Medical Research Council report on intimate femicide in South Africa

Social Development

05 February 2014
Chairperson: Ms H Makhuba (IFP)
Share this page:

Meeting Summary

The Medical Research Council (MRC) firstly outlined the composition and functions of its Gender and Health Research Unit, a premier research institution in the Southern hemisphere, and noted a number of studies across a number of countries. It had undertaken a study into gender based violence and femicide in 1999, with another in 2009, focusing on intimate femicide in South Africa, and the study methodology and design had allowed for comparisons to be drawn between the two studies, to try to assess trends. It was pointed out that government departments had no reliable databases at the moment.  The study had been conducted, in both cases, at the same proportion of large, medium and small mortuaries, with interviews also with investigating officers and comparison of records at police stations.  There had been several challenges encountered, including the absence of data in many cases, with missing CAS numbers, incomplete registers at police stations that did not identify the causes of death, or link them to domestic violence, other missing information, missing mortuary files, or those that could not be traced although they were noted in the register, failure to link mortuary and police information and missing post-mortem reports. It was stressed that the findings that would be outlined were also hampered by the fact that in 18% of cases police could not identify a perpetrator, so it was not known whether it was or was not an intimate partner.

The study showed an overall reduction (measured by numbers) in female homicide, and an overall reduction in the numbers of intimate femicide cases. However, there had been a rise in the proportion of intimate femicide cases to female homicide, with intimate femicide remained the leading cause for female homicides in South Africa. It was estimated that three woman were killed by a intimate partner per day in South Africa in 2009 (which was likely to be an under-estimate given that 18% of murder perpetrators were not identified). This pointed to the fact that work being done to try to prevent intimate partner violence, and generally to prevent gender based violence, was not effective; if it had been, then a larger reduction in intimate femicide would have been seen, instead of the 1.4% increase in intimate partner violence. It was also important to note that in the case of non-intimate femicide, rape homicides had increased. There was no evidence that police investigations had improved. Convictions had also decreased. The significant reduction in gun murders (50% decrease likely) suggested evidence that the gun control legislation was working, but there was an increase in stabbings.

The 2009 homicide study had also included child homicides for the first time, although the results of this study were not being presented today, and it had determine the national incidence of child homicide, looked at gender patterns, and look to the links between homicides and child abuse.   

Members asked question about whether the causes of violence and the reasons for such brutality had been identified, the effect of societal issues on gender based violence, whether the study covered violence against men, how the information was being disseminated, and whether MRC could ensure that it formed the basis of actions taken. They asked why the requirements of the Domestic Violence Act were not apparently indicating this information at an early stage, what was found in the registers held by the South African Police Service, whether preventative measures were effective, and noted that the decrease in gun killings was no doubt due to effective gun legislation. They questioned if unemployment of perpetrators, and women being financially dependent on their partners, played a part. They also asked about collaborative research.

Members then adopted the draft Committee report on its oversight visit to Gauteng. 

Meeting report

Medical Research Council Report:  Understanding Intimate Femicide in South Africa
Professor Naeemah Abrahams, Specialist Researcher: Gender and Health Research Unit, Medical Research Council, apologised for the absence of Professor Rachel Jewkes, who was unwell.

Professor Abrahams outlined the structure of the Medical Research Council (MRC) and said that its Gender and Health Research Unit (G&HRU) was the premier research unit on gender-based violence and health in the southern hemisphere, and was led by Rachel Jewkes, a National Research Foundation “A-rated” scientist. This Unit It had been founded in January 1995 and had published over 160 peer reviewed journal articles, 30 book chapters and 50 reports. The G&HRU had undertaken and supported research in South Africa, Namibia, Uganda, Tanzania, Kenya, Ethiopia, China, Papua New Guinea, Sri Lanka, Indonesia, Cambodia and Bangladesh. Its research areas included describing the nature and magnitude of gender-based violence (GBV), understanding the health impact, developing an evidence-based health sector response to GBV, primary prevention of GBV and researching the inter-sectoral response to GBV, including the criminal justice sector.

The Unit led two global initiatives. Firstly, its Sexual Violence Research Initiative ( aimed to increase awareness of sexual violence as a priority public health problem, through evidence-based communication and information. The SVRI was a global network of over 3 000 researchers and practitioners working in the field of sexual violence and gender based violence. Secondly, it was also leading an extensive DFID programme, entitled “What works?” - Global Programme to prevent Violence Against Women and Girls”. Furthermore, the G&HRU also ran a programme on evidence based in-service training on post-rape care and Gender Based Violence. The MRC also had a number of publications that impacted on global policy on intimate partner violence.

The MRC, in doing this study on Intimate Femicide, had compared studies done also in 1999 with those of 2009 study.

She outlined the definitions used,  which included:
- Female Homicide- Killing of a woman
- Femicide- killing of a woman
- Intimate femicide (IF)-Killing of a woman by an intimate partner (current and ex husband/boyfriend/same sex partner)
- Non intimate femicide (NIF)- Killing of a woman by  a non-partner (stranger/acquaintance/family member etc )
- Suspected Rape Homicide - Evidence of rape present (forensic and/or police observations)
- Intimate suicides -Perpetrator suicide following the murder
- Convicted- convicted for homicide or culpable homicide- with the figure calculated from all those that were charged with these crimes.

Professor Abrahams noted that the killing of women by intimate partners was the most extreme consequence of intimate partner violence. South Africa had limited monitoring tools to measure the impact of the laws, policies and programmes that addressed gender based violence. G&HRU would only know if its programmes were effective if it saw a decline in GBV.

In 2004 the G&HRU released the findings from their first national female homicide study, which reflected 1 999 deaths. Of these, it was found that 50% of all female homicides were perpetrated by intimate partners, or four women killed by intimate partners per day. However, this had been an underestimate because in 18% of cases police could not identify a perpetrator.

The 1999 study had been replicated in order to achieve some comparisons over time. The second study had also been done because South Africa did not have homicide data bases that linked the victim and the perpetrator, in the Departments of Home Affairs, Police or Justice and Constitutional Development. The investigation of data from 2009 would allow the G&HRU to compare results with the 1999 study, in order to:
- determine whether there had been a shift in the overall number of females killed and females killed by a intimate partner
- determine shifts in other patterns such as mechanism of death -  in particular firearm homicides, homicides following rape, and convictions

The 2009 homicide study had also included child homicides for the first time. This would:
- determine the national incidence of child homicides in SA
- look at gender patterns (for victims and perpetrators)
- look at child (those under 18 years) homicides associated with abuse (sexual and non sexual abuse) However, that particular study was not being presented today.

Ms Bongwekhazi Rapiya, Researcher, G&HRU, presented on how permission was obtained for the study, and the study design. The 2009 study design was similar to the 1999 Female Homicide Study. The mortuary-based prevalence study, with all mortuaries in country included in the sampling frame, was chosen to allow for a national estimate. Mortuaries were stratified based on size (as with the 1999 study), but 38 mortuaries were sampled in 2009, compared to 25 in 1999, which reflected a change in the service composition. The 2009 Sample studied five out of the eight large mortuaries, 13 out of the 33 medium size mortuaries, and 20 out of the 81 small mortuaries

She also spoke to the three phases of data collection. The first was collection of data from mortuaries, the second phase was collecting data from the South African Police Service (SAPS) and the third phase was using the blood alcohol results from laboratories.

Professor Abrahams presented the results. There had been a decrease in female homicides between 1999 and 2009. There had been a decrease in the numbers of intimate homicides as well, although this had not decreased by the same degree as the overall female homicides. However, the proportion of intimate femicides to female homicides had increased by 6.5%. She noted the population rates for 1999 and 2009. She said that these figures showed that there was, in 2009, a 2.6% greater likelihood of a rape homicides among non intimates in 2009, compared to 1999. The mechanism of death results showed that killings by guns had decreased, however deaths caused by stabbing had increased. There had been a 50% decrease in the likelihood of a gun-related homicide in 2009, compared to 1999. There had been a 30% decrease in the likelihood of a conviction among non intimate femicides in 2009, compared to 1999. There had been an overall increase in intimate partner violence by 1.4%.

Challenges in conducting medico-legal research included missing mortuary files, and at some mortuaries files could not be traced (after cases were identified in the register). There had also been missing Post-mortem (PM) reports. In some cases PM reports were handed to the police, but no duplicates retained in mortuary files, and these outstanding PM reports had delayed the criminal investigations. Outstanding toxicology results had meant that in a number of cases investigations were pending the outcome of the results, and no investigations had been conducted or cases opened. There had also been incorrect CAS numbers documented. Deaths at hospitals had often meant that cases had never been opened.

The challenges in collecting data from SAPS included missing CAS numbers, as well as the fact that some police stations had not identified these cases in their registers, or information was missing from files. She noted that it was critical to establish linkages between mortuary data and police cases. At one Eastern Cape mortuary,  77cases had been identified but CAS numbers had only been available for seven of these cases. A Gauteng mortuary had had 18 cases where the police station number and CAS numbers were missing. Much time and effort had to be put into tracing these cases.

She said that all regions provided support to the researchers, to access police stations and investigating officer (IO) data. This support had been essential in tracing cases and decreasing the number of cases that initially could not be traced. It had been a challenge to set up interviews with the investigating officers. There were also several challenges around obtaining the blood alcohol concentration (BAC) results from forensic laboratories. In Gauteng, North West and Free State, the laboratory servers had crashed in March 2012, and the data was currently being retrieved.

She concluded by saying the comparison of two identical surveys – ten years apart - had provided G&HRU with an opportunity to monitor GBV. She reiterated that there had been an overall reduction in female homicide. However, the numbers of intimate femicide cases had decreased to a lesser extent. Intimate femicide remained the leading cause for female homicides in South Africa. It was estimated that three woman were killed by a intimate partner per day in South Africa in 2009, but she reiterated that this was likely to be an under-estimate, since one fifth of all murders never had their perpetrator identified. There was evidence that work being done to try to prevent  Intimate Partner Violence and Gender Based Violence was not effective, for a bigger reduction in Intimate Female murders should have been recorded. It was also important to note that in the case of non-intimate Femicide, rape homicides had increased. There was  no evidence that police investigations had improved. Convictions had also decreased. The  significant reduction in gun murders suggested evidence that the Gun Control Legislation was working. However, the increase in stabbings was noted again.

She suggested that homicide surveys in the future should use this survey as a monitoring tool in the absence of good national data. This was because it provided insight on progress made towards gaining gender equity in South Africa. G&HRU recommended that the survey be repeated every five years from 2014 onwards.

The Acting Chairperson thanked the presenters and stated that this was a problem that Members hoped could be tackled as a collective.

Cause and brutality of killings
Mr M Magagula (ANC) asked if the research looked at the cause of ‘family related homicides’. If the MRC was looking at the causes then it should surely be suggesting some answers.

Professor Abrahams replied that this study was about numbers and was merely identifying the size and descriptors of the problem. She said, however, that other research had been done on what contributed to GBV, in a number of other studies. One of her colleagues had visited correctional centres to ask about femicide, and she had concluded that there was no one single cause of why men abused women but it was rather a multi-faceted issue. She said that although some key issues had been identified, it would take much time to talk through the root causes of violence. These causes were the same across many parts of the world.

Mr Magagula asked what the reason was for killing and raping so brutally, and wondered if this was perhaps due to substance abuse. He suggested that issues of masturbation and devil worship needed to be looked into.

Mr R Bhoola (MF) also spoke to the unregulated porn industry and its effect.

Professor Abrahams replied that she did not have the whole answer on this issue. She agree that there had been links established between viewing porn and an increase in masturbation (at a later stage), which was sometimes identified as a cause of rape. Studies showed that masturbation by children should not always be treated as a negative thing, but increased instances of masturbation at the later stages of life could have a negative effect. She reiterated that this particular study had not looked at the reasons people killed, but looked more towards number and description. The MRC had come across witchcraft killings, but not necessarily those linked to devil worship.

Men being abused
Mr M Magagula raised the issue that men were also being killed by women.

Professor Abrahams replied that GBV spoke to violence that happened according to a woman’s gendered position so it focused on the position of women. She understood that men were also victims of violence within society, but in these cases, gender was not the primary construct that caused the violence, and it was thus not properly described as falling within “gender based” violence.

Dissemination of information
Ms N Gcume (ANC) asked if the manual produced by MRC had been shared with other entities such as NGOs within the urban and rural areas. She asked if there had been instances of capacity building and awareness raising within these communities

Ms F Khumalo (ANC) asked if this information was being spread to others

Professor Abrahams replied that the information had been disseminated, including to other departments such as the Department of Health. The information was not always utilised and the MRC could not force entities to use the information, given its position as a research council.

Ages of intimate violence victims
Ms Gcume asked for the age ranges of the victims.

Professor Abrahams replied that the ages for intimate violence victims generally ranged between 34 and 40 years.

Preventative measures
Mr Bhoola said he was concerned that preventative measures were not effective. He asked if the MRC could share some innovative ideas about what could be put in place to lower the instance of such violence. He suggested that some of the issues that affected people could be  related to the erosion of family values or the huge rise in the use of drugs in communities.  

Psychology of men
Mr Bhoola asked if the psychology of men was being dealt with, noting that when some perpetrators entered the parole system then they often committed another similar crime.

Professor Abrahams replied that it was unfortunate that this was the case, and that often men were imprisoned again because in their first incarceration they had not been rehabilitated.

Position of SAPS
Mr M Waters (DA) asked if MRC had raised with SAPS the fact that statistics were not readily available. He said that surely the Domestic Violence Act compelled police to record the statistics of domestic violence. This led to the question of how well the domestic violence register was being implemented across the country, and whether that register was capable of being used as an instrument to effectively record these matters.

Professor Abrahams replied that since 1999 the MRC had said to SAPS that there would have been no need for another body to undertake studies, if SAPS simply recorded, in some way, that cases involved intimate violence. A further problem was that the Domestic Violence Act registers only recorded what was happening with “live bodies” but when a woman died as a result of this violence, the case was often disconnected from the Act.

Unidentified perpetrators, decrease of convictions and weapons used
Mr Waters said he was worried to see that the number of unidentified perpetrators had increased to nearly one in four, and asked why this was the case.

Professor Abrahams replied that she did not know why this had been the case.

Ms Khumalo asked why there had been a decrease in convictions.

Professor Abrahams replied she did not know but again, this was worrying.

Mr Waters said it was interesting to see the drop in the number murders caused by guns, which showed the positive impact of good gun legislation.

Underage girls drinking
Mr Waters asked how many underage girls had been drinking when they became victims.

Professor Abrahams replied she did not know the answer for this, as the study had not concentrated on this area.

Unemployment and femicide
A Member raised the point that there had been suggestions that high numbers of rape cases were related to high unemployment, and questioned if this was correct.

The Acting Chairperson asked if there was anything in the MRC research that looked at the dependency of women on their abusive spouses.

Professor Abrahams stated that unemployment had not been shown to be a significant cause, as often perpetrators were employed. It was sometimes a case that women were originally attracted to men by the money they had, and sometimes men used their income as a form of asserting their masculinity.

Minors and dependency of women
The Acting Chairperson asked if the MRC researched minors being sexually abused.

Professor Abrahams replied that a similar review was being conducted for children.

Collaborative efforts
The Acting Chairperson asked how the research done could assist other departments such as the Department of Social Development, and if MRC’s research could help government overall in building databases. She argued that there should be forums set up where stakeholders came together to show what had been done and to decide upon a possible way forward. She felt that researchers at the MRC should not simply research but make sure the findings were used.

Professor Abrahams replied that MRC often thought about how its research was used and how it could be more useful to government. She said that there was a need to talk to people who could bring about change and MRC could bring the necessary ammunition to effect changes. The discussion needed to be on who was responsible for formulating the plan and who was to monitor it.

Other Committee Business: Adoption of draft Report on Oversight Visit to Gauteng
Members agreed, having studied the draft Report on the Committee’s oversight visit to Gauteng, that it was comprehensive and included the correct recommendations.

Ms Khumalo said that the recommendations needed to be forwarded to the office of the Minister.

The report was adopted.

The meeting was adjourned.


  • We don't have attendance info for this committee meeting

Download as PDF

You can download this page as a PDF using your browser's print functionality. Click on the "Print" button below and select the "PDF" option under destinations/printers.

See detailed instructions for your browser here.

Share this page: