Meredith Loken. Security Studies. Volume 26, Issue 1, January-March 2017.
Conflict-related rape is a critical issue of human security and a threat to post-war prosperity. Yet we know little about why it occurs and what can prevent it. Patterns of rape vary widely in scope, scale, purpose, and form between and within conflicts. Recent studies find that state forces are more likely than non-state armed groups to rape noncombatants. One explanation focuses on the participation of women in rebel organizations. Some scholars posit that the substantial inclusion of women in armed groups discourages wartime rape. Elisabeth Jean Wood directly makes this link, asserting, “Armed groups with a high proportion of female combatants engage less in sexual violence.” Women are more likely to fight in irregular and otherwise non-governmental forces than in state militaries or paramilitary groups. Female-intensive insurgencies, including those in Sri Lanka and El Salvador, employed low levels of rape against noncombatants or prohibited it altogether. Some women note fear of military rape as part of their motivation for joining rebel organizations. The simple conclusion is that rebel groups commit less rape because they have higher proportions of female cadre. A number of scholars and international organizations encourage the increased participation of women in public militaries to combat rape and other forms of civilian violence for this reason.
This article challenges that argument, finding that the participation of female fighters has no significant impact in constraining an armed group’s propensity to rape. Using an original dataset of women’s involvement as combatants in civil wars from 1980 to 2009, I test the relationship between women’s participation and rape. Ordered probit and rare events logistic regression models demonstrate that conflict-related rape is as likely in female-intensive groups as it is in those with low proportions of women. Neither does women’s participation significantly impact the severity of rape; female-intensive organizations are no less likely to commit high levels of rape during conflict.
This finding is important. It disrupts the broader intuition of how gender matters to security. It further problematizes what R. Charli Carpenter calls the “strategic frame” of women as a vulnerable population at the margins of war. As Iris Marion Young notes, the logic of masculinist protection of women and children “illuminates the meaning and effective appeal of a security state.” Security studies broadly neglects the empirical reality of gender. Most medical and population-based surveys of rape do not ask respondents for their assailants’ gender. Much scholarly work on even the most studied cases of conflict-related sexual violence fails to note the range of women’s experiences. Amidst the narratives of women suffering brutal rape in the Democratic Republic of the Congo (DRC), women’s participation as combatants and rapists has been largely erased. Lynn Lawry and colleagues estimate that female fighters compose nearly 48 percent of combatants in the DRC. Women may participate in 90 percent of armed groups.
Not only are many of these armed groups implicated in sustained rape campaigns in the country’s eastern region, but an estimated 41 percent of female and 10 percent of male rape victims reported being victimized by female perpetrators or mixed-gender groups. In a piece for Time, Jessica Hatcher interviews women raped by women in the DRC. Marie, then thirty-six, notes that the female perpetrators fought with male fighters over who would have her and, upon winning, sexually and psychologically abused her for four days. She remembers, “When I saw women, I thought I was saved.”
I refute the outmoded argument that the presence of female combatants lessens the likelihood of rape. This work will be of interest to scholars of human and international security intent on understanding and preventing civilian targeting and gender-based violence in conflict areas. Empirically sound research on women’s participation in armed conflict and wartime rape is critical to the development of disarmament, demobilization, and reintegration (DDR) programs; clinical, psychosocial, and legal interventions for survivors; rape prevention; international judicial development; and new theoretical investigation of the causes and consequences of wartime sexual violence.
The assumption that women’s participation lessens violence draws heavily from the false notion that women’s presence in combat roles is infrequent. This article begins with a discussion of conventional depictions of women as peripheral war actors and an overview of women’s empirical contributions to civil conflicts over the past three decades. It then explores the proposed theoretical mechanisms by which women are expected to mitigate conflict-related rape. A series of large-n cross-national studies using new data tests the hypothesis that female-intensive armed groups commit less rape, followed by an interrogation as to why women’s participation does not discourage rape. I argue that organizational factors, not individual characteristics, drive violence in armed groups. Female fighters do not reduce rape because women are subject to same organizational pressures as male cadre. The work concludes with a discussion of the empirical, theoretical, and policy implications of my findings and offers recommendations for future research.
Women in War
War is a forum of deeply entrenched gendered essentialisms. The theoretical frames of female civilians and male combatants classify women as exclusively innocent and vulnerable during conflict. As Carpenter notes, women’s permeation of the combatant sphere and the civilian status of many men empirically distort these camps. Maintenance of the war apparatus requires the desire to protect civilians. Historically, women have been valorized as objects of protection: vessels of the nation, symbols of state history, and trophies of masculine success. Women’s place as “the girl worth fighting for” is perpetual in narratives of conflict. Violent women interrupt this stereotype—they are not “the peaceful, war-resistant, conservative, virtuous and restrained women that just warriors protect from enemies.” In reality, the relationships between women and conflict are complex.
Popular response to women’s political violence is a melding of “awe-inspired fascination and deeply disdainful judgment.” As demonstrated by Wafa Idris of Palestine, Pauline Nyiramasuhuko of Rwanda, and Lynndie England at Abu Ghraib, the worldwide captivation with violent women illustrates our collective inability to imagine women as militarized. The very need to specify female combatants is indicative of the symbolic position of women as unconventional actors during conflict. In an article on the Rwandan genocide, Nicole Hogg asks if female participants are “mothers or monsters,” illustrating stark categorical exclusivity. As Laura Sjoberg and Caron Gentry contend, “Women’s violence is often discussed in terms of violent women’s gender: women are not supposed to be violent.” Often dismissed as deviant and depraved, women’s politically violent contributions are largely ignored or entirely erased from historical memory.
Hyper-feminized in public view, female fighters are frequently depicted with intense focus on their physical attractiveness and sexuality rather than their political autonomy. As bell hooks argues, “such a devaluation of the [militarized] roles of women constructs a false notion of female experience.” News media scrutinized women of the Liberation Tigers of Tamil Eelam (LTTE) as irregularly masculine and questioned their virginity. A VICE News documentary on the heavily female and socialist Party of Free Life of Kurdistan (PJAK) in Iran invites viewers to come along for a “springtime frolic with the lovely lady guerrillas of the Kurdish Liberation Movement.” It suggests that “few things unsettle the male mind like a lady in arms.” In a 2008 National Public Radio (NPR) interview, the male host asks a Liberian ex-combatant, “And I have to say that you’re very feminine, you know you have a lovely face. You’re lovely—you’re a very lovely woman. But do you remember why you became a soldier?” Western commenters appear equally infatuated with the “jihadi brides” and “terror groupies” supporting the Islamic State (IS) and IS’s all-female Al Khansaa Brigade. A photo essay titled “10 smokin’ hot, trained killers of the Israeli army” entices readers to “meet the armed and dangerously sexy women [from] the northeast side of the Gaza Strip.” Even in wartime, women are not permitted to be dangerous, unless they are dangerously sexy.
This perception is not limited to media portrayals of violent women, but is internalized by military institutions. In 2005, the Russian Ministry of Defense held the second Miss Russian Army competition. The pageant, named “Beauties in Shoulder-Straps,” was in part a bid to solicit male recruits. According to one contestant, “The main goal of each girl is to increase conscription to the armed forces.” A Defense Ministry spokesman noted, “When the girls arrived in Moscow … the young men who were there were climbing the walls and saying: ‘We’re running straight to the military commissariat, we want to serve!'”
Politically violent women are spectacles for public critique and consumption, but this misrepresents their direct involvement in civil conflict. Data suggests that women composed a substantial proportion of armed combatants in nearly a quarter of civil wars fought over the past thirty years. Women’s involvement in non-combatant and support roles is even greater. Women are often essential members of combat units—the Kurdistan Workers’ Party (PKK) would likely be unable to sustain operations without female fighters. Between 25 and 30 percent of the Eritrean People’s Liberation Front (EPLF) and a third of fighters in the Tigrayan People’s Liberation Front (TPLF) were women during the Ethiopian Revolution. Dara Kay Cohen estimates that female combatants composed 24 percent of the Revolutionary United Front (RUF) in Sierra Leone’s civil war.
During El Salvador’s civil war, women composed nearly a third of Farabundo Marti National Liberation Front (FMLN) militants. Women also constitute more than 30 percent of the Revolutionary Armed Forces of Colombia—People’s Army (FARC-EP) and other leftist groups in the country. The communist Naxalites in India boast over 33 percent female membership and afford several high ranking combat positions to women. Despite media focus on women in Sub-Saharan African rebel groups as exclusively sex slaves, 71 percent of women who went through DDR programs in Liberia and approximately 60 percent of young female survey respondents in Uganda reported their primary or secondary role as combat fighters. PJAK guerrillas claim gender parity in membership. The organization markets itself as fronted by women soldiers drawn to the group’s radical feminist underpinnings, vaunting that its women are among the fiercest and cruelest of all fighters.
Theorizing the Women Question
Despite women’s violent participation in domestic conflict, some scholarship suggests that women have a pacifying effect on armed organizations. A number of scholars hypothesize that a high proportion of female combatants may reduce, eliminate, or otherwise impact rape perpetrated against civilians. United Nations Security Council Resolution 1325 (UNSC 1325) of 2000 emphasizes conflict’s disproportionate impact on women, particularly in terms of rape and sexual violence, and calls for gender mainstreaming in armed institutions as a solution:
Reaffirming the important role of women in the prevention and resolution of conflicts and in peace-building, and stressing the importance of their equal participation and full involvement in all efforts for the maintenance and promotion of peace and security, and the need to increase their role in decision-making with regard to conflict prevention and resolution…. [The Security Council] Urges Member States to ensure increased representation of women at all decision-making levels in national, regional and international institutions and mechanisms for the prevention, management, and resolution of conflict…. Further urges the Secretary-General to seek to expand the role and contribution of women in United Nations field-based operations, and especially among military observers, civilian police, human rights and humanitarian personnel.
US Senator Barbara Boxer recently introduced legislation to actively recruit women into military and security forces because, as she notes, “when women are deployed … there are fewer allegations of sexual abuse and exploitation.” Claudia Card argues that women should construct independent military organizations in order to disrupt the symbolic importance of gender and rape during war. She contends, “One way to undermine [rape] is for women to have the same access to weapons and military training as men have presently…. Suppose women entered military institutions in large numbers … what is the likelihood that males would rape in war if they fought side by side with equally trained and armed females?” Wood similarly notes, “An army for whom females comprise a high fraction of combatants may be constrained in its use of sexual violence.”
Marie’s introductory comments demonstrate that this view is internalized among civilians as well. In a similar story, a mixed-gender group in the DRC raped Valarie, then 17, so brutally that her uterus was destroyed. In an interview, she noted that the assault “is an unforgettable wound” because rape by women is so “incomprehensible.”
Why should female-intensive armed groups be expected to commit less rape? Theorists provide four mechanisms: the relative passivity of women, the substitution of civilian victims for in-group cadre, the progressive political objectives of female-intensive armed groups, and women’s disruption of misogynistic socialization practices. The following section reviews the theoretical and empirical evidence for each.
The Civilizing Effect
The most traditional explanation for women’s perceived mitigating impact on war is that women as a gender are less bellicose than men. The idea that women have a natural affinity for non-violence has a long history echoed in attempts to include and restrict women from conflict. Mahatma Gandhi writes, “If non-violence is the law of our being, the future is with women.” Francis Fukuyama argues that women are not biologically predisposed to intense competition for dominance and that “in no area is sex-related difference clearer than with respect to violence and aggression…. What is bred in the bone cannot be altered easily by changes in culture or ideology.” In a speech at the United Nations, war historian Gerard De Groot contends that unlike men, women are not naturally inclined toward violence, and he advocates the recruitment of women into peacekeeping missions for this reason. Citing their “calm and conciliatory” effect, he suggests that female fighters should be recruited but must remain “womanly.” Carol Cohn and colleagues note that in UNSC 1325, “women are essentially victims, peace-builders, and peace-makers.” The UN justifies female inclusion in war through women’s value as pacifiers.
Literature of the gender and international security tradition notes the dichotomy between women as mothers of the nation and men as their protectors. Judith Hicks Stiehm argues that the state’s monopoly on legitimate violence depends on this segregation, She contends that state interest in maintaining this dichotomy helps explain why comparatively few women act as defenders in state militaries or use weapons. Some scholars and feminist-pacifist organizers believe that potential for motherhood results in natural passivity. Discussing female fighters of the Rwandan genocide, Dr. Ezechiel Sentama of the University of Rwanda Centre for Conflict Management notes, “I was able to see it was really bad women, I could say that they forgot their role as far as the Rwandan culture is, but actually the role, the role of woman we know. A woman is a mother. She is somebody who seems to have much more compassion than men, you know. So they forgot their role of women.” Sara Ruddick concludes, “Although mothers are not naturally peaceful, maternal practice is a ‘natural resource’ for peace politics.” She calls peacefulness a trait “latent in maternal practice.”
Women have engaged in active non-violent resistance in civil wars spanning decades, including in Palestine, Tibet, and Eritrea. In East Timor, women involved in active non-violence testified that they felt these duties were more important than combat roles. Opinion polls also demonstrate that women are less likely to support war than men.
Substitution
A theory of substitution proposes that male fighters in female-intensive armed groups are less likely to abuse civilians because they have sexual access to women in their unit. Card makes a strategic argument: if women entered military institutions en masse, there would be fewer civilian women to rape. However, most theories of substitution assume that rape is rooted in sexual desire and suggest that female combatants are either sexually violated or willing sexual participants and thus satisfy men’s need for sex. Jonathan Gottschall argues that sexual desire plays a causal role in wartime rape, noting that young women are most biologically sexually desirable and subsequently most likely to be sexually violated. Gottschall explains wartime rape as social influences mitigating a biological need: biology is shepherded through approved sociocultural avenues. Substitution theories suggest cultural norms will direct men to meet biological needs by raping women in their units instead of civilians.
Some military officials assume that an inadequate supply of sexual partners leads to rape. In Nanking (now Nanjing), China during the Second Sino-Japanese War, Japanese commanders established military brothels in an attempt to offset widespread rape. According to Wood, at least one military official argues that rape in the eastern DRC is pervasive because fighters are too poor to pay sex workers. Michele Leiby and Cohen note forced marriage in the LRA as another example of in-unit abuse dissuading civilian targeting. The high prevalence of “blue on blue” rape within the US military may also support a substitution explanation, and some have argued that women’s increasing military presence invites more in-unit rape in general.
Group Type and Objective
Women may simply be more likely to join militaries that espouse a feminist or emancipatory agenda. Violent political organizations supporting gender equality platforms are more likely to include women in their ranks. Indeed, some female Naxalites credit the organization with offering women opportunities to cross traditional gender boundaries. Sandinista women of Nicaragua were reportedly drawn to the group by its strong women’s organization and revolutionary feminist policies, while the leader of the LTTE once professed, “The ideology of women[‘s] liberation is a child born out of the womb of our liberation struggle.” The People’s Liberation Army (PLA) of Nepal considers women a basic foundation of the revolutionary force and requires every guerrilla squad to include at least two female combatants. It forms all-female militias and positions women high in party leadership. These female-intensive insurgencies explicitly prohibit or use very little rape during civil war, leading some to theorize that self-proclaimed emancipatory groups may be less likely to rape.
An absence of rape may reflect group political ideology. Wood notes that many Marxist–Leninist groups “engage in significant levels of other forms of violence against civilians but rarely engage in sexual violence.” Some scholars contend that non-state armed groups in general may encourage female combatants to join the fight for future gender equity. Miranda Alison argues that in ethno-nationalist military organizations, the anti-state nationalisms of rebel groups create ideological and logistical space for women to participate that do not exist within institutionalized state groups or the sociopolitical status quo. In this sense, the liberatory opposition to the state doubles as a liberatory practice of shedding some gender norms and, by extension, the rules of war. Anti-state rebel groups are more likely to be receptive to the nontraditional roles of women as combatants and thus may be more amenable to the tenets of gender equality. In rebel groups more broadly, women’s emancipation may be literally and symbolically related to emancipation from the state’s ideology and may manifest in the recruitment of female cadre.
Socialization
Socialization arguments suggest that the inclusion of women in military units may disrupt misogynistic training practices that include rape. War is a distinctively social enterprise: people participate in war in groups. Small-group bonding is essential to effective combat units, and male bonding through practices enforcing masculinity and militarism are historically seen as a vital part of this socialization. J. Ann Tickner argues, “Militaries must work hard to turn men into soldiers, using misogynist training … to teach men to fight. Importantly, such training depends on the denigration of anything that could be considered feminine.” The phallic shape of weapons including knives, arrows, spears, guns, and missiles has withstood decades of new technologies and is central to military socialization. In basic training for the US Army, soldiers famously chant, “This is my rifle [holding up rifle], this is my gun [motioning to penis]; one’s for killing, the other’s for fun.”
Commanders may use rape instrumentally as an attachment agent to vilify femininity, build camaraderie, and establish loyalty. Bülent Diken and Carsten Bagge Laustsen contend that Serb factions in the Yugoslav wars employed rape as a “rite of initiation” to ensure that fighters would not leave their units for fear of having to reconcile their actions. They identify this as a “brotherhood of guilt.” Cohen similarly argues that rape, especially gang rape, is used as a socialization mechanism to bond forcibly recruited fighters with initially low cohesion. She notes that gang rape is the most common type of rape related to conflict and that the communal aspect of group sexual violence may boost morale. However, her evidence from Sierra Leone indicates that women were frequently reported as participants of gang rape. Contrary to conventional opinion that women will interrupt rape as a bonding tool, she argues that female fighters have little effect on group violence because they are integrated into these instrumental practices.
Assessing Women’s Impact
Women’s relative passivity, the sexual substitution of out-group women for in-group cadre, the regularity of women joining sympathetic groups, and female interruption of misogynistic socialization should result in female-intensive groups committing less rape than their predominantly male counterparts. However, while the mitigating impact of female combatancy on conflict-related rape is theoretically established, it has been empirically untested. There are important studies of the influence of women’s active presence in war. These include tremendous case study research focusing on women’s activity in specific armed groups or conflicts. Still, the extent to which women have actively participated in civil wars spatially and cross-temporally has rarely been comprehensively evaluated.
To address this lacuna, I created a systematic, cross-national measure of women’s participation as combatants in armed conflict to statistically explore the relationship between female combatants and conflict-related rape. I collected an original dataset of women’s participation as combatants in state and rebel armed groups in civil wars from 1980 to 2009, consisting of 86 major internal conflicts over 983 conflict years. This new data begins to fill the gaps in existing research on conflict-related rape.
Data Collection and Coding
I produced four female combatancy (FC) variables coded across a three-level scale: none/low, moderate, and high. Ideally, the three-level coding scale would capture greater variation of women’s participation within armed forces, but part of the difficulty in assessing women’s influence in war is that comprehensive data is extremely limited. Consequently, the three-level scale, while crude, is the most appropriate coding choice. Table 1 presents these variables:
Table 1. Female combatancy variables.
| Variable | Description |
| FC | Female combatancy in rebel groups (a) |
| Female combatancy in state forces (b) | |
| FC2 | Female combatancy in rebel groups including girl soldiers and/or mixed roles (c) |
| Female combatancy in state forces including girl soldiers and/or mixed roles (d) |
As further discussed below, the primary FC variables offer the most conservative estimation of women’s combat participation. The FC2 variables are more generous, accounting for potential bias created by excluding unclear or theoretically complicated cases.
Due to the difficulty in obtaining data, female combatancy variables are coded by armed group type, not armed group. This coding procedure is established in the literature for scarce data. Information on women’s involvement in major armed organizations is severely lacking. Data on women in smaller or less prominent groups and militias, which are not well documented in general, is nearly nonexistent. The majority of data available on women’s participation reports female involvement during the conflict as a whole, the proportion of women during demobilization, or the general influx of women at a non-specific point (for example, near the beginning of the war). For this reason, I code the female combatancy variables by conflict, not by conflict year (for example, in Guatemala, female combatancy in rebel groups is coded as “moderate” for every conflict year). Available data does not yet enable a cross-national, conflict-year based analysis.
Building from scratch, this dataset draws from a variety of sources to code the female combatancy variables, including extant qualitative accounts of female fighters, demobilization reports, news articles, and reports from national governments, the United Nations, the US Department of State, and human rights organizations including Amnesty International and Human Rights Watch. Example sources used in coding include press releases from the Israel Defense Forces, Alison’s fieldwork and interview-based research on the Irish Republican Army (IRA), the US Office on Colombia’s report on sexual violence against women in the FARC-EP, Lynn Lawry and colleagues’ epidemiological survey of rape victims in the DRC, demobilization reports from Liberia, newspaper articles from Bosnia-Herzegovina, NATO’s national report on women in the Turkish armed forces, and multi-conflict accounts of women’s violence from Margaret Gonzalez-Perez and Sjoberg and Gentry.
Relying on restricted data and a breadth of case studies presents two limitations: discrepancies in the available quantitative indicators of female combatancy by conflict and the conflation of women and child soldiers. To address the first limitation, I used a mixed qualitative and quantitative coding scheme. As noted in Table 2, highly female-intensive groups are coded using a quantitative cut point of 20 percent of total cadre or a series of qualitative indicators. I established the quantitative indicator because the organizations most widely touted as female-intensive, such as the RUF, LTTE, and PKK, are composed of approximately 20 to 30 percent female combatants. In every moderately or highly female-intensive case, some quantitative estimates are available. Qualitative indicators are used secondarily to establish women’s roles in armed groups and their contextual importance.
Table 2. Coding rubric for FC variables.
| Level | Coding Rubric | Armed Groups (examples) |
| 0 Low/none | No verifiable reports of moderate or significant female combatancy:
• No reported data on women’s participation or • Quantitatively identified as less than 10 percent of armed cadre and/or • Qualitatively described as “limited” “not significant,” “low,” “non-active roles,” “supportive roles,” “non-violent resistance,” “relative absence” and/or • Indistinguishable from child soldiers or explicitly labeled as child or girl soldiers and • When possible, verifiable by more than one data source |
FARF (Chad) Nagorno-Karabakh (Azerbaijan) |
| 1 Moderate | Verifiable reports of moderate female combatancy:
• Quantitatively identified as 10–19 percent of cadre and/or • Qualitatively described as “moderate” • Identified as active armed combatants separate from other conflict roles including porters, cooks, sex workers, or support and • Distinguishable from child soldiers, not explicitly labeled as child or girl soldiers and • When possible, verifiable by more than one data source |
URNG (Guatemala) Bosnia–Herzegovina (state forces) |
| 2 High | Verifiable reports of significant female combatancy:
• Quantitatively identified as 20 percent of cadre or higher and/or • Qualitatively described as “significant number,” “significant contingent,” “in large numbers,” “large participation,” “prominent,” or “found widely” and • Identified as active armed combatants separate from other conflict roles including porters, cooks, sex workers, or support and • Distinguishable from child soldiers, not explicitly labeled as child or girl soldiers and • When possible, verifiable by more than one data source |
EPLF/TPLF (Ethiopia) PJAK (Iran) |
Table 3. Regression results, conflict-related rape in civil war, 1980–2009.
| Conflict-level Rape | Rape by Rebels | Rape by State Forces | |
| Ethnic war | −0.136 (0.128) | 0.182 (0.173) | −0.120 (0.118) |
| Extrajudicial killings | 0.269 (0.110) | ||
| Magnitude of state failure | 0.036 (0.106) | 0.250 (0.077) | 0.051 (0.080) |
| Conflict aim | −0.139 (0.109) | −0.201 (0.179) | −0.121 (0.116) |
| Female combatancy (rebel) | −0.132 (0.145) | ||
| Genocide (rebel) | −0.091 (0.075) | ||
| Abduction (rebel) | 0.635 (0.156) | ||
| Forced recruitment (rebel) | 0.033 (0.166) | ||
| Contraband | 0.601 (0.295) | ||
| Female combatancy (state) | −0.122 (0.215) | ||
| Genocide (state) | 0.043 (0.078) | ||
| Press-ganging (state) | 0.513 (0.203) | ||
| Conscription (state) | −0.095 (0.169) | ||
| Troop quality | −0.107 (0.114) | ||
| Female labor force participation | −0.001 (0.005 | 0.003 (0.005) | −0.006 (0.005) |
| Polity | −0.014 (0.019) | −0.007 (0.014) | −0.010 (0.020) |
| Fertility rate | 0.114 (0.074) | 0.070 (0.114) | 0.007 (0.090) |
| Duration | −0.0003 (0.007) | −0.007 (0.009) | −0.004 (0.007) |
| Year | 0.091 (0.0005) | 0.084 (0.001) | 0.101 (0.001) |
| Population (log) | 0.187 (0.072) | 0.090 (0.098) | 0.214 (0.069) |
- 30002 p< 0.05;
- 30003 p< 0.01.
Combatants are identified as separate from other conflict roles including porters, cooks, sex workers, and organizational support. Individual sources use varying thresholds for designating women as combatants; I identify them using both the descriptor of “combatant” and available qualitative narratives of militarized roles. Combatant fighters may move in and out of a myriad of positions including support, transport, and logistics, and they may even return to civilian life during the war. Therefore, the FC variables are conservatively coded to reflect cases where women’s roles as combatants are verifiably distinct. The FC2 variables account for cases in which women’s roles are mixed or indistinguishable from other supportive duties.
Documentation of female fighters often overlaps with that of child soldiers, likely because a large proportion of fighters in civil wars are young and many girls are kidnapped into conflict. Adult fighters may enter the conflict as girls. To avoid the theoretical entanglements in differentiating adult women from girls in war, I conservatively coded as non-female intensive/low the FC variables in which women’s involvement is not differentiated from that of girl soldiers. I coded as moderately or highly female-intensive the FC2 variables in cases where female involvement is mixed, indistinguishable, or overwhelmingly that of child fighters. This iteration of my variables reflects a liberal sample that considers these groups moderately or highly female-intensive to account for the possibility of unwarranted case exclusion and is a robustness check. I follow Mia Bloom in establishing the age of fifteen as the designation of adulthood during war.
Alternative Explanations
Other recent work explores a host of explanations for variation in conflict-related rape. Cohen statistically tests many of them using the best proxies available. I include these alternate hypotheses in my models to ask, when accounting for previously evaluated explanations for variation, if the participation of female fighters has a significant impact on group rape. To avoid reproducing work optimized by others and to maintain focus on my key independent variable, I briefly discuss alternative explanations. Scholarship by Wood, Leiby, and Cohen can be consulted for more in-depth analyses of them.
Some hypothesize that rape is more common in certain types of conflict (namely ethnic conflicts, genocides, and secessionist movements) or correlates with other forms of violence. Rape may be part of a larger campaign of ethnic cleansing or annihilation. Rape may be a tactic of secessionist groups intended to signify the end of inter-group relationships. In this analysis, Ethnic war (Fearon and Laitin) measures whether wars are ethnically motivated. Death magnitude (PITF) measures the relative levels of killings by both states and insurgents during genocides and has been recoded to account for non-genocide conflicts and years. In conflict-level analysis, extrajudicial killings (CIRI) reports the relative levels of lethal violence by year. Conflict aim (Fearon and Laitin) delineates armed groups aimed at secession versus autonomy, allowing an exploration as to which are more likely to employ rape. Another explanation is that the state weakness simply affords combatants inclined to rape the opportunity to do so. Magnitude of state failure (PITF) measures the degree of state, military, and judicial regulation and infrastructural oversight. Troop quality (Pickering) reflects cohesion within state militaries.
I also test Cohen’s hypothesis that armed groups relying on forced recruitment are more likely to use rape as a socialization and bonding agent. Four variables composing forced recruitment are used to measure recruitment practices: conscription and press-ganging for state forces; abduction and other forms of forced recruitment for insurgent groups. I evaluate the hypothesis that insurgent groups with substantial contraband funding may be less reliant on civilian populations for support and more likely to use violence against them. Drugs (Fearon and Laitin) measures the significance of contraband funding for insurgents. I use female labor force participation as defined by the World Bank to test the hypothesis that gender inequity in peacetime increases the likelihood of rape during war. Measuring latent phenomena like gender inequality is difficult, but labor force participation is an established proxy among many social science and economic researchers, including the US Agency for International Development (USAID).
Measuring Rape
I use the data on rape in civil wars that Cohen collected from US Department of State Human Rights Country Reports. Rape is coded across three variables on a four-point scale based on levels of severity by conflict year and include (1) the highest level of rape insurgent groups perpetrated in each year, (2) the highest level of rape state actors perpetrated in each year, and (3) the highest level of rape in the conflict year overall. Cohen denotes levels of rape as none, occasional, moderate, and widespread. Like the female combatancy indicators, this variable is coded by armed group type (rebel or state forces).
Results
To evaluate the impact of female combatancy on wartime rape, I estimate three ordered probit regressions that include the female combatancy variables, alternative hypotheses, and controls. I cluster standard errors to account for lack of independence between conflicts. I further estimate rare events logistic regressions to test the relationship between women’s participation and the most widespread level of rape. I repeat all models using the FC2 variables. As illustrated in Figures 1 and 2, the results suggest that women’s participation as combatants has no significant impact on the propensity of armed organizations to rape. For insurgent groups and state forces, the coefficients for all iterations of the female combatancy variables are negative but insignificant at p < 0.01 in the ordered probit regressions (Table 3). This indicates that excluding cases in which women’s participation is largely that of girl soldiers or in non-combatant roles has no significant impact. Female combatancy is still insignificant at the lowest p value in the rare events logistic regression and is positively correlated. These results suggest that women’s involvement as combatants does not significantly impact the likelihood that armed groups will commit higher levels of rape during civil conflict, demonstrating that female-intensive armed groups are not less likely than other organizations to commit rape in civil wars cross-nationally.
While women composed a moderate or high proportion of combatants in at least 21 percent of rebel groups during this time period, only 4.7 percent of state forces can be classified as moderately or highly female-intensive. Of these cases, only one—Bosnia–Herzegovina—is a state force in the traditional sense. The other female-intensive public militaries in my data are the DRC, Liberia, and Rwanda, for which the distinction of “state actors” is theoretically complicated because these forces may be far more widespread, less regulated, and less representative of a conventional government. The sample is too small and uncertain to draw responsible conclusions about female combatancy and state forces; with a larger sample, correlations between women’s active participation in armed state organizations and rape can be more confidently established. Nonetheless, the consistent insignificance of the female combatancy variables across models suggests that women’s participation does not lessen rape, contradicting the notion that their presence in armed organizations explains why rebel groups commit less rape than state forces.
The full findings of the regression largely support those in the literature. When accounting for women’s participation, armed groups are more likely to commit rape when states are weak and when insurgents rely on civilian populations. Rape likely correlates with other forms of lethal violence during the war, although ethnic war and genocides are not significant estimators of rape by either rebel or state forces. Nor do secession or conflict aim predict levels of rape. Finally, fertility rates and female labor force participation are not significant. These results remain unchanged when employing the FC2 variables.
Figure 1. Likelihood of rape committed by rebel forces
When accounting for female combatancy, abduction by rebel groups and press-ganging by state forces remain significant indicators of rape. To thoroughly assess Cohen’s socialization hypothesis, I ask, “Does the level of women’s participation impact whether groups use rape as a bonding mechanism for forcibly recruited fighters?” I produced a fourth model that includes an interaction term composed of rebel female combatancy and use of abduction by armed groups. The predicted probabilities (Figures 3 and 4) demonstrate that while armed organizations recruiting through abduction may be more likely to perpetrate rape, the percentage of women in the group has no effect on this probability, regardless of whether the rebel group uses abduction. A fifth model (not shown) exploring the characteristics of female-intensive armed organizations, finds that female-intensive rebel groups are no more likely to use abduction than other groups. The extent to which groups abduct fighters or use rape to socialize abducted members is unrelated to the gender composition of the unit.
Figure 2. Likelihood of rape committed by state forces
Women’s participation does have some effect when we consider the interaction between female combatancy and state press-ganging. As illustrated in Figure 5, the percentage of women in state forces has a slight but not substantial impact on the probability of rape for militaries using press-ganging. However, in cases of units that do not use press-ganging, women’s participation has a significant impact. The data suggests that female-intensive state forces refraining from press-ganging are more likely to use occasional, moderate, and widespread rape when they have higher proportions of women. While this finding is intriguing, the sample of female-intensive state forces is again so small and theoretically varied that these results should be treated with caution. Bosnia-Herzegovina is the only case in which a female-intensive group did not use press-ganging in at least some conflict years, indicating that one case may drive the results. Regardless, across models there is no evidence that women’s participation correlates with less rape.
Figure 3. Likelihood of rape committed by rebels using abduction
In the rare events logistic regressions, I find that female combatancy is unlikely to impact the severity of rape during conflict. Female-intensive rebel and state forces do not employ lower levels of rape than their counterparts. Rebel groups that forcibly recruit members may be more likely to perpetrate the most widespread practices of rape, as do resource-heavy factions that operate in areas of low oversight. Female labor force participation is positively and significantly correlated in this regression, indicating that gender inequity in peacetime may increase the likelihood of high levels of rape by rebels during conflict. It is notable that genocide by rebels is a highly significant indicator in their committing the highest levels of rape. But because there are only two cases of rebel groups committing genocide in this sample—the second conflict in Angola and a one-year conflict in Burundi—it is possible that this finding is statistically representative of an uneven sample and not substantive phenomenon. These results remain unchanged when employing the FC2 variables.
Figure 4. Likelihood of rape committed by rebels without abduction
Re-Thinking Rape
Arguments advocating female combatancy as a method for discouraging rape (hereafter “theory of female combatancy”) suggest that women and men respond differently to military situations and that individual characteristics guide behavior during war. Cohen addresses these assumptions directly, arguing—as Diken and Laustsen do—that rape may be a tactical bonding tool. She notes the deficient evidence of women’s passivity and contends that armed factions persuade forcibly recruited fighters, including women, to participate in gang rape to strengthen cohesion. While instrumental socialization practices explain variation in the use of rape across organizations, this hypothesis is under-theorized. Cohen emphasizes the social risks of refusing to participate in costly behavior, but she neglects the wide social psychological and political science literature on conformity and social identity.
Instrumental socialization theories fail to explain women’s rape and other civilian violence beyond that intended as a bonding mechanism. How do women respond to military institutions when instrumental socialization practices are not used? How can we explain women’s independent violence in civil war? Why doesn’t female combatancy reduce rape? Organizational factors—primarily culture—drive violence in armed groups, operating through a variety of mechanisms including obedience to commands, group social identity, and norm internalization. Because these mechanisms encourage conformity irrespective of individual characteristics or instrumental bonding practices, the presence of female fighters does not reduce the incidence of rape. I follow Elizabeth Kier in defining organizational culture as “the set of basic assumptions, values, norms, beliefs, and formal knowledge that shape collective understandings.”
The argument is threefold. First, the propensity for individuals to obey commands and conform to group norms —mechanisms offered by social psychologists to explain individual behavior—are particularly strong in armed groups because of their hierarchical structure and high levels of external threats. Fighters conform to group preferences, including those that conflict with personal presences, even in the absence of a direct order. Second, as international security scholars emphasize, organizational factors take precedence over individual dynamics when explaining the culture of armed groups and violence against civilians. Cross-disciplinary empirical evidence supports this notion, pointing out that previously excluded groups, such as homosexuals and African-Americans, have been integrated into armed organizations without disrupting unit cohesion or altering group violence. Third, organizational cultures of militaries are often replete with hyper-masculine and misogynistic norms; women are socialized into these cultures that promote sexual violence. Incorporating women into armed institutions does not disrupt the misogynistic, masculine culture of those groups because women are subject to the same organizational dogmatism as the men. Women’s violence—including rape—against civilian targets supports this claim.
Identity and Conformity
Militaries mandate that rank and file cadre commit violence to further organizational goals. Why do ordinary individuals perpetrate violence during war? Social psychologists offer two discrete explanations: obedience and conformity. Several renowned experiments, such as those of Stanley Milgram, Philip Zimbardo, and Jerry M. Burger, demonstrate that individuals obey authority and inflict violence with little resistance even if it conflicts with personal values. There is little evidence that women are less violent than men in these circumstances. Women in Milgram’s tests, Zimbardo’s shock experiments, and Burger’s study obey commands at levels virtually identical to men. Milgram notes that individuals possess an “extreme willingness to go to almost any lengths on the command of an authority.” Soldiers from highly regulated institutions such as the US military and the Wehrmacht have explained violence as “just following orders.”
Rebel groups, however, are often loosely structured. Insurgents move in and out of combatant roles, sometimes returning to civilian life for long periods of time. Fighters may be embedded in their communities during combat. Leadership may be fractured and commanders vary in strength and oversight. Obedience to authority may explain some individual violence, but is limited in these cases. Rather, pressure to conform to group norms may be at play. Evidence suggests that individuals will still conform to group behavior in the absence of a direct command, as well as in armed groups with lower degrees of regulation. Individuals are likely to adopt the majority opinion or norms of a group, even when they believe that the norm is wrong. Their preferences will change when a group norm shifts. If a group norm has not emerged, they will still conform to their colleagues’ opinions. This phenomenon can drive group behavior in the absence of a direct order from above. Groups often punish deviants because achieving group goals depends on group identity and conformity.
Group identity is easy to create. Henri Tajfel’s experiments demonstrate that individuals grouped together based only on their estimation of the number of dots on a slide will favor their group members and discriminate against an out-group. There is no gender difference: women discriminate as robustly as men. Tajfel and John Turner use social identity theory to explain why individuals in these “minimal groups” discriminate against others. Researchers contend that feeling positively about one’s group or negatively about another group strengthens in-group identification and encourages discrimination. Subsequently, groups will compete with one another, maximizing their differences rather than seeking mutual gain.
Social identity theory is a useful framework for understanding group behavior, but distinctions between armed factions are much sharper than those based on dot calculations. During conflict, the risks are high. Emotional and discriminatory characteristics divide armed organizations based upon political ideology, religion, ethnicity, desire for power, economics, experiences of repression, and fear. It is not surprising that in-groups and enemies create such defined boundaries during war. Armed organizations face powerful external threats and have a vested interest in the group identity of their cadre. It is not just that women and men feel more positively about their group when faced with danger, they also conform more strongly to organizational norms. As Kier argues, “greater hostility in the organization’s external environment [increases] the potential for organizational dogmatism.” This explains why women and men participate in violence even in cases with fractured leadership or in the absence of a direct command. Women are equally subject to organizational doctrine.
Explaining Individual Violence in Armed Organizations
Subsequent to such dynamics, international security scholars prioritize organizational explanations for civilian violence, arguing that organizational factors drive behavior and group decision-making. The theory of female combatancy, which emphasizes the individual, is inconsistent with their studies’ theoretical and empirical findings. For example, Jeremy Weinstein argues that when and how rebel groups abuse non-combatants is contingent on organizational structure and the types of economic and social endowments they can access. Amelia Hoover Green contends that the ability of commanders to control their soldiers dictates violence. Reed Wood views civilian violence as a blunt instrument that rebels use when they lack capacity to compel loyalty through benefits. Sexual violence scholars conclude that state militaries are most likely to rape when rebels constitute a significant threat to state power but have not yet reached dominance; groups use sexual violence when it advances political goals; and commanders can curb rape.
Hoover Green observes, “Recruits enter armed groups with widely varying ‘preferences’ over violence, but socialization processes break down these initial preferences and build, in their places, norms and preferences that better serve group goals.” She emphasizes desensitization and conformity pressures as key mechanisms in this process. She further demonstrates how armed groups take otherwise average fighters and increase their preferences for violence, particularly rape. Conflict actors internalize this view as well. One Liberian civilian notes the influence of simply having a weapon on the behavior of insurgents: “We the women are suffering the most during war. We are completely powerless. We are powerless because the men have weapons. It’s only due to their possession of weapons and not because they are men. The weapons give them power. This is the reason why they can rape and harass the women.”
Armed groups do not have to work hard to build shared norms and objectives among fighters. Fear, necessity, and ideology encourage individuals to rely on and identify with their organizations. Consequently, they commit violence. Instrumental socialization theories are incomplete as evidenced by the fact that combatants rape and commit violence in female-intensive groups where forced recruitment is infrequent. Rather, instrumental bonding practices that include rape are likely the result of organizational values, norms, and objectives. Instrumental socialization is not the cause of women’s integration into wartime violence, but may be the result of organizational dogmatism that encourages conformity and sexual violence more broadly.
This explains why women participate in or fail to curtail rape in groups where they are not forcibly recruited or part of instrumental socialization practices. For example, in Liberia women compose between 20 and 40 percent of armed groups and are accused of civilian targeting and rape. However, only 12 percent of female and 21 percent of male ex-fighters reported being abducted. The majority of fighters reported fear or desire to protect their families as motivation for joining an armed faction. Female ex-combatants report a strong social identity resulting from the belief that their group was the best vehicle for survival. One fighter notes, “He [the Commander] told the people to give me water and then he gave me clothes to wear. They gave me food, everything. So at that day I felt at home. I feel at home that very day.”
Another remembers, “My parents got killed. So I decided at that time, because the only way to help yourself, or to get food for your family is to be a soldier.” Liberian ex-combatants also detail a group identity built from animosity toward other factions. Liberian rebel and political leaders stoked ethnic and religious flames to solidify hatred for the enemy and to encourage soldiers to fight. As one female ex-fighter remembers, “I went fighting because of my religion. You see if you are a Muslim or Mandingo in this country, they say you don’t belong in this country. So I had to fight.” Threat and the intensity of experience lead individuals to identify strongly with their groups and conform to organizational norms that may contradict personal preferences for non-violence. Combatant groups have integrated women without disruption because individuals, regardless of gender, are socialized toward group preferences for violence.
Kier observes that integrating women and other previously unrepresented groups into the US military has not interrupted the organization or generated any observable changes. Some Liberian insurgents note the relative ease of integrating effective female commanders. As one recalls, “Some boys are laughing in the beginning when women want to join, but if the girl proves herself, the men will look up to her. The men can as much obey a male general as a woman general. Men have respect for good generals and they are also afraid of them so they will obey.”
In fact, incorporating women has not altered practices of rape or civilian targeting in most organizations. Even the most ideologically progressive and female-intensive factions commit civilian violence, and many of them rape. Heavily female insurgencies including those in Liberia, the DRC, India, Nepal, Colombia, Sierra Leone, and Burundi rape non-combatants. And as Tickner notes, introducing women to the US military has not altered its masculine culture or mitigated problems of sexual violence. After the release of photographs depicting female soldiers’ sexual abuse of men detained at Abu Ghraib, Barbara Ehrenreich confessed, “Secretly, I hoped that the presence of women would over time change the military, making it … more capable of genuine peacekeeping…. A certain kind of feminism, or perhaps I should say a certain kind of feminist naiveté, died in Abu Ghraib.”
Women in combatant groups commit brutal violence. Many local populations have implicated women in violence even more extreme than that of their male counterparts. Hutu female combatants played an unprecedented role in the Rwandan genocide, with one witness recounting watching a woman carrying a baby murder another woman holding her child. Helene Cooper recalls a female Liberian solider admiring another woman’s baby, saying, “Oh, what a fine baby! I’ve killed two like him today.” Women committed 80 percent of suicide bombings in the Russian–Chechen separation wars, often killing women and children. Female LTTE fighters were thought to be more violent and frightening than male members, taking part in intimate violence, suicide bombings, and assassination plots. Speaking to a reporter during the Yugoslav wars, Bosnian soldier Mirsada Hromo noted the pleasure derived from her violence: “It’s a nice feeling to kill a man, especially when you know he is going to kill you. You get this special feeling when you see him walking toward you, wanting to kill you and you just shoot him.” To this, another female soldier responded, “Maybe we should charm them so they’ll walk a little closer.”
Female fighters also rape. Women participated in approximately one in four incidents of reported gang rape in Sierra Leone, often raping women with sticks and bottles. Victim testimony suggests that female-perpetrated gang rape was also practiced against Tutsi men in Rwanda. According to IBUKA, a Rwandan survivors’ advocacy group, there are numerous reports of female combatants forcing young Tutsi boys into sex. Sjoberg’s forthcoming book on women’s “genocidal rape” explores motivations and details a diversity of cases spanning conflicts in Darfur, Serbia, Rwanda, Nazi Germany, and the Armenian Genocide. Liberian female fighters are accused of raping women and men, including penetrating victims with guns and mutilating men’s genitals. Some of these fighters committed acts of sexual violence to avenge rape.
Integrating women has not reduced rape. In fact, female-intensive groups often subject their female fighters to sexual violence. This is true of factions with egalitarian political agendas and of those that rape civilians. Maoists of the PLA sexually assaulted young women recruited into the group and punished female cadre and noncombatants raped by the opposition; commanders also took sexual slaves. In Sierra Leone, some RUF members carved the group’s initials into the breasts of girls who attempted escape. Some female FARC-EP fighters report being expected to remain sexually available to their male counterparts, and they recall forced abortions. In Liberia, an estimated 75 percent of young fighters reported sexual abuse or exploitation before and during conflict. Coulter and colleagues note that while high-ranking female commanders were sometimes spared abuse, they experienced sexual violence and discrimination before achieving their leadership positions.
Theoretical and empirical evidence from social psychology and political science suggest that groups with a propensity for rape will do so regardless of gender composition. An easy conclusion is that gender simply washes out of the violence equation. However, in order to explore the effects of women in combat, it is important to consider armed groups as uniquely gendered institutions and question how their organizational cultures encourage rape. The theory of female combatancy holds that women’s participation will profoundly alter military violence. This hypothesis undervalues the organizational importance of hyper-masculinity in military culture, neglects the situational influence of misogynistic narratives on all fighters, and fundamentally misunderstands the purposes of rape during armed conflict. Integrating women into militant organizations does not strip away violent values or halt their reproduction. Rape will continue to be a practice of many military institutions until commanders explicitly proscribe it and abandon a culture of misogynistic masculinity.
Bringing Gender Back In
Feminist international security scholars have written extensively on the importance of masculinity to the maintenance of armed institutions. They conclude that group cohesion and effectiveness in war depends on a valorized misogynistic masculinity. I do not argue that militaries must rely on misogynistic culture to encourage individuals to fight, but four decades of research emphasizes the extent to which these norms, practices, and experiences shape military culture. Jennifer Mathers writes that the “notion of a hierarchy or privileging of (what is construed as) the masculine over (what is construed as) the feminine is very important because it helps explain why militaries need to preserve their status as distinctively masculine.” This is because, as Lauren Wilcox notes, armed groups are “institutions that mold men into the values of warrior masculinity” to garner obedience. Hoover Green concludes that within these units, commanders valorize violence among recruits and demand obedience through behavioral conformity and traditional masculinity. She argues, “The result, in many cases, is a set of cultural norms that valorizes violence in general and sexual violence in particular —and that strongly discourages criticism of group norms, goals, or actions.”
As Maurice Berger and colleagues attest, “Far from being just about men, the idea of masculinity engages, infects, and shapes everyone.” In order to integrate individuals into the greater armed group identity, militaries capitalize on notions of manhood that both reinforce warfare as categorically male and degrade femininity. This involves instrumental mechanisms including misogynistic training and bonding practices, but it is also rooted in the larger discursive context of armed organizations as sites of masculinization through violence. It is for this reason that rape is a productive tool during war. Rape, like war, seeks power and domination. Rape during conflict serves many perverse functions: it may deactivate communities, force civilians to abandon territory, or be used for punishment, retribution, torture, intimidation, revenge, socialization, humiliation, or to improve perpetrator group morale. Acquiring power through domination is culturally and historically a masculine enterprise, illustrated best by the rape and pillage model of warfare that has survived centuries. Adam Jones notes that militarization bolsters masculinity through the expression of violence—war demands violence and victory. This reification of masculinity through domination feminizes the victim regardless of gender.
A theory of female combatancy misunderstands rape as a function of heterosexual maleness. Many predominately male groups, including those in Afghanistan, Chad, India, China, Azerbaijan, Croatia, Pakistan, and Morocco commit little or no rape. The largely male Sri Lankan Armed Forces disproportionately rape men. Some female-intensive insurgencies, including those in Guatemala and El Salvador, refrain from rape. Heavily female rebel groups in the DRC and Liberia rape female combatants and civilians. The FARC-EP subjects female fighters to violence while mostly sparing noncombatants. Rape is not a function of group composition or individual preference, it is the result of organizational factors including culture, structure, and commander oversight. Historically, military institutions use misogyny, gender-based violence, and dominance as mechanisms of cohesion and survival. Nicole Detraz summarizes, “Adding women to an institution that depends on and furthers militarized masculinities without challenging those masculinities will likely to do nothing to fundamentally change the institutions themselves.” As Cynthia Enloe contends, lauding women’s combat participation as feminist or emancipatory fails to consider militarism’s violent, masculine harms to which women in armed groups are subject.
Recommendations
The UN and many scholars recommend gender mainstreaming as a solution for conflict-related rape and gender based violence. This proposal assumes that women’s presence in armed factions reduces their proclivity for rape. My findings suggest that integrating female combatants into armed groups does not lessen the incidence of rape in civil conflict. Rebel groups do not commit less rape than state forces because they have more female fighters. Adding more women to public military institutions is unlikely to substantively impact the use of rape during war. Arguments for women’s pacifying effect create a misleading framework for understanding women’s wartime experiences. Future research must interrogate those experiences contextually. In some cases, women’s involvement may be directly related to a lack of rape. In other cases, women are the rapists. As Paige Whaley Eager observes, “Women’s relative agency in various organizations and movements which utilize political violence and terrorism varies tremendously.”
The finding that the presence of female combatants does not curtail rape has considerable theoretical, empirical, and policy implications for research on conflict-related rape and programs intended to prevent it. Operating under the assumptions that women’s presence discourages rape and that women are not sexual offenders, researchers, international governmental organizations, and aid workers have produced academic work, population-based sexual violence and health surveys, military strategies, DDR programming, and survivor outreach that ignore the roles women play as armed combatants.
For example, UNSC 1325 is the first UN resolution to specifically mention women and the first legal document requiring member states to protect women and girls from sexual and gender-based violence in conflict. It is an important document, but one that adopts false assumptions about women’s wartime experiences. It recommends gender mainstreaming as a primary solution for rape. Propagating the female combatancy hypothesis disadvantages women’s capacity for reintegration and security during and after conflict, which in turn renders peace-building and post-war reconciliation woefully incomplete. How can we make durable peace recommendations, adjudicate war crimes, intervene in incidents of rape, or assist in post-conflict development programming when we ignore the empirical importance of women’s violence?
A large literature over the past two decades recognizes the shift in tactics of warfare that blurs the distinction between combatants and non-combatants and understands conflicts fought between often non-uniformed, non-traditional military actors including civilians and children. However, while the international community has focused on the plight of child soldiers, the unique experiences of women during war have largely been neglected. The one-dimensional representation of women as passive, uncommon war actors is rendering them invisible. When women’s participation is recognized, the results are promising. For example, in Angola and Sierra Leone, female fighters were classified as dependents in DDR initiatives and thus were prevented from receiving state benefits provided to combatants. Women constituted just 6.5 percent of DDR participants in Sierra Leone, despite estimates that female fighters composed 24 percent of the RUF alone. In contrast, the executive director of Liberia’s DDR programs noted that they were explicitly gender-sensitive and enabled women who completed demobilization to receive reintegration benefits including school enrollment and vocational training. Women composed nearly 20 percent of total cadre demobilized in that country.
The policy implications of these findings are not unique to the military apparatus. Clinical, psychosocial, and legal interventions for victims of conflict-related rape cannot be properly instituted without understanding women’s roles as perpetrators. As noted, the majority of population-based epidemiological studies on rape do not ask for the offenders’ gender. Women actively participate as rapists in a number of conflicts. As Lori Girshick notes, “women-on-women sexual violence is an invisible form of sexual violence because of our denial that women are sexual perpetrators.” The doctrinal infallibility of the notion that men are perpetrators and women are victims is so engrained that post-violence support for male victims of sexual violence is often non-existent. Medical workers may not be trained to look for signs of sexual violence in men or may disbelieve victims’ stories. Dismantling these gendered stereotypes improves survivor access to healthcare and judicial redress.
Theoretically, the findings suggest that we are asking the wrong questions about gender and war. Currently, we are asking if gender impacts war processes as an external shock, as a theoretical intruder. What happens when we take war and add gender? As this article demonstrates, we should view gender as integral to processes of violence and conflict and theorize it as such. Instead of asking if gender modifies violence, ask how does gender modify violence? How does gender change or conform to violent processes? Under what circumstances does gender have a substantive impact? Under what conditions will women rape? Under what conditions does women’s presence mitigate sexual violence? Under what conditions does it aggravate them? How do these new theoretical avenues improve our understanding of international security?
Empirically, research regarding women’s motivations for joining militant groups, in-unit roles, use of violence, political imprisonment, and post-war social reintegration is deficient. Some of the most female-intensive rebel groups globally, including the PJAK, remain almost entirely unstudied. Security studies and sexual violence researchers must explore new avenues for research about female fighters and work to document and understand women’s violence during war. While perpetrators of rape are overwhelmingly male, women’s violence must be recognized and understood as real violence if we are to explain rape’s variations and make recommendations to prevent it. Academically, recognition of—and a focus on—female combatants will enable more comprehensive data on women’s participation in civil war to be collected. My results suggest that practically, advocating further militarization of women in an attempt to pacify the lethality or intensity of conflict, as De Groot and UNSC 1325 do, may be a poor and largely ineffective practice. My findings refute Card’s argument that programs intended to train and militarize women will undermine rape during conflict.
As women increasingly become central actors in internal conflict, security studies must recognize the critical roles that gender plays in shaping war processes. That gender matters in international politics is well established, but the conditions under which it plays a distinctive role are ripe for debate and reimagination. While women’s inclusion in armed groups does not have a mitigating impact on conflict-related rape, their active involvement as combatants requires us to delve more deeply into the complexities of gender and militarization. We should not rely on outmoded assumptions of gender as framework for new theoretical ground. Moving forward, we must not continue asking the question of whether or not gender plays an important role in international security. We should direct our attention to how gender matters, when, and why.