Last month, Human Rights Watch released a report confirming that the Burmese security forces “have committed widespread rape against women and girls as part of a campaign of ethnic cleansing against Rohingya Muslims in Burma’s Rakhine State” since 25 August 2017. The report found that the actions of the military, border police and ethnic Rakhine militias amount to crimes against humanity under international law.

Although there is no legally agreed definition of ethnic cleansing, the description developed by a UN Commission of Experts holds significant sway. They described ethnic cleansing as ‘a purposeful policy designed by one ethnic or religious group to remove by violent and terror-inspiring means the civilian population of another ethnic or religious group from certain geographic areas.’ The events of September and October have certainly served to remove Rohingyas from northern Rakhine State.

In early December, the International Committee of the Red Cross, one of the very few international organisations with access to effected areas inside Myanmar, reported that “formerly energetic communities and village tracts are suddenly empty. Life continues for those that remain, but in certain parts of Maungdaw and Sittwe, there is a pervasive sense of absence.”

MSF have now also released extensive testimonies from survivors on the killing, arson and sexual violence they have experienced. Ninety percent of the survivors of sexual violence treated by MSF were attacked after 25 August. Fifty per cent of survivors are under the age of 18, including several under the age of ten.

Early reports of the number of pregnant women and new mothers in the refugee population could have been used as an indicator of increased conflict related sexual violence and ethnic cleansing. While the accuracy of the data and a heightened tendency for pregnant women to flee may both affect the analysis, the matrix of indicators of conflict-related sexual violence developed in response to UN Security Council Resolution 1888 identifies an increase in cases of unwanted pregnancy as an indicator of ongoing sexual violence. International non-governmental organisation Ipas, began increasing its response to the crisis, providing trainers to ‘provide on-site training for health workers in postabortion care’.

That same matrix also reminds us that the increasing ‘statements from doctors, war surgeons, gynaecologists and/or medical NGOs that they are increasingly seeing rape-related injuries’ are an indicator of ongoing sexual violence. As early as September, doctors from the International Organization for Migration, as well as a range of UN agencies and non-governmental organisations, reported high numbers of patients with physical injuries that are consistent with violent sexual attacks, including forced penetration and lacerations to the vagina.

In the two weeks immediately proceeding the crisis, the lead UN agency on sexual and gender based violence (UNFPA), provided services to 3500 Rohingya refugee women who had been sexually assaulted. It is incredibly difficult to gather large-scale data on sexual violence in emergencies, but we do know that only 7 percent of women subjected to sexual violence during the conflict in East Timor reported it, and only 6 percent of rape victims during the Rwandan genocide sought medical treatment. If the women and girls who have reported to those health clinics represent 6 percent of victims, they would be the tip of just one iceberg comprising 58,300 women and girls.

If we had used a gendered lens to analyse the unfolding crisis from the outset, we would have had a better comprehension of the ethnic cleansing that was occurring from the outset. This could have better informed humanitarian and international legal responses. A new comment in the Australian Journal of International Affairs unpacks reported figures of pregnant women who are seeking refuge in Bangladesh in an attempt to understand some of the gendered dimensions of the conflict. It proposes possible reasons for the presence of a high proportion of pregnant and lactating women in the refugee population, and goes on to reflect on indications of increased conflict-related sexual violence and ethnic cleansing. It shows that, while failings in the quality of data in emergencies mean it cannot be relied on as the basis for rigorous conclusions about the gendered nature of conflict, when taken with qualitative reports, and compared with other emergencies, gendered data can be used to build a better understanding of the conflict.

In the first two weeks of the Rohingya crisis, UNICEF reported that an unprecedented portion of the refugees fleeing to Bangladesh were children. The Chief of Child Protection for UNICEF in Bangladesh, Jean Lieby announced that preliminary data showed that 60 percent of the arriving refugees were children, who were often unaccompanied. Such extreme family separation can be an indicator of the degree of chaos and of rate of adult deaths. This second indication is reinforced by the fact the next largest age group of refugees are the elderly. UNICEF also reported that 67 percent of the refugees are female. Combined, this could indicate fighting age males had been targeted in Myanmar.

In mid-September, the Bangladesh Ministry of Health reported that approximately 70,000 of the Rohingya refugees who have arrived since August were pregnant or new mothers. This would represent a staggering 20.8 percent of the female population. Despite the high birth rates among Rohingya communities, we know that Rohingya women have an average of 3.8 children in their lifetime, we would expect only 6.9 percent of the female Rohingya refugee population to be pregnant or breastfeeding. The Bangladesh home minister has said that 90 percent of the refugee women have been raped. That would equate to over 335,600 people.

Peter Bouckaert, emergencies director at Human Rights Watch, said that “the level of hatred and extreme violence—especially towards women and children” is driven by dehumanisation and racism. Because the Rohingya have been described as “too dirty” for soldiers to rape, he believes there is no doubt that “the majority of the women who were raped were killed.” The organisation has drawn connections between what is happening to the Rohingya in Rakhine State and what occurred during the genocide in Rwanda.

Indeed, a comparison of the above data on the Rohingya refugees to that which we saw in the aftermath of the Rwandan genocide provided an early indicator of ethnic cleansing in Rakhine State. UN reporting in the aftermath of the Rwandan genocide indicated that the genocide had so rapidly altered the demographics that 60–70 percent of the population was female. It was estimated that the Rwandan refugee population exceeded pre-war fertility. In the immediate aftermath, this was presumed to be the case because of the high number of men who were killed during the genocide. Adolescent and adult males under the age of 45 were the primary targets in the early stages of the Rwandan genocide. There were also “indications that attempts to exterminate women, girls and the elderly eventually encountered significant popular opposition”. But sexual violence was a key feature of the Rwandan genocide. Although, as in the Rohingya case, the majority of rape victims were then killed, most recent estimates indicate that in excess of 20,000 Rwandan children were born from genocidal rape.

It is hoped that the new comment in the Australian Journal of International Affairs, and the general analysis described here, will serve as a reminder of the importance of gendered, cross-disciplinary research to accurately understand forces of peace and conflict in the world, and to inform appropriate policy responses such as humanitarian assistance and international legal action.