Earlier this month the Afrikaans pop singers and activists Steve Hofmeyr and Sunette Bridges, along with Amanda de Lange of the Freedom Front Plus, laid a complaint to the South African Human Rights Commission against the fact checking publication Africa Check and the Medical Research Council over claims to the effect that white women were more likely to be murdered by their (white) male partners than unknown (often black) men.

The complaints related to two claims made by Africa Check researchers in response to Hofmeyr's allegation, made on his website and Facebook pages in the middle of 2013, that Afrikaners were being "killed like flies" under the ANC government.

In an article which appeared on Africa Check on June 24 2013 Nechama Brodie took issue with Hofmeyr's claims in a long feature article. She was subsequently quoted in Rapport, the following weekend, as saying "A White woman has a better chance of being murdered by her lover or partner than by an unknown Black man. We should not overlook the real problem."

In an article published a few weeks later by Africa Check Lisa Vetten repeated this claim, citing MRC research on intimate femicide in South Africa. She wrote: "Claims that white women are likely to be murdered by "unknown black males" amount to racial scare-mongering. The vast majority of women who are murdered in South Africa die at the hands of their husbands, boyfriends and lovers."

These two statements are, Hofmeyr, Bridges and De Lange claim in their complaint, "inaccurate and devoid of any statistical or actual, factual research, but an offensive insult to ‘white' South African men in general." In their support of these claims they wrote that they have been able to track down only 49 cases of white males killing their female partners from 1996 to the present, compared to hundreds of documented cases where white women have been killed by "unknown" black men. They appended a list of these alleged murders to their complaint.

Rapport this weekend ran an article stating that this list was full of mistakes. In a post on the Mail & Guardian's Thought Leader website this weekend Thorne Godinho also criticised Bridges, Hofmeyr and De Lange for their complaint to the SAHRC. He stated that "It is problematic - and startlingly racist - that people are promoting the notion that white men are less likely to hate women or engage in acts of hatred against women."

Yet before dismissing Bridges and Hofmeyr's complaint out of hand it is worth examining the basis of Africa Check's claims.

Vetten in her article cites as her authority two MRC studies on female homicide in South Africa carried out in 1999 and 2009. The 1999 study estimated, Vetten wrote, that "approximately half of all women murdered in 1999 died at the hands of their partners." The 2009 study, found that due to a decline in murders of women by strangers, "intimate femicide had become the leading cause of female homicide in South Africa."

Hofmeyr and Bridges' claimed at the time to have counted 55 white women killed by unknown black males between May 2012 and June 2013 as compared to just one case of a white woman being killed by their partner (Oscar Pistorius' shooting of Reeva Steenkamp.)

Vetten dismissed this argument, based as it was on a non-representative study, saying it "cast no doubt on the research showing the majority of women to be murdered by their intimate partners and, by extension, that the majority of white women are murdered by their white male partners."

The obvious question here is whether the conclusions of the overall conclusions of the MRC studies can be applied to white women (and their white male partners) in particular?

The 1999 study was conducted retrospectively in 2002 and 2003 by sampling national mortuary data. In about 70% of the 1 052 cases the researchers sampled were they able to establish from police dockets the likely relationship between the female murder victim and the killer. The MRC estimated that overall there were 3 793 murders of women in South Africa in 1999.

The following is a table of the MRC's estimate of the total number of "Murders by Intimate Partners" (Intimate Femicide) vs the "Murders by Others" (Non-Intimate Femicide), in 1999, broken down by race[1] The table from which it is extracted is reproduced below. This, incidentally, comes from a paper on which Vetten is listed as a co-author:

As can be seen although black African, Coloured and Indian women were, according to the MRC's estimates, more likely to be murdered by their partners - than by others - the same did not apply to white women. This group was, according to the study, over twice as likely to be murdered by "others" than by their partners. (It would be surprising if there was no divergence simply on the basis of the differing age profile of the white population compared to the black and Coloured population.)

The later MRC survey, based upon a sample of 903 cases gathered from 2009 mortuary data, claimed a decline both in the number of estimated murders of women by their partners (from 1 349 to 1 024), and the estimated number of murders of women by others (1 335 to 768). The MRC said there had been 2 363 murders of women in 2009 - a decline of 36,7% from their 1999 estimate.[2] The number of murders of white women by their partners was estimated at 28, down from 53 in 1999. However, this survey did not provide a racial breakdown of the murder of women by others - as they had done in the 1999 study.

The reason for this, according to Naeemah Abrahams of the MRC, was due to the small sample sizes. She told Politicsweb "When we explored the calculation of the rates [of non-intimate femicide] by race we found the estimates became unstable because of the small numbers and we preferred to not present rates where the confidence intervals were wide. This would have been the case for Indian and white women mainly."

Thus, while the survey of 1999 murders supported the view that white women were, uniquely, much more likely to be murdered by others than their partners, the 2009 survey had no reliable results on this either way. If the estimate of 28 white women killed by their partners in 2009 is accurate, then it is likely that the findings of the 1999 survey still hold given the (apparently greater) number of cases of murder of white women by others that made their way into the press in that year.

The two MRC surveys cited by Vetten and Africa Check as their authority thus do not support the claim that the "majority of white women are murdered by their white male partners."

What one can say, assuming the 1999 MRC survey estimates are accurate, is that in 1999 white women were much less likely to be murdered by their partners than black African and Coloured women and somewhat less likely to be murdered by others.

Footnotes:

[1] Naeemah Abrahams, Rachel Jewkes, Lorna J Martin, Shanaaz Mathews, Lisa Vetten and Carl Lombard, "Mortality of Women From Intimate Partner Violence in South Africa: A National Epidemiological Study", Violence and Victims, Volume 24, Number 4, 2009

[2] According to SAPS statistics the total number of murders during that period declined by 25,5% in that period - from 22 604 in 1999/2000 to 16 834 in 2009/2010. There were 2 457 reported cases of murder of adult women in 2009/2010, 14.6% of the total. There do not appear to be SAPS statistics for the number of women murdered in 1999/2000.

Click here to sign up to receive our free daily headline email newsletter