Not Every Civil War Results in Mass Rape
“Our analysis of the behavior of the Muslim communities demonstrates that the morale, will and bellicose nature of their groups can be undermined only if we aim our action at the point where the religious and social structure is most fragile. We refer to the women, especially adolescents and to the children. Decisive intervention on these social figures would spread confusion … thus causing first of all fear and then panic, leading to a probable [Muslim] retreat from the territories involved in war activity.”
In 1991, Yugoslav special services officers provided this analysis at a meeting of army officers involved in the Bosnian conflict. This became known as the infamous “Brana Plan.” Though the quote doesn’t explicitly refer to rape, many have interpreted the meeting to be one of the only existing instances of military officers stating their intention to use mass sexual violence as a part of its overall military strategy.
Indeed, between 20,000 and 50,000 women were victims of sexual violence during the war in Bosnia. This was one of the worst cases of mass rape in the history of civil war. But it wasn’t the only one.
During Pakistan’s civil war in Bangladesh in 1974, an estimated 200,000 to 400,000 Bengali women were raped by Pakistan security forces. Between 250,000 and 500,000 Tutsi women were raped during the genocide in Rwanda. And more recently, the Burmese military has been accused of systematic rape of Rohingya women during counterinsurgency operations in Burma’s Rakhine state.
But not every civil war results in mass rape. In fact, only 64 percent of the 91 civil wars between 1980 and 2012 featured wide-scale or systematic rape, according to data from Harvard Kennedy School associate professor Dara Kay Cohen’s recent book Rape During Civil War.
It would seem from the preceding examples that mass sexual violence is a “weapon of war” specifically used to ethnically cleanse populations. Many scholars, policy experts and human rights organizations have argued as much.
For instance, here’s how the “Brana plan” worked during the war in Bosnia.
“Irregular Serb or Bosnian-Serb forces enter a village … pull several girls and women from their homes and rape them in public, thereby shaming them and creating terror among the town’s population,” Cohen writes. “The irregular forces, or ‘Chetniks,’ then depart, leaving the civilians in a chaotic state. Several days later, regular soldiers from the Serb ‘Yugoslav National Army’ arrive and offer the terrified inhabitants safe passage to a refugee camp or U.N. ‘safe area,’ if they agree to depart immediately and take nothing with them.”
The population of the village would be so terrorized by the mass rape and massacres that they would readily submit to the JNA’s request to leave. One scholar describes this “terrorize” and then “cleanse” tactic as “a highly economic weapon in the kind of ‘ethnic cleansing’ aimed at ridding a territory,” of its population.
The “terrorize” and “cleanse” tactic has occurred during other civil wars as well. For instance, the tactic seems to explain the Burmese military’s pattern of mass rape against Rohingya civilians.
Since the late 1960s the Burmese military has fought insurgents mainly by removing civilians from villages providing aid to insurgents. Doing so allows the Burmese military to cut off insurgents’ supply of intelligence, recruits, resources and a base from which to plan attacks. To clear these villages the Burmese military destroys much of the infrastructure, and either relocates the population to areas under government control, or simply forces civilians to flee.
During these operations massacres, rape, and gang rape have been common for decades. The operations against the Rohingya are no different. While forcing Rohingya to flee their villages Burmese troops have reportedly raped women accused of supporting the insurgency. There are also some reports that Burmese troops raped women after announcing that the Rohingya population must flee to Bangladesh.
But mass rape and ethnic cleansing do not always go hand-in-hand during civil war. Of the 32 civil wars featuring ethnic cleansing in Cohen’s dataset, the government committed mass rape in 17 of those cases, and rebels committed mass rape in six.
At top — refugees in South Sudan. United Nations photo. Above — a Rohingya refugee camp. Adryel Talamentes photo
Rape as genocide
A subset of the ethnic cleansing argument is that rape is a tool used to carry out genocide. Not all ethnic cleansing operations can be defined as genocide. In fact, genocide occurred in only five of the 32 cases of ethnic cleansing in Cohen’s dataset.
One scholar finds several reasons militaries utilize rape as a tool to perpetrate genocide. In some cases, women are killed through rape, either through injuries or the intentional transfer of HIV/AIDs—though reports of this are scarce.
Sexual violence seems to be used more as a tool to utterly destroy the social fabric of populations targeted for genocide. Rape is often done in front of family members or community members in order to destroy social bonds between the victims and the audience. In many societies women who are raped become outcasts, and some even die due to lack of support from their communities.
Other women are forced to become pregnant in order to ensure that future generations are identified by the perpetrator’s ethnicity. This is especially the case in societies where the child’s ethnicity is identified by the father. These women and their children often become social outcasts.
Surprisingly, however, not every genocide in Cohen’s dataset also featured mass rape. Seven out of the 19 civil wars featuring genocide involve few reports of rape, or none at all.
U.N. troops in South Sudan. United Nations photo
Rape as a tool for territorial control
Other work by scholars, and human rights organizations show that mass rape is often used by militaries attempting to control a population within a territory.
For instance, over a three-day period in 2014, Sudanese military forces raped approximately 194 women in a town called Tabit in North Darfur, according to a Human Rights Watch report.
The initial reason for the operations are somewhat unclear. However, former Sudanese soldiers’ testimonies clearly indicate rape was used as a tool to punish the population for allegedly supporting rebels.
“One soldier who later defected told Human Rights Watch that they had been ordered to search for and punish rebel supporters in the town because of information that rebel forces deployed outside the town were planning to attack the [Sudanese military] base,” according to the report. “Two soldiers who had participated in the operations said that superior officers had ordered them to ‘rape women’ because the women were rebel supporters.”
Rape by rebel and government forces in eastern Democratic Republic of Congo in the early 2000s seems to have also been intended to, at least in part, punish “individuals, families, and communities for allegedly sympathizing with or supporting the ‘enemy’,” according to a USAID report.
During its occupation of Timor-Leste between 1975 and 1999, the Indonesian military was infamous for its use of rape as mechanism to control territory during counterinsurgency operations.
“Women clearly identified as members or supporters of Falintil [rebels] were … targets of sexual violence,” according to Timor-Leste’s Commission for Reception, Truth and Reconciliation report. One women described being “detained, tortured and subjected to repeated rape because she was suspected of providing food to Falintil.”
In addition, “[w]omen thought to have strategic information about the location of [rebels], or who were thought to have links to those in the mountains, became targets of rape,” according to the report. Women would be raped during interrogation to extract information on rebels. Others would be raped prior to being forced to help search for rebels in villages.
Another common practice for Indonesian forces was to rape the wives of elusive rebels, according to the report. It seems the intention of this form of “proxy rape” was to keep fence-sitters from joining the rebellion, and to punish active rebels.
Clearly, some militaries have found rape to be a tool that helps them terrorize a population into submission. Officers seem to perceive that rape makes cleansing operations easier. While others find “terror through rape” prevents the population from collaborating with rebels, or forces the population intro collaboration with the government.
Of course, mass public executions might have the same terrorizing effect. This seems especially the case with performative executions such as beheadings, drownings, or burning people alive. However, based on the literature, mass rape should be seen as essentially an additional tool used to terrorize a population into submission.
A survivor of sexual violence in Democratic Republic of Congo. David Axe photo
Rape as a tool of combatant socialization
The “rape as a weapon of war” argument is a compelling one.
Dara Kay Cohen’s 2016 book doesn’t totally discount this argument. However, she argues that a fuller explanation for civil war rape involves answering three major puzzles.
First, all militaries fundamentally desire the ability to easily control civilian populations during civil war. Thus, using rape as a weapon to terrorize the population into submission should be quite common in civil wars. As stated earlier, large-scale or systematic wartime rape is featured in 64 percent of the 91 civil wars in Cohen’s dataset.
In addition, Cohen points out that there is scant evidence that commanders actually order their soldiers to rape. And she argues that scholars should not assume that rape was a part of an intentional military strategy simply because it was widespread, or used during cleansing or control operations.
The second puzzle is that gang rape, defined as two or more persons raping an individual, is far more prevalent in civil war than in peacetime. During peacetime, gang rape accounts for roughly two to 27 percent of reported rape, while gang rape tends to account for 75 percent of reports during wartime, according to research cited by Cohen. This means wartime rape arguments should be able to explain the group aspect of such violence, according to Cohen.
Finally, “the perpetrators of wartime rape [in many civil wars] are armed groups comprised of ordinary people, not sociopaths or members of highly trained militias.” Additionally, some of the perpetrators of rape during civil war are women. Thus, a full account of wartime rape can’t simply rely on the sociopath argument, or that men have an innate desire to rape and take advantage of the lawlessness of war.
Cohen explains wartime rape as a form of combatant socialization. “Insurgent groups that depend on abduction and states that depend on press-ganging are more likely to perpetrate rape than groups that use more voluntary methods of recruitment.”
Her argument starts with the premise that military units that are made up of forcibly recruited members suffer from poor social cohesion. Faced with the reality that poor cohesion severely threatens a group’s survival in a war zone, “[m]embers of the group form social bonds by participating in acts of rape and these bonds are strengthened and reproduced through the process of recounting the violence in the aftermath,” according to Cohen.
Cohen draws on studies from other fields to argue that the high-risk nature of rape is the mechanism that increases the social cohesion of the group. Essentially, by sharing in the same high-risk activity the group breeds “comradery through adversity.”
Cohen argues that the social cohesion formed through rape does not increase the fighting abilities of the armed group. Rather, cohesiveness simply allows the group “to increase trust among people who may otherwise be predisposed to fight each other, and to create a sense of collective responsibility that reduces attempts at desertions or mutinies and allows the armed group to endure.”
Importantly, the use of rape to socialize a group is not necessarily a conscious choice by commanders.
Instead, Cohen argues that all armed groups usually have a few men that maintain a propensity to rape. War allows these men the ability to carry out rape.
When a group is poorly socialized, other members of the group join in. The bonds formed through rape are collectively seen as a positive, so commanders allow the practice to continue, according to Cohen.
Alternatively, armed groups that go through basic training, or consistently form bonds through battle are able to form social bonds without the use of rape. Thus, even when rape occurs, it does not become a consistent practice, according to Cohen.
This is especially the case because rape is a rather costly practice in terms of battlefield effectiveness, Cohen argues. In short, men can’t fight when they contract debilitating—and frequently untreated—STIs. Cohen’s interviews with ex-combatants from the Sierra Leonean rebel group, the RUF, make this point clear.
Members of the rebel group reported not being able to fight—or even to walk, in some cases—because of the pain from STIs. Because they were based in the jungles, the RUF was forced to kidnap doctors and nurses and raid pharmacies to treat the many sick fighters.
The severity of STIs resulting from rape even led one RUF commander to kill two infected soldiers unable to flee their hideout. The commander “was concerned that the men would be tortured and killed by the army if they were left behind, and there was no way to carry them along.” Cohen reports hearing many other similar stories.
Rape, especially gang rape, carries other risks as well. For instance, gang rape takes time, and pulls combatants’ attention away from guarding their area of operation from enemy attack. Additionally, Cohen found that gang rape also has an averse psychological effect on many of the fighters.
Again, for well socialized armed groups these risks cause commanders to restrict troops from rape. But for poorly socialized groups, rape is common because of these risks. High-risk group behavior strengthens social bonds. Thus, commanders may either order the practice, or more commonly, simply allow it to continue, according to Cohen.
Cohen tests her argument both statistically and through careful case studies of armed groups in civil wars where widespread rape did and did not occur. The case studies draw on a range of sources, the most important of which are interviews with former combatants.
The statistical portion of Cohen’s book analyzes a dataset she constructed using U.S. State Department Country Reports on Human Rights Practices. Cohen and a team of researchers used the reports to identify the scale of rape in each civil war on an annual basis. She uses a four-point scale to measure the “relative magnitude of rape.”
A three on the scale was coded when a State Department reports described the rape as “massive” or “massive in scale.” Or rape was described as a “weapon of war” and described as “widespread,” “common,” “often” and other related terms.
A “two” on the scale was coded when reports describe rape as “widespread,” “common,” “often,” or other related terms. Or when rape was described as a “weapon of war” but without any of the previous descriptors.
A “one” was coded when reports note that rape was described as “isolated,” terms were used like “some reports,” etc. A “zero” was coded when the State Department recorded no reports of rape.
Obviously, measuring rape in this way has its problems. As Cohen notes, using key terms to gauge the magnitude of rape is not as fine-grained a measure as numeric measures of victims of rape. But unlike with civilian deaths, accurately counting the number of rapes in a war zone is notoriously difficult because the practice is often hidden, and reliable counts are not unavailable.
Cohen notes that relying on State Department reports could introduce some biases. Thus, she also went to great lengths to cross-validate her measures, to include checking her coding against a similar dataset on wartime sexual violence.
Cohen statistically tests her argument against several others. This includes the “rape as a weapon of war” theory discussed earlier. To test this theory, she includes variables measuring whether a war included ethnic cleansing, or genocide.
Cohen also tests the argument that poorly trained militaries and those lacking professionalism have a higher propensity to rape.
Another theory included in the statistical tests is that highly resourced insurgent groups are more likely to rape. This theory comes from Jeremy Weinstein’s book Inside Rebellion. Weinstein reasons that highly resourced rebels are more likely to abuse civilians because they are not as likely as poorly resourced groups to need willing support from local populations.
Additionally, highly resourced groups attract opportunists who are simply looking to exploit the chaos of war to enrich themselves and act as they please. Poorly resourced groups, on the other hand, only attract those devoted to the political aims of the rebellion, and are thus easier to control, according to Weinstein.
Cohen tests three additional argument. The first is that state collapse simply allows predators the free reign to rape. Another theory argues that countries with greater gender inequality see higher rates of wartime rape. Finally, Cohen tests the commonly held notion that ethnic wars are more likely to feature widespread rape.
Ultimately, she finds that forced recruitment—the key proxy measure for the combatant socialization argument—is both positively and significantly correlated with wartime rape for both government and rebel forces.
Interestingly, arguments about highly resourced rebels, and state failure are also positively and significantly correlated with wartime rape. There is no evidence of an association between ethnic war or country-level measures of gender inequality.
The “rape as a weapon of war” argument is largely unsupported by the statistical tests. Ethnic cleansing does not reach statistical significance for either state or rebel forces.
However, she does find a significant, but negative correlation between genocide and rebel-perpetrated rape in civil war. This means that rebel-perpetrated genocide actually “deceases the likelihood of rape.”
Cohen reasons that this result might be explained by the fact that rebels who perpetrating genocide are likely to see sex with the target group as a form of self-pollution.
Cohen made clear in her book and during an interview with War Is Boring that while her combatant socialization theory explains variation better than many existing theories, it doesn’t explain all instances of wartime rape.
“The type of rape I study is essentially public gang rape,” Cohen pointed out.
As noted earlier, this form of indiscriminate rape by groups of fighters is probably the most common type of rape in war. But Cohen explained that future scholarship can focus in on different types of rape in war.
For instance, the Sexual Violence in Armed Conflict dataset cited earlier codes whether widespread rape was reported to be targeted or “selective.” This means the sources that the researchers used noted whether armed groups seemed to be intentionally targeting certain groups for rape.
This includes rebel collaborators and ethnic groups. Cohen agrees that focusing in on studying the cases where rape seems to be explicitly targeted would allow researchers to better understand the “rape as a weapon of war” argument.
However, Cohen points out that the selective wartime rape that occurred in Rwanda and Bosnia was “targeted in the sense that it’s targeted towards one ethnicity, [but] it’s still not necessarily ordered … and there’s some type of social dynamic to it because it was often [reported to be gang rape].”
This means that future researchers delving deeper into the “rape as a weapon of war” theory need to take seriously the social dynamics that are so common during instances of wartime rape.