CORRECTION: A previous version of this report rated as unproven the second claim that at least half of murdered women in South Africa die at the hands of their intimate partners. This was incorrectly based on the proportion of cases where a perpetrator had not been identified. Experts involved with research on femicide have clarified that the missing data isn’t bound to change the overall pattern much.
“South Africa femicide rate is 5 times more than the global rate,” Mthethwa, the minister of arts and culture, added. Key findings from the “many surveys which have been done on this matter” included that “in South Africa, every 8 hours a woman is killed and at least half of these women die at the hands of their intimate partners”.
A reader drew our attention to the claims. Africa Check then contacted Mthethwa’s spokesman to get his source, but at the time of publication, it was outstanding.
(Note: Spokesman Lisa Combrinck has since responded, saying: “There is no further need to share the sources with you of the other references you request, as you have gone to similar sources that were used for the most part.”)
To get the latest figures on femicide, Africa Check spoke to Major-General Sally de Beer, the South African Police Service’s head of corporate communication.
Between April and December 2016, the police recorded a total of 14,333 murders, De Beer told Africa Check. Of these, 1,713 were women.
This works out to a woman being murdered every 4 hours in South Africa. (Note: When women murdered by intimate partners are considered, the most recent data points to one murder every 8 hours, as we explain next.)
De Beer told Africa Check that “the crime registrar office of the South African Police Service does not keep intimate-partner-violence statistics, as this is not a legally defined crime”.
The most recent data we have on intimate partner femicide in South Africa is based on 2009 autopsy data of women aged 14 years and older from mortuaries across the country.
Researchers, mainly from the South African Medical Research Council’s Gender and Health Research Unit, corroborated the data by interviewing police investigators “to verify the cause of death, to identify relationships with perpetrators and to collect other crime investigation data”. The sample of cases was then weighted to be nationally representative.
For the study’s purpose, intimate femicide was defined as the murder of women by their “current or ex-husband or boyfriend, same-sex partner or a rejected would-be lover”.
The published study indicated that the police could identify the perpetrator in 1,792 of an estimated 2,363 cases. Of the cases where a perpetrator had been identified, more than half of the murders (57.1%) were by an intimate partner.
Though the perpetrator was unknown in 22% of the cases, this was not thought to skew the results significantly, two experts involved with the study told Africa Check.
The researchers made the statistical assumption that the perpetrator data was missing at random and that “one would get the more or less the same intimate/non-intimate distribution in the cases with the missing data”, Prof Carl Lombard, director of the MRC’s biostatistics unit and study co-author, said.
“There are all sorts of statistical ways to deal with missing data, but if you have the majority of the data… then the results won’t change [in important ways],” Dr Edward Mills told Africa Check. He is associate professor in the department of health research methods, evidence and impact at McMaster University in Canada and edited the journal article.
The researchers’ data for 1999 showed that in 50.2% of cases where a perpetrator was identified (18.5% of the total), the victim was killed by an intimate partner. These studies “suggest a relatively stable but possibly increasing proportion of intimate partner femicides”, criminologist Anine Kriegler told Africa Check. She is co-author of the book A Citizen’s Guide to Crime Trends in South Africa and based at the University of Cape Town.
“This is unlikely to have seen dramatic change,” she noted. “I would say [this claim] qualifies as ‘correct’, given available data.”
Stephanie Burrows, a technical officer in the World Health Organisation’s Violence and Injury Prevention programme, told Africa Check that the organisation’s latest murder estimates provide data for 2015.
The organisation collected data using a questionnaire which was completed by respondents “working on violence prevention” in various ministries and institutions. Data was collected for the years 2000 to 2010.
Using this, the organisation projected murder figures for 2015. Burrows noted that the estimated global rate of femicide for 2015 was 2.4 per 100,000 women. South Africa’s rate for the same year was 9.6 per 100,000 women. This would mean that South Africa’s rate is 4 times that of the global average when considering the latest estimates.
When actual data for 2010 is considered, we see that the country’s femicide rate was then 5.1 times higher than the global average.
|Year||SA rate/100,000 women||Data source||WHO global average rate/100,000|
The rate of femicides needs to be contextualised within the broader murder rate. Abrahams told Africa Check that a high femicide rate was plausible as we “have high homicide anyway”.
The Institute for Security Studies’ Crime and Justice Information Hub manager, Lizette Lancaster, added that South Africa’s “murder rates more than halved” between 1994/95 and 2011/12. “Therefore, it is very likely that our femicide rate also decreased. Of course, the overall murder rate increased by almost 13% in the past 4 years.”
© Copyright Africa Check 2017. You may reproduce this piece or content from it for the purpose of reporting and/or discussing news and current events. This is subject to: Crediting Africa Check in the byline, keeping all hyperlinks to the sources used and adding this sentence at the end of your publication: “This report was written by Africa Check, a non-partisan fact-checking organisation. View the original piece on their website", with a link back to this page.