that it was the same calculation that had been discussed with Africa Check earlier, only for 2016.
Politicsweb has meanwhile highlighted a different set of figures. In a 2025 analysis of the US executive order on South Africa, Dave Steward wrote that on average 69 farmers per year were murdered between 1994 and 2020, or just under 1,800 over the period. The same analysis cited AfriForum’s calculated murder rate of 156 per 100,000, against a national murder rate of 45 per 100,000, but again did not question it.
The 69 per year is a count of victims over a given period. The 156 per 100,000 is a calculated rate.
Death rate = victims ÷ population of the same group × 100,000.
The number does have a big problem. Stats SA counted 40,122 commercial agricultural units for 2017, but a unit is an enterprise, not a person. AfriForum recorded 49 farm murders in 2023, during 296 attacks. If you divide 49 by 40,122, you get 122 per 100,000, but that is then 122 per 100,000 agricultural units, not per person.
As at 30 June 2023, Stats SA calculated and counted a total of 770,181 persons working in the agricultural sector. That is almost 19 times more than the 40,122 commercial units used earlier as the denominator. These are, in other words, workers who live and work on the same farms as farmers and consequently share the same risks. These people, who for all practical purposes are just as vulnerable as the farmers and farm owners themselves, must be included. Leaving them out of the calculation artificially inflates the rate. And that undermines the point that Roets and others are trying to make.
Divide 49 by 770,181, and you get about 6.4 per 100,000.
That is still not an official farm murder rate, because it does not yet include owners, family members, smallholding residents or visitors. And the numerator and denominator in this calculation come from different sources. What it does show is how drastically the outcome changes depending on which group you choose as the denominator.
A little perspective
Lex Libertas’s own list of names covers victims from 1990 to 2026 (thus a period of 36 years). The 2,203 verified names on the list over that period work out to about 61 victims per year, or roughly 0.17 per day.
South Africa has remained one of the most murderous countries in the world over the same period. In 2025/26, for example, the official crime statistics recorded 23,096 murders nationwide (about 63 per day). That means the entire 36-year farm murder register, with every name, represents less than a tenth of one year’s total murder figure.
Even if one uses the police’s recent annual figure of 49 to 54 farm murders, that is still barely a fifth of one day’s nationwide murder figure, every year, over and over.
This does not mean the suffering on farms and of farmers is any less real. It does mean that a rate of 156 per 100,000, more than 20 times the national average of the time, is not reconcilable with the statistics. And that over the period, farm murders consistently made up a very small fraction of the country’s overall, exceptionally high level of violence.
Genocide?
The debate about “genocide” in turn also deserves caution. It is propagandistic when critics pretend that everyone who stresses the seriousness of farm murders is necessarily claiming a “white genocide,” and then dismiss the whole safety problem by refuting only that extreme label. AfriForum did use the language of the Genocide Convention in its criticism of the “Kill the Boer” slogan, but has not consistently described farm murders as a proven white genocide. In fact, the organisation formally objected when that position was attributed to it, and still does so whenever it happens.
The Press Council’s ruling on this dispute shows why the distinction matters. One can therefore reject the genocide label, criticise the exaggeration in some farm murder statistics, and still acknowledge that farm attacks are a serious safety problem.
Reducing the whole debate to a choice between “genocide” and “no problem” replaces analysis with emotion and propaganda.
To mourn and to calculate
Someone who has lost her father, mother and sister does not need a figure of 156 to make her loss greater; the loss is already irreparable. The crosses commemorated real people who died violently, and that deserves respect, regardless of which number is written next to them.
But respect for the dead also demands honesty about the figures spread in their name. The crosses do not prove that 3,000 white farmers were murdered, nor that white farmers experience a murder rate of 156 per 100,000. Unfortunately, Ernst Roets’s figures don’t add up for me; if, for example, there are only 3 workers for every farmer, then the death rate on farms is lower than the national average.
Whoever truly cares about farm victims should demand the same strict standard from everyone who uses their names and numbers, including, and precisely, from the campaigners on their own side.