The Impact of Women Legislators on Humanitarian Military Interventions

Abstract

In this article, we contend that the current gender and conflict literature ignores the context of military decisions and thus underestimates the support of women for certain types of military interventions. We argue that the issues related to humanitarian crises are likely to provoke support from women. Consequently, as more women enter elected positions in state legislatures, the more likely a state will become involved in a humanitarian military intervention. To test our argument, we compile a data set of humanitarian military interventions and women legislators from 1946 to 2003. A series of estimation approaches and robustness tests support our assertion that more women legislators impact the likelihood that a state will become involved in a humanitarian military intervention. Our research has specific implications on the role of gender in conflict processes and more general implications on the connection between domestic political processes and foreign policy decision making.

Keywords

humanitarian intervention domestic politics foreign policy military intervention

As of January 2015, women constituted 22 percent of the world’s parliaments and legislatures. This number nearly doubles the percentage of women legislators from 1995 (Inter-Parliamentary Union and UN Women 2015). This changing representative dynamic has implications not only for domestic policies but also for international relations. While there is no general consensus in the conflict literature, some research has demonstrated that women, whether in the general public or in government, are less likely to support military action. If true, more women in government may decrease the use of military force.

However, previous examinations of the relationship between gender and conflict largely ignore the reasons why states become involved in conflict. States use force for many reasons, suggesting that different motivations may generate varying degrees of support across a society. In this article, we argue that some types of military interventions receive more support from women than other types of interventions. Specifically, the issues related to humanitarian crises are likely to provoke support from women. Consequently, as more women enter elected positions in legislatures, the more likely a state will become involved in a humanitarian military intervention.

Our argument advances the idea that the motivation for the use of military force can originate from varying domestic sources. As a result, the association between domestic politics and the use of force varies across different types of military interventions. Some scholars already examine how the motivations of some conflict—whether it be territorial, policy based, or regime change motivated—can affect the proclivity and nature of conflict (Huth and Allee 2002; Senese and Vasquez 2008; Moon and Souva 2016). Other research examines how specific domestic groups have varying degrees of support for conflict (Synder 1991; Jentleson 1992; Jentleson and Britton 1998; Kirshner 2007; Narizny 2007; Lobell 2008). Our research combines these approaches, examining how domestic political processes intersect with different types of military interventions.

Our analysis has implications on the study of gender and conflict. Most of the existing literature on gender and conflict treats conflict as a homogenous phenomena, where the decisions to use force or to increase military spending are not put into context. In this study, we focus on the salience of particular issues—specifically the prevention of sexual violence and the protection of children—in humanitarian crises, and find reasons why women legislators are likely to push for military interventions during humanitarian crises. By examining specific military interventions, motivated by specific issues, we find evidence that women legislators increase the likelihood of military action for humanitarian reasons. Our analysis helps reconcile the inconsistent empirical relationship found between gender and conflict. Some research finds that gender equality or women in legislators have a pacifying effect on conflict (Marshall and Ramsey 1999; Caprioli 2000; Caprioli and Boyer 2001; Goldstein 2001; Regan and Paskeviciute 2003). Other research finds either a null relationship (Bjarnegård and Melander 2013) or even a positive relationship between women leaders and conflict (Caprioli and Boyer 2001; Koch and Fulton 2011). We contend that exploring the heterogeneous motivations of conflict helps define the relationship between gender and conflict. The salience of the prevention of sexual violence and the protection of children motivate women legislators to advocate for the use of force for humanitarian reasons.

The rest of this article is structured as follows. First, we review the literature on gender and conflict. Then, we present our argument to explain why more women legislators increase the likelihood of humanitarian military interventions. To test our argument, we compile a data set of humanitarian military interventions and women legislators from 1946 to 2003. A series of estimation approaches and robustness tests support our assertion that more women in legislatures increase the likelihood that a state will become involved in a humanitarian military intervention. We conclude with a discussion of our analysis and suggest possible avenues for future research.

Gender and Military Action

There is an extensive literature that examines the relationship between gender and conflict issues.¹ With important exceptions, the general consensus is that women are less supportive of military action than men. This finding can be found in connection to the general electorate (Smith 1984; Shapiro and Mahajan 1986; Conover and Sapiro 1993; Togeby 1994; Caprioli 2000; Caprioli and Boyer 2001; Nincic and Nincic 2002; Eichenberg 2003; Eichenberg and Stoll 2012), to attitudes toward women (Hudson et al. [2008] 2009; Caprioli 2000; Caprioli and Boyer 2001), and to women in government (Koch and Fulton 2011).

Some scholars argue that the women’s attitudes toward the use of force may be a result of women being conditioned by society to be maternal caregivers and peacemakers (Ferris 1993). Others argue that women are more loyal to liberal political parties, which tend to oppose the use of military force (Koch and Fulton 2011). Alternatively, there may be a reverse causal story, as more women in legislative bodies are a symptom of a society’s values that would favor a more peaceful foreign policy (Caprioli 2000; Caprioli and Boyer 2001).

However, there is evidence that the relationship between gender and international relations is not as straightforward as the conventional wisdom claims. Holsti and Rosenau (1990) find little evidence of a gender gap in foreign policy attitudes among American elites. In addition, other scholars have focused on the stereotypes associated with women and foreign policy that might prompt women officials to have more hawkish attitudes (Lawless 2004; Swers 2007, 2013). Koch and Fulton (2011) find that the presence of women executives increases conflict behavior and defense spending.

We argue that this mixed evidence can be reconciled with the belief that women are less supportive of the use of force. In abstract conditions, women are less supportive of military force, but when women are asked about specific military action, they are sometimes as supportive of force as men (Conover and Sapiro 1993). For example, the gender gap closes when considering the use of force in the Middle East (Wilcox, Hewitt, and Allsop 1996; Tessler and Warriner 1997; Tessler, Nachtwey, and Grant 1999). This suggests that the issues surrounding military action matter in terms of generating support from women. By extension, the issues surrounding military action should affect how women in government support or influence decisions to use force. In the next section, we outline an argument that women in legislatures will be more supportive of humanitarian military interventions, thus when more women are in government, a state will be more likely to engage in humanitarian military action.

Our Argument

Although many assume women are less supportive of military interventions than their male counterparts, we question why this relationship holds over all types of military interventions. Some military interventions, specifically humanitarian interventions, are designed to save at-risk lives, which by and large are women and children. We expect that the vulnerability of women and children in humanitarian crisis situations will be a salient issue for women legislators in other countries, who will advocate for humanitarian interventions to protect these at-risk groups.

Humanitarian Crises

War may result in humanitarian crises, where the blurring of front lines and home fronts endangers the general population. Given that men are more likely to be official participants of conflict, they are more likely to be harmed directly in combat. However, conflict has severe indirect costs, as “most victims of war die in silence” (Plumper and Neumayer 2006, 723). Plumper and Neumayer estimate that war affects women’s life expectancy more negatively than men’s. Reduced access to necessary resources such as food, clean water, and health services produces stronger indirect effects than the direct combat effects of war.²

War can also displace people from their homes, forcing the general population to seek safety in refugee camps. The resource disparity between males and females caused by crises can be exacerbated within these refugee camps. As a result of this resource disparity, females were reported to be three and a half times more likely to die than males in camps (Plumper and Neumayer 2006). A general lack of resources and vulnerability to infectious disease accounts for this difference in mortality rates. In addition, refugee camps do not protect women from sexual violence. Economic inequality and lack of access to resources leave women reliant on men and more vulnerable to economic manipulation and exploitation. For example, in Sri Lanka in the late 1990s, incidents of rape and other types of sexual abuse were reported to have been committed by police and the paramilitary personnel assigned to provide security to the population displaced by the civil conflict (Fisher 2010).

Outside of these refugee camps, women face threats of sexual violence in conflict environments. Militant groups adopt sexual assault as a tactic to terrorize and demoralize the opposing population. In Yugoslavia, Bosnian women were systematically raped, sometimes even forced into mass “rape camps,” by Serbian men to undermine the mental and psychological welfare of a community (Niarchos 1995). The use of sexual assault to instill fear in the refugee communities was a tactic carefully planned by forces like the Janjaweed in Darfur, and its purpose was to destroy a community’s sense of security, immobilize a community with the fear of attack if they tried to leave, and emotionally and physically damage the women of the community to prevent them from later producing children.

Even outside of armed conflict, women face risks during humanitarian crises. For example, women had a mortality rate one and a half times higher than men during the 1995 Kobe, Japan earthquake, five times higher in the 1991 Bangladesh floods, and up to four times higher during in the Southeast Asia 2004 tsunami (Seager 2005; Mazurana, Benelli, and Walker 2013). A natural disaster can cause the general displacement of people, creating a scarcity of basic necessities. In addition, natural disasters create an environment of chaos and resource disparity, which can permutate into sexual violence. In the aftermath of the tsunami in 2004, opportunistic perpetrators of sexual violence took advantage of the turmoil, preying on girls and women who were extremely vulnerable to abuse. Sexual violence, harassment, and domestic violence continued during disaster recovery. Similarly, following Hurricane Mitch in Nicaragua and Honduras, coerced prostitution became a problem in some rural areas among adolescent girls (Fisher 2010).

Women Legislators and Humanitarian Military Interventions

Because humanitarian crises—either born out of armed conflict or natural disasters—threaten the safety and security of women and children, we argue that women in other countries, specifically women legislators, will be more supportive of military interventions during these crises. When these humanitarian crises present themselves, women legislators are more likely to act as transnational surrogate representatives for women outside their constituency.³

A legislator will be more likely to feel responsibility for nonconstituents when a representative shares experiences with the surrogate constituents that are alien to the experience of the majority of the legislature (Mansbridge 2003). Thus, a sense of surrogate responsibility is more likely to emerge when women legislators share experiences or empathy with the surrogate constituents in a way that men legislators do not or cannot. Humanitarian crises generate two shared experiences for women across most cultures that incite empathy and provoke support for military interventions: protection of children and prevention of gender-based violence. While men can have the same level of empathy, they are less likely to be politically motivated by this empathy. US Representative Nydia Velazquez articulated this view of surrogate representation, arguing that women legislators “have a responsibility that goes beyond women in America …. We have a responsibility to protect women elsewhere” (Carroll 2002, 57).

Women are not a homogenous political group with uniform interests and beliefs. Only a small part of the experiences of women (or any group) “is reflected in the experience of any particular individual” (Weldon 2002, 232). While women’s interests and experiences are diverse, they do share some common experiences of being women. As the expected primary caregiver of children across most societies, women have a shared experience in protecting children. In addition, women are cross-nationally more vulnerable to sexual violence, and thus have a shared motivation to prevent further occurrence of violence. For example, US representative Carrie Meek stated that “[a] woman who is raped in Bosnia by the troops who were there, or a woman who is raped in Haiti by coup members over there, a woman who has her genitalia mutilated in certain African societies.… I am partially responsible for that, for some amelioration of that” (Carroll 2002, 57).

These shared experiences of women are often highlighted as women legislators’ policy priorities. For example, in 2010, new British Home Security and veteran member of parliament, Theresa May, stated that “[v]iolence against women is and always has been a priority for me and my ambition is nothing less than ending it.”⁴ Australian Senator Christin Milne articulated that women’s security should be the center of any state’s national security plan, stating, “You can’t have a secure Afghanistan unless Afghan women are safe.” She urged Australia to use its position “to raise and support this issue of empowerment of women in Afghanistan.”⁵

Our argument that women have shared experiences about the protection of children and preventing sexual violence is consistent with public opinion results. Women are relatively more sensitive to humanitarian concerns and to the loss of human life (Eichenberg 2003). Cross-nationally, women in over 180 countries identified violence against women as a vitally important issue (Weldon 2002). Protection of children and prevention of gender-based violence make humanitarian crises particularly salient to women legislators. As a result, we expect that women legislators are more likely to push for military interventions during these crises. These legislators can affect the process through their votes or through alternative legislative processes. In 2011, US Representative Carolyn Maloney, along with several female cosponsors, introduced legislation titled the “International Women’s Freedom Act” that explicitly requires the President to “oppose [human rights] violations and promote the human rights of women” through diplomatic or military means (https://www.govtrack.us/congress/bills/112/hr418, accessed September 1, 2014).

Previous empirical work has examined the voting patterns of women and men and has found there is a consistent empirical relationship between the gender of legislators and voting on legislation that affects children and gender violence. Survey research suggests that women officials feel a responsibility to legislate on behalf of women issues. In 2001, 85 percent of women in the US state legislatures that responded to a survey agreed that “women legislators have a special responsibility to represent women’s concerns within the legislature.” In addition, two-thirds of the women respondents stated that they worked on legislation directly linked to women issues (Center for American Women and Politics 2001). Women are more likely to vote in favor of legislation that protects, aids, or educates children with programs that increase welfare programs for children or increase funding for education (Thomas 1991). This relationship holds for cross-national legislation as well, as more women in legislatures lead to increased development assistance to other states (Breuning 2001). Women are also more likely to vote for stricter laws related to gender violence like laws on domestic abuse, sexual assault and rape, and on foreign women’s rights policies (Childs 2006; Swers 2006).⁶ Women legislators across parties are more likely to discuss women’s interests in floor debates (K. Pearson and Dancey 2011). Women legislators also tend to vote more as a block when it comes to these issues (Swers 1998, 2002; Dolan 1997).⁷

Women legislators are not only more likely to speak out and act on issues related to the protection of children and gender violence, women legislators are effective in legislating on these issues. As a result, we expect that more women legislators lead to more humanitarian military interventions. Women legislators will not always advocate for military interventions. There are other policy alternatives available instead of military interventions that can be considered. For example, a state may choose to disperse foreign assistance. However, the effectiveness of foreign aid assistance during humanitarian crises is limited if at-risk groups cannot access it. During the Rwandan crisis in 1997, the UN high commissioner for refugees admitted that criminals had to be bribed in order for aid to reach women and children (Weiss 2005, 101). Thus, resources are diverted away from groups that need assistance because of the power asymmetries within refugee populations. Without security commitments, it is difficult to ensure that aid disbursement would help at-risk groups. Thus, we focus on governments’ policies related to humanitarian military interventions.⁸

We outline three main mechanisms that connect numbers of women legislators to humanitarian military interventions. First, we expect that as more women enter a legislature, they will form strategic coalitions that will grow more influential with increased numbers. Growing strategic women coalitions should not only increase voting influence, but it may sway men’s behavior, causing both male and female legislators to pay more attention to humanitarian crisis issues. We expect that the probability that men are influenced increases as more women enter office. Some scholars note that the opposite effect can occur; as more women enter office, men legislators will concentrate their efforts to defeat policies that favor women (Childs and Krook 2009). We suspect that this potential backlash is less likely to happen over humanitarian military interventions because this policy is less likely to be either identified as a women’s issue or a threat to men’s dominant position in government.

Women legislator’s ability to influence men on this issues is necessary, given that men have traditionally dominated legislative processes across most countries. If men do not perceive that the politics surrounding humanitarian military interventions are threatening their position of political dominance in governance, then we expect that they will not oppose these types of interventions on the grounds of their relevance to women’s issues. Similarly, we expect that men legislators will be more likely to be influenced by women legislators if humanitarian military interventions are not framed or perceived as women issues, but rather around broader issues of security and stability.

The second reason why we expect more women legislators will impact humanitarian military intervention decisions is that the shared experiences of child protection and gender violence not only increase the probability that women legislators will become transnational surrogate representatives for women and children outside their borders, but we expect it also increase the possibility that a critical actor emerges to advocate for humanitarian military interventions. We expect instances where critical women legislators may overcome the constraints placed by their minority status. These are the actors “who act individually or collectively to bring about women-friendly policy change” (Childs and Krook 2009, 127). We expect that identity issues and shared experiences motivate actors to become critical actors. In terms of humanitarian military interventions, we expect women to be more likely to emerge as critical actors because of the salience of the relevant issues—child protection and gender violence prevention.

While critical actors on issues surrounding humanitarian military interventions do not necessarily have to be women, they will more likely be women. While men may share the same empathy as women legislators, we expect some divergence on how much men are willing to advocate for humanitarian military interventions. We think that more women in legislatures will increase the likelihood that a critical actor emerges, which is consistent with our theoretical expectations. In addition, while critical actors’ effect on policy outcomes may be independent of the number of women legislators, we expect that the likelihood that a critical actor emerges is a function of the number of women legislators. In other words, more women legislators increase the sample from which a critical actor may emerge.

Finally, we expect women legislators to be more effective in legislating on humanitarian military intervention issues. Gender bias in the electoral process prompts only the most talented female candidates to succeed in elections. Thus, the women who do win elections will perform their job better (Anzia and Berry 2011). This expectation is consistent with research in American Politics that finds that US Congresswomen are more likely to deliver federal spending to their home districts (Anzia and Berry 2011), and women are more likely to get their sponsored legislation further through the legislative process (Wiseman, Volden, and Wittmer 2013). This is also consistent with research that finds that legislatures with higher proportions of women introduce and pass more bills on women’s issues than legislatures with lower proportions of women (Bratton 2005; Thomas 1991, 1994). While this empirical evidence comes within the American political context, the gender bias prevalent in many countries’ suggests more general empirical implications. Thus, we expect that women in many legislatures are highly effective in advocating on issues that are of interest to women.

Empirical Implications

The protection of children and gender violence are not new issues or recent characteristics of humanitarian crises. However, more women in positions of power in government is a relatively new phenomenon.⁹ As a result, women legislators are in a better position to advocate for military interventions in humanitarian crises to protect children and prevent sexual violence against women. We expect that as more women gain power in a country’s legislature, that country will be more likely to carry out humanitarian military missions.

From this discussion, we derive the following hypothesis:

The more women legislators in a state’s government increase the likelihood that a state will participate in a humanitarian military intervention.

Some scholars question the theoretical and empirical connection between women in legislatures and conflict issues. Bjarnegård and Melander (2013) find that although communist countries tend to have more women in legislatures, this had no effect on intrastate conflict in these countries. The authors question the essentionalist view that women are less violent than men, and they argue that women legislators have varied interests and priorities. Thus, it is hard to connect numbers of women to specific policy outcomes. We agree with Bjarnegård and Melander’s (2013) concerns, but at the same time think that women in legislatures can have an important causal effect on policies where outcomes have higher political salience for women.

Conflict and peace issues concern many different dimensions that may prompt varying policy positions from women. We do not question that women legislators are a heterogeneous group of political actors with varied preferences. We also do not assume that women are more peaceful than men. However, if the protection of children and prevention of sexual violence are salient issues related to conflict, we expect that women are more likely to push for humanitarian military interventions. In addition, as more women gain seats in legislatures, we expect that states will be more likely to participate in humanitarian military interventions.

Our theoretical framework is consistent with Bjarnegård and Melander’s (2013) empirical results. Our theoretical mechanism focuses on the relationship between women legislators and conflict as conditional on the issues of the protection of children and prevention of sexual violence. If the issues surrounding a conflict are dominated by other issues, we would expect that women legislators would have more varied preferences toward policy options. For example, Bjarnegård and Melander examine intrastate armed conflict. We expect that intrastate of conflict to be dominated by strategic, material, ideological, ethnic, and other concerns. In these types of nonhumanitarian conflicts, our theory derives no clear theoretical relationship between women in legislatures and the proclivity of the use of force.

Another issue with our theory is whether women legislators actually affect policy outcomes. Some scholars suggest that a critical mass of women legislators is needed in order for women to impact policies. This threshold is commonly thought to be at 35 percent representation, though it varies across studies (Kanter 1977). We do not expect this critical mass threshold to be relevant for our theoretical expectations for two reasons. First, critical mass is often expected because women legislators face backlash from their male counterparts. As more women enter office, male officials will fight harder against an agenda that favors women’s issues. It is not until women reach a critical mass, can they affect policy. However, we do not expect men to advocate against humanitarian military interventions because it is unlikely that men would see them as advancing women issues. Additionally, we find it plausible that men may support humanitarian military interventions, but for different reasons than women (i.e., state security concerns instead of human security concerns). We only expect that women legislators are more likely to advocate for humanitarian military interventions than their male counterparts.

The second reason we do not believe that critical mass is the appropriate theoretical model for our article is that critical mass has little supporting empirical evidence. Childs and Krook (2009) review a host of studies that fail to find a critical mass relationship between women representation and policy outcomes.¹⁰ This lowers our expectation that the correct theoretical relationship between women in legislatures and humanitarian military conflict is defined by critical mass. Ultimately, we believe that it is an empirical question, one that we examine in the next section.

Empirical Analysis

To test our hypothesis, we compile data on humanitarian military interventions and women in legislatures. Our data extends from 1946 to 2003.¹¹ Our unit of observation is country-year. Given the binary nature of our dependent variable, the main analysis utilizes logistic regressions.

Measuring Intervention

Our dependent variable is a binary measure of international military interventions (IMIs) taken from F. S. Pearson and Baumann (1993) and Kisangani and Pickering (2008). These authors define intervention as the “movement of regular troops or forces of one country inside another, in the context of some political issue or dispute” (F. S. Pearson and Baumann 1993, 1). In order for a state to be considered as an intervener, its troops must have a compellent or deterrent role. Thus, states that provide troops to oversee infrastructure rebuilding or to provide medical relief are not coded as interveners.¹²

The advantage of the IMI data set on military action over other conflict data sets is that IMI codes the motivations of conflict, such as economic, territory, strategic, social, regime changes, domestic disputes, and humanitarian reasons.¹³ To test our hypothesis, we focus on interventions that are motivated by humanitarian and social issues. The IMI data set defines social interventions as interventions designed to “protect socio-ethnic factions or minorities.” Humanitarian interventions are designed to “save lives, relieve suffering, distribute foodstuffs to prevent starvation and so forth apart from the protection of minorities” (as captured in the social intervention measure). We assume that social and humanitarian issues will disproportionately affect women and children, making these issues more salient to women legislators in other countries.¹⁴ The dependent variable is coded as 1 if a state initiates a military intervention for humanitarian or social reasons in a given year and is coded as 0 otherwise.¹⁵

Key Explanatory Variable: Women in Legislatures

As argued above, humanitarian or social issues disproportionately affect women and children, making these particular issues more salient to women in other countries’ legislatures. More women in legislatures increase the probability that a state will intervene in another state for humanitarian or social reasons. To measure the percentage of women in state legislatures, we rely on data from Paxton, Green, and Hughes (2008), which extends from 1946 to 2003. While the use of these data presents some temporal limitations, our focus on within-state variation, instead of between state variation, mitigate our concerns that the lack of data post 2003 is somehow biasing our results. In addition, we examine a sample of post–Cold War states below and find consistent results to our main inferences.

One implication of critical actors within legislatures is their effect on policy outcomes may be independent of the number of women legislators. This might counter our linear assumption of the relationship between women legislators and humanitarian military interventions. We, however, do not believe that this would threaten our inferences from our statistical results. If critical actors, independent of numbers of women legislators, were the causal driving force between gender and humanitarian military intervention, we would expect the coefficients of our analysis to be biased downward. In other words, if critical actors totally negated the importance of representative numbers, we would expect a null result. Thus, we expect that our test provides a conservative estimation on the effect of women in legislatures.

Theoretically, we expect that more women legislators in a state’s government increase the likelihood that states will participate in humanitarian military interventions. Instead of focusing on women within specific political parties, we expect that issues of child protection and sexual violence prevention increase the probability that women will work together, across party lines, to pressure their government to intervene militarily. Rwandan legislator Connie Bwiza Sekamana explained that when it relates to women issues, “we unite as women, irrespective of political parties. So we don’t think of our parties, [we think of] the challenges that surround us as women” (quoted in Powley 2007, 6).

This study focuses on women in legislatures as opposed executive positions of power for two reasons. First, the rarity of women as executives may limit the generalizability of the empirical inferences. Second, women in executive positions face pressures to combat gender stereotypes and thus will act more “hawkish” (Koch and Fulton 2011). The intersection of the stereotype pressures and humanitarian military interventions is an interesting dilemma for women in power, but outside of the scope of this article. We expect that women in legislatures, as opposed to women in the executive, face less pressure to contradict negative stereotypes. This idea is supported by a quote from Republican US representative Marge Roukema:

I didn’t really want to be stereotyped as the woman legislator.… I wanted to deal with things like banking and finance. But I learned very quickly that if the women like me in Congress were not going to attend to some of these family concerns, whether it was for jobs or children, pension equity, or whatever, then they weren’t going to be attended to. So I quickly shed those biases that I had and said “Well nobody else is going to do it; I’m going to do it.” (Quoted in Carroll 2002, 55)

Descriptive Data

To gauge the initial plausibility of our theory, we explore the basic structures of our data. Figure 1 plots humanitarian, geopolitical, and the total number of interventions on the same graph. We narrowly code geopolitical interventions as military force motivated by territorial, strategic, or military-diplomatic concerns (as defined by the IMI data set). The different colored segments in Figure 1 show the proportion of countries in each year that intervene militarily in another country for humanitarian reasons, geopolitical reasons, and all other reasons (i.e., economic and policy change, etc.). The graph shows that humanitarian interventions are a smaller proportions of all interventions except in 2003. In addition, there is no clear temporal trend in the data except the paucity of interventions in the early part of the data and a spike of humanitarian interventions in 2003. We further explore these temporal anomalies in the robustness section.

Figure 1.

Military interventions, 1946–2003.

Cross-sectionally, there are fifty-nine countries that have been motivated to militarily intervene in another country for humanitarian reasons. As shown in the map in Figure 2, these countries are geographically, politically, and economically diverse.¹⁶ However, when we examine the number of humanitarian interventions of a country, we observe some countries are more likely to intervene in multiple years, such as the United States, France, and the United Kingdom. In sum, there appears to be temporal, within, and between variations in our dependent variable.

Figure 2.

Humanitarian military interveners, 1946–2003. The gray areas indicate countries that have not intervened.

Turning to our main explanatory variable, we find sufficient levels of variance in the percentage of women in governing legislatures. The left panel in Figure 3 shows the distribution of women in legislature over the duration of our sample. While the modal observation is 0, the data extend up to 50 percent of the legislature. The right panel shows how the mean percentage of women in legislatures has changed over time, with a general increase.

Figure 3.

Women in legislatures, 1946–2003.

We subset the sample to illustrate the variation of women in legislatures within some countries. Figure 4 demonstrates that there is variation across counties and within countries. Some countries, such as the United States, show subtle increases of women in government, while the end of apartheid led to more dramatic changes in South Africa.

Figure 4.

Women in legislatures, 1946–2003.

Our theory asserts that the variance in these variables are connected, specifically that as more women enter legislative office, the more likely a state will militarily intervene for humanitarian reasons. However, since women in government is a nonrandom phenomena, the effect of having more women in government on humanitarian military interventions is potentially nonignorable.¹⁷ To address this potential issue, we turn now to multivariate analysis in an attempt to model the causal effect of women legislators as being conditionally ignorable.

Multivariate Analysis

Given that women in legislatures is a nonrandom phenomena, we cannot identify an average causal effect without conditioning the data. We attempt to condition a causal effect by controlling for potential confounders and by controlling for unit-specific heterogeneous factors. To model alternative causal pathways of humanitarian military interventions, we include a set of control variables (Pearl 2000; Morgan and Winship 2007). We expect that these covariates may change the likelihood of states intervening in humanitarian issues, but should not completely explain the decision to intervene.

In this section, we describe the variables we include in all models that could possibly affect women in legislatures and humanitarian interventions. We then describe the estimation approaches we use to identify the effect of women legislators on humanitarian interventions. Finally, we describe our results and various robustness checks.

Control Variables

We first include a measure for military capacity, capabilities, which combines Correlates of War National Material Capabilities (v4.0) data on energy, military troops, industrial production, military spending, and urban population. Given the left skewed nature of this measure, we take the natural log of the capabilities measure. We expect that states with high levels of military capacity are more able to execute military interventions for humanitarian reasons. It may also be the case that capabilities—given its components—is correlated with some latent development characteristic that explains the levels of women in legislatures. More developed countries may be more likely to have more women in positions of power.

Next, we include a regime-type variable to capture the democratic levels in a country. We use Cheibub, Gandi, and Vreeland’s (2010) binary regime classification which distinguishes between democracies and nondemocracies. We expect that democracies are more likely to have women in positions of power and may be more likely to be motivated to intervene in humanitarian and social issues. We examine alternative measures of democracy in the robustness section.

Finally, we include the variable years peace that measures the years since the last humanitarian military intervention by a state.¹⁸ We expect that states more recently involved in a humanitarian intervention will be more likely to be involved in another intervention. In addition, it may be the case that recent humanitarian interventions prompt more women to seek office. By including this measure, we control for this potential alternative causal path.

We limit the amount of control variables in our preferred model specification to decrease the likelihood of multicollinearity (Achen 2006; Clarke 2005; Schrodt 2014). However, in the robustness section below, we include additional control variables to account for alternative theoretical explanations. As shown below, these additional controls do not affect the substantive inferences from the main results.

Primary Estimation Approach

Our analysis is potentially impacted by unobserved unit-specific factors that help explain both the percentage of women in legislatures and humanitarian military interventions. For example, colonial history, culture, communist legacy, or other historical state characteristics may make it more (less) likely for a state to have more women in government and also be more (less) likely to become involved in a military intervention for humanitarian or social reasons.¹⁹ In other words, women “representation implies very different things in different settings” (Bjarnegård and Melander 2013, 559).²⁰ Not modeling this unobserved or unmeasured heterogeneity opens the possibility of omitted variable bias.

One solution to this potential heterogeneity problem is using a “fixed-effects” estimator. This estimation approach provides a within-state comparison of the effect of women in legislatures on humanitarian interventions. A limitation of this estimation approach is that since 125 of the 184 countries in the data never intervene for humanitarian reasons, these observations would be dropped from the analysis. This deletion would increase the possibility of inconsistent estimates (King 2001). A random effects estimator provides a logical alternative, which allows the full sample to be analyzed. The trade-off is a stronger assumption about the independence between the unobserved unit-specific factors and the observed covariances (Wooldridge 2001, 252). While we cannot test this assumption, we run both random and fixed-effects models and find the same substantive results. The Hausman test reveals that the fixed-effects model is more appropriate than a random effects model (on the limited sample). As a result, we rely on fixed-effects modeling, with the understanding that we are truncating our sample. To ensure that this truncation is not biasing our results, we also run our analysis using random effects.

Employing a fixed-effects estimator helps us analyze states that may have a low baseline proclivity to use force and instead prefer to employ alternative strategies. For example, Nordic countries, which traditionally have high levels of representation of women and little history of military intervention, favor alternative strategies to military intervention, such as humanitarian assistance and international organization involvement. This does not necessarily contradict our expectation since we are interested in an average effect. We expect that more women in legislatures increase the likelihood of humanitarian military interventions, although this likelihood may begin at a low baseline rate. Thus, we expect that increases to the already high representation of women should increase the likelihood that these states initiate humanitarian military interventions.

In addition to unit-specific effects, we account for temporal dynamics in our analysis. We expect that states recently involved in humanitarian military interventions are more likely to become involved again. Following Carter and Signorino (2010), we use the squared and cubed value of years since humanitarian intervention to account for temporal dependence in our main models. We discuss in the robustness section additional analysis that addresses temporal dynamics.

Results

Model 1 in Table 1 examines how women in legislatures affect the probability of humanitarian military interventions using a fixed-effects estimator. The coefficient of interest, women in legislatures, is expected to be positive, as more women legislators within a state should increase the probability that the state in question initiates a humanitarian military intervention. The issues surrounding humanitarian crises—protection of children and prevention of sexual violence—are more likely to provoke a response in women legislators in other countries. As more women enter legislatures, the more likely we will observe humanitarian military interventions. The women in legislatures estimate is positive and statistically significant, consistent with our argument that more women in office will result in more humanitarian interventions.

Table 1.

Logit Examining Humanitarian Military Interventions, 1946–2003.

	Model 1	Model 2	Model 3	Model 4
	Fixed effects	Random effects	Advanced democracies	Other interventions
	β/SE	β/SE	β/SE	β/SE
Women in legislature (%)	0.07*** (0.02)	0.04** (0.01)	0.10*** (0.03)	−0.01 (0.01)
Democracy (CGV)	1.09 (0.60)	0.99*** (0.27)		−0.12 (0.23)
ln(military capabilities)	−1.39*** (0.39)	0.45*** (0.07)	−0.40 (0.50)	0.22 (0.19)
Years peace	0.11* (0.05)	0.04 (0.05)	−0.01 (0.08)	−0.10** (0.03)
Years peace²	−0.01** (0.00)	−0.01** (0.00)	−0.01 (0.00)	0.00 (0.00)
Years peace³	1.36*** (0.35)	1.03*** (0.27)	1.14 (0.62)	0.00 (0.00)
Constant		−2.09*** (0.57)
ln( $σ_{u}^{2}$ )		−0.55 (0.47)
Log likelihood	−344.2	−532.6	−147.0	−980.1
AIC	700.4	1,081.3	304.1	1,972.2
N	2,509	6,356	723	3,915

Note: β = coefficient estimates; SE = standard errors; CGV = Chebiub, Gandhi, and Vreeland; AIC = Akaike information criterion.

*p < .05.

**p < .01.

***p < .001.

Substantively, the odds ratio (not shown) derived from the logit model estimates that a 1 percentage point increase in women representation in a given state increases the likelihood that a state is involved in a humanitarian military intervention by 7 percent. We also provide a visual representation of the predicted probability of humanitarian military given women in legislatures. Given that our fixed-effects estimation provides a within-unit comparison, we examine predicted probabilities as a function of states’ deviation from their mean levels of women in legislatures.²¹ Figure 5 shows that states with levels of women in legislatures below their mean (i.e., negative deviations) have substantively low probabilities of initiating a humanitarian military intervention. Conversely, as the level of women in legislatures exceeds a state’s mean level (i.e., positive deviations), the predicted probability of humanitarian interventions increases. Across the range of deviations demonstrates a seven-fold increase in the probability of humanitarian interventions. This estimated effect is quite large, given that previous gender and conflict research would either expect no effect or even the reverse directional relationship.

Figure 5.

Predicted probabilities of humanitarian interventions, 1946–2003.

Model 2 uses a random-effect estimator. As mentioned above, the limitation of the fixed effects estimation strategy is that states that have never intervened for humanitarian or social reasons are dropped from the sample, which amounts to 61 percent of the sample. The advantage of fixed-effects estimation is that it controls for time-invariant unit-specific confounders (i.e., history, culture, colonial history, communist legacy, etc.). The results of the random-effects mode are consistent with the fixed-effects model. We also replicate the other models in this table using random effects and find substantively the same results.²²

The next model, model 3, examines the role of regime type. The theory outlined above implicitly argues that the relationship between women in legislatures and humanitarian military intervention holds across regime types. While democracies may have institutional rules that allow legislative members more influence over the foreign policies, nondemocracies also create institutional and party structures that can influence policy (Svolik 2012). Nondemocratic regimes have incentives to develop these types of representative mechanisms because they allow regimes to survive longer than regimes that do not develop these institutions (Geddes 1999). However, Bjarnegård and Melander (2013) argue that authoritarian communist regimes install more women in legislative bodies that have little influence over policies. We agree that there may be differences in the impact of women legislators across political systems. However, the effect of having more women legislators within a political system—which we can estimate using unit fixed-effects—should still theoretically affect the proclivity of humanitarian military interventions. Nonetheless, to ensure that we are analyzing comparable units of analysis, we limit our sample to the twenty-two most developed democracies, consistent with Koch and Fulton’s (2011) analysis. Again, we find the more women legislators increase the likelihood of humanitarian military interventions. We further examine the role of regime type in the robustness test section.

Finally, model 4 in Table 1 provides a type of placebo test for our theory. As we argued and tested above, we expect that more women legislators increase the probability of a humanitarian military intervention. We, however, expect women in legislatures to have no effect on the propensity of military interventions motivated by nonhumanitarian or nonsocial reasons. To test this null expectation, we change our dependent variable of interest to all nonhumanitarian military interventions in model 4. Model 4 is designed to ensure that the explanation of the results in the previous models is consistent with our theory and not with the alternative explanation that women are sometimes more hawkish in order to combat negative stereotypes. There is no evidence from the placebo model that women are more or less likely to support nonhumanitarian military interventions, further suggesting that humanitarian military interventions, and not interventions in general, are a function of women in legislatures. We not only find a statistically insignificant result in model 4, but the 95 percent confidence intervals of the estimated effect size in model 4 do not overlap at all with the confidence intervals from the estimated effect in model 1. This supports our claim that model 4 is both a statistically and substantively null result (Rainey 2014).

In sum, the results in Table 1 support our theoretical expectations that more women in legislatures increase the probability that a government militarily will intervene in another country for humanitarian purposes. While our results are robust to several alternative models, we turn now to additional robustness checks to ensure that our inferences are not a function of model specification or other research design decisions.

Additional Robustness Checks

This section describes robustness tests that provide additional supporting evidence and address potential threats to our main inferences. A more detailed description of these tests and the associated tables and graphs are delegated to the supplemental appendix. The results are consistent with our main inference, but warrant some discussion here as they have implications for future research on this topic.

The first set of robustness tests examine and control for additional factors that either help us conditionally model a causal effect or rule out potential confounding factors. The most important confounding factors to control for are societal or governing attitudes that would make it more (less) likely to elect women into the state legislature and be more (less) sympathetic toward international humanitarian or social issues. We attempt to control for these potential factors by including variables that measure women’s rights. Specifically, we include a political rights variable, a social rights variable, and an economic rights variable. These data are drawn from Cingranelli, Richards, and Clay (2014). The inclusion of these variables has no substantive effect on our inferences.

In addition, we include variables that count the number of years since women gained the right to run for office or count the number of years since women first gained at least 10 percent of the seats in the legislature.²³ Both of these variables attempt to measure not only states’ respect for women’s rights but also how well these rights are consolidated into a state’s political culture. We assume that higher levels in the added control variables represent more consolidated attitudes toward women’s rights. Again, these added variables do not change our inferences.

Finally, we control for government ideology. It may be the case that left-leaning governments are more likely to have women officials and are also more likely to be involved in humanitarian interventions (Koch and Fulton 2011). To account for this possibility, we use a weighted mean score of political orientation of the government, drawing our data from Lowe et al. (2011), where positive values indicate more right orientated governments.²⁴ While using this measure does limit the sample to democracies (and therefore we omit the democracy covariate from the analysis), it increases the likelihood that a logical confounder is not omitted. The inclusion of this control variable does not affect our inferences. All of these models with added confounders are found in the supplemental appendix.

We next turn our attention toward time. The descriptive statistics reveal potentially interesting temporal trends in the data. To ensure that these temporal trends do not threaten the validity of our results, we examine time in the supplemental appendix. First, we lag women in legislatures to ensure the temporal ordering of our analysis. Next, we examine the effect of women in legislatures on humanitarian military interventions for the post–Cold War era only (1990–2003). Given the change in polarities in the international system with the end of the Soviet Union, it is likely that the nature and the quantity of military interventions change.²⁵ Then, we control for annual heterogeneity instead of modeling the temporal trends with a cubic duration polynomials. We do this by examining a model with both unit- and year-fixed effects. Finally, we omit 2003 from the analysis. In Figure 1, we observe that there is a spike in humanitarian interventions as a result of IMI coding the 2003 Iraq War as a humanitarian intervention (as well as a strategic intervention). To ensure that this coding decision does not drive our results, we omit 2003 from the analysis. For all these tests, we find consistent results.

Next, we reconsider the role of democratic institutions. As a robustness check, we restricted our analysis to advanced democracies in model 3 in Table 1 to ensure we are analyzing like-units. In this section, we further examine the role of democracy in two ways. First, we include an alternative regime type measure, Polity, drawn from Marshall and Jaggers (2010). Polity is a twenty-one ordinal regime-type measure where higher values of Polity indicate more democratic regimes. The inclusion of this variable does not change our results. Next, we interact women in legislature with both of our binary and ordinal regime type variables. What we find does not change our inferences, but suggests that further research may be needed to further disentangle the role of institutions. We present the marginal effects of the interaction in Figure 6 (the table results are reported in the supplemental appendix). The top panel in Figure 6 shows that a change in women legislators (a standard deviation increase from its mean value) increases the likelihood of humanitarian military interventions in democracies. The same cannot be definitively said for nondemocracies given that the confidence intervals overlap with 0. Similarly, the bottom panel in Figure 6 shows that a change in women legislators (a standard deviation increase from its mean value) increases the likelihood of humanitarian military interventions in regimes with Polity scores higher than 4. Regimes with scores lower than 4 (nondemocracies) have no clear effect. However, we cannot statistically distinguish the effects between democracies and nondemocracies, as demonstrated by the overlapping confidence intervals. So while democracies demonstrate a positive effect and nondemocracies demonstrate a null effect, these effects are not statistically different from one another.

Figure 6.

Marginal effects of increase women in legislature.

We set out to test whether women in legislature had an average effect on humanitarian military interventions, across a broad sample of states. We find evidence that this is indeed the case. The conditional results demonstrated in Figure 6 suggest an avenue that future research considers this topic. Though outside the scope of this article, we speculate that there may be interesting variation within both democratic and nondemocratic regimes that may further condition the effect of women in legislatures.

Other robustness tests we employ include examining the linear functional form of women in legislature. We find that our linear specification provides the best model fit, though there is some evidence that a natural log transformation of our variable of interest may also be appropriate.

In conclusion, our results support our hypothesis that more women in legislatures will increase the probability that a state will undertake a humanitarian military intervention. These results are consistent across several model specifications. More discussion on all the robustness analysis is found in the supplemental appendix.

Conclusion

This article contends that women in national legislatures are a driving force in the decisions to use military force for humanitarian reasons. Our empirical analysis consistently supports this argument as we find that more women in legislatures are associated with a higher proclivity of humanitarian military action.

Our research has implications for the women and politics literature. Within this literature, the debate continues about differences in descriptive and substantive representation of women. While we acknowledge that women are not a homogenous legislative block with uniform interests and beliefs, we have focused on two issues that generally unify women’s interests: protection of children and prevention of sexual violence. The salience of these issues increases that probability that women act as transnational representative surrogates, which affects a state’s propensity to use military force for humanitarian reasons.

Our research also has implications on the domestic politics and conflict processes literature. In a general sense, our research contradicts the conventional wisdom that women are less supportive of the use of force than men. We argue that previous research in this area focused on the use of force in the abstract, ignoring the possibility that issues regarding the use of force could generate women’s support.

Finally, our research has policy implications as more and more women enter office. Our results suggest that as more women enter national legislatures, the more likely states will engage in humanitarian military interventions. Militaries in regimes where women are expected to take more legislative power should thus adapt grand strategies to better deal with humanitarian missions. Given the increase in women representation around the world, it is likely that humanitarian missions will be more prevalent in the near future. However, more women in legislatures may open opportunities for women in higher levels of office, such as executive or cabinet positions. As demonstrated by Koch and Fulton (2011), women in higher levels of government face pressure to combat negative stereotypes about security issues. It is unclear how the pressures to combat stereotypes affect decisions related to humanitarian military interventions, as opposed to other types of military interventions.²⁶ In addition, military interventions may have the opposite intended effect, hurting more women and children, rather than saving women and children (Peksen 2011). Additional research on the role of women in government and military interventions may help disentangle these dynamic political processes.

Footnotes

Authors’ Note

Data and replication files are available online at . Please direct questions regarding replication to pshea@uh.edu.

Acknowledgments

The authors thank Sara Angevine, Kelly Dittmar, Ryan Kennedy, Justin Kirkland, Johanna Luttrell, Paul Huth, and two anonymous reviewers for their helpful comments and suggestions.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Supplementary Material

Supplementary material for this article is available online.

Notes

References

Achen

C. H.

2006. “Let’s Put Garbage-can Regressions and Garbage-can Probits Where They Belong.” Conflict Management and Peace Science 22 (4): 327–39.

Angevine

2014. “Women’s Rights Foreign Policyin the US Congress.” PhD thesis, Rutgers University, New Brunswick, NJ.

Anzia

S. F.

Berry

C. R.

. 2011. “The Jackie (and Jill) Robinson Effect: Why Do Congresswomen Outperform Congressmen?” American Journal of Political Science 55 (3): 478–93.

Bjarnegård

Melander

. 2013. “Revisiting Representation: Communism, Women in Politics, and the Decline of Armed Conflict in East Asia.” International Interactions 39 (4): 558–74.

Bratton

K. A.

2005. “Critical Mass Theory Revisited: The Behavior and Success of Token Women in State Legislatures.” Politics & Gender 1 (1): 97–125.

Breuning

2001. “Women’s Representation and Development Assistance: A Cross-national Study.” Women & Politics 23 (3): 35–55.

Caprioli

2000. “Gendered Conflict.” Journal of Peace Research 37 (1): 51–68.

Caprioli

Boyer

M. A.

. 2001. “Gender, Violence, and International Crisis.” Journal of Conflict Resolution 45 (4): 503–18.

Carroll

S. J.

2002. “Representing Women: Congresswomen’s Perceptions of Their Representational Roles.” In Women Transforming Congress, edited by Rosenthal

C. S.

, University of Oklahoma Press: Norman, OK, 50–68.

10.

Carter

D. B.

Signorino

C. S.

. 2010. “Back to the Future: Modeling Time Dependence in Binary Data.” Political Analysis 18 (3): 217–92.

11.

Celis

Childs

Kantola

Krook

M. L.

. 2008. “Rethinking Women’s Substantive Representation.” Representation 44 (2): 99–110.

12.

Center for American Women and Politics. 2001. “Women State Legislators: Past, Present, and Future.” Accessed September 10, 2015. http://www.cawp.rutgers.edu/research/topics/documents/StLeg2001Report.pdf

13.

Cheibub

J. A.

Gandi

Vreeland

J. R.

. 2010. “Democracy and Dictatorship Revisited.” Public Choice 143:67–101.

14.

Childs

2006. “The Complicated Relationship between Sex, Gender and the Substantive Representation of Women.” European Journal of Women’s Studies 13 (1): 7–21.

15.

Childs

Krook

M. L.

. 2008. “Critical Mass Theory and Women’s Political Representation.” Political Studies 56 (3): 725–36.

16.

Childs

Krook

M. L.

. 2009. “Analysing Women’s Substantive Representation: From Critical Mass to Critical Actors.” Government and Opposition 44 (2): 125–45.

17.

Cingranelli

D. L.

Richards

D. L.

Clay

K. C.

. 2014. “The CIRI Human Rights Dataset.” Version 2014.04.14. Accessed August 20, 2015. http://www.humanrightsdata.com.

18.

Clarke

K. A.

2005. “The Phantom Menace: Omitted Variable Bias in Econometric Research.” Conflict Management and Peace Science 22 (4): 341–52.

19.

Conover

P. J.

Sapiro

. 1993. “Gender, Feminist Consciousness, and War.” American Journal of Political Science 37 (4): 1079–99.

20.

Dodson

D. L.

2006. The Impact of Women in Congress. New York: Oxford University Press.

21.

Dolan

1997. “Support for Women’s Interests in the 103rd Congress.” Women & Politics 18 (4): 81–94.

22.

Eichenberg

R. C.

2003. “Gender Differences in Public Attitudes toward the Use of Force by the United States, 1990-2003.” International Security 28 (1): 110–41.

23.

Eichenberg

R. C.

Stoll

. 2012. “Gender Difference or Parallel Publics? The Dynamics of Defense Spending Opinions in the United States, 1965-2007.” Journal of Conflict Resolution 56 (2): 331–48.

24.

Ferris

1993. Women, War, and Peace. Technical report. Uppsala, Sweden: Life & Peace Institute.

25.

Fisher

2010. “Violence against Women and Natural Disasters: Findings from Post-Tsunami Sri Lanka.” Violence Against Women 16 (8): 902–18.

26.

Geddes

1999. “What Do We Know about Democratization after Twenty Years?” Annual Review of Political Science 2:115–44.

27.

Goldstein

J. S.

2001. War and Gender. Cambridge, UK: Cambridge University Press.

28.

Hawkesworth

Casey

K. J.

Jenkins

Kleeman

K. E.

. 2001. Legislating by and for Women: A Comparison of the 103rd and 104th Congresses. Brunswick, NJ: Center for American Women and Politics.

29.

Holsti

O. R.

Rosenau

J. N.

. 1990. “The Structure of Foreign Policy Attitudes among American Leaders.” The Journal of Politics 52 (1): 94–125.

30.

Hudson

V. M.

Caprioli

Ballif-Spanvill

McDermott

Rose

Emmett

C. F.

. (2008) 2009. “The Heart of the Matter: The Security of Women and the Security of States.” International Security 33 (3): 7–45.

31.

Huth

Allee

T. L.

. 2002. “Domestic Political Accountability and the Escalation and Settlement of International Disputes.” Journal of Conflict Resolution 46 (6): 754–90.

32.

Inter-Parliamentary Union and UN Women. 2015. “Women in Politics: 2015.” Accessed September 10, 2015. http://www.ipu.org/pdf/publications/wmnmap15_en.pdf.

33.

Jentleson

B. W.

1992. “The Pretty Prudent Public: Post Post-Vietnam American Opinion on the Use of Military Force.” International Studies Quarterly 36:49–73.

34.

Jentleson

B. W.

Britton

R. L.

. 1998. “Still Pretty Prudent Post-cold War American Public Opinion on the Use of Military Force.” Journal of Conflict Resolution 42 (4): 395–417.

35.

Kanter

R. M.

1977. “Some Effects of Proportions on Group Life: Skewed Sex Ratios and Responses to Token Women.” American Journal of Sociology 82 (5): 965–90.

36.

King

2001. “Proper Nouns and Methodological Propriety: Pooling Dyads in International Relations Data.” International Organization 55 (2): 497–507.

37.

Kirshner

2007. Appeasing Bankers: Financial Caution on the Road to War. Princeton, NJ: Princeton University Press.

38.

Kisangani

E. F.

Pickering

. 2008. “International Military Intervention, 1989-2005.” Inter-university Consortium for Political and Social Research, Data Collection 21282, University of Michigan, Ann Arbor.

39.

Koch

M. T.

Fulton

S. A.

. 2011. “In the Defense of Women: Gender, Office Holding, and National Security Policy in Established Democracies.” The Journal of Politics 73 (1): 1–16.

40.

Lawless

J. L.

2004. “Women, War, and Winning Elections: Gender Stereotyping in the Post-September 11th Era.” Political Research Quarterly 57 (3): 479–90.

41.

Lobell

S. E.

2008. “Second Face of Security Strategies: Anglo-German and Anglo-Japanese Trade Concessions During the 1930s.” Security Studies 17 (3): 438–67.

42.

Lowe

Benoit

Mikhaylov

Laver

. 2011. “Replication Data for: The Manifesto Project Data Extended to Include the Logit Scales and Standard Errors.” Accessed August 20, 2015. https://hdl-handle-net.web.bisu.edu.cn/1902.1/17073

43.

Mansbridge

2003. “Rethinking Representation.” American Political Science Review 97 (4): 515–28.

44.

Marshall

M. G.

Jaggers

. 2002. Polity IV Project: Political Regime Characteristics and Transitions, 1800–2002. Version p4v2010e. College Park, MD: Center for International Development and Conflict Management, University of Maryland. Accessed January 24, 2014. http://www.systemicpeace.org/inscrdata.html

45.

Marshall

M. G.

Ramsey

. 1999. “Gender Empowerment and the Willingness of States to Use Force.” Unpublished research paper, Center for Systemic Peace. Accessed September 10, 2015. http://www.members.aol.com/CSPmgm

46.

Mazurana

Benelli

Walker

. 2013. “How Sex- and Age-disaggregated Data and Gender and Generational Analyses Can Improve Humanitarian Response.” Disasters 37 (s1): S68–82.

47.

Mendelberg

Karpowitz

C. F.

Goedert

. 2014. “Does Descriptive Representation Facilitate Women’s Distinctive Voice? How Gender Composition and Decision Rules Affect Deliberation.” American Journal of Political Science 58:291–306.

48.

Moon

Souva

. 2016. “Audience Costs, Information, and Credible Commitment.” Journal of Conflict Resolution 60 (3): 434–458 doi: 10.1177/0022002714545222

49.

Morgan

S. L.

Winship

. 2007. Counterfactuals and Causal Inference: Methods and Principles for Social Research. New York: Cambridge University Press.

50.

Murdie

Peksen

. 2014. “The Impact of Human Rights INGO Shaming on Humanitarian Interventions.” The Journal of Politics 76:215–28.

51.

Narizny

2007. The Political Economy of Grand Strategy. Ithaca, NY: Cornell University Press.

52.

Niarchos

C. N.

1995. “Women, War, and Rape: Challenges Facing the International Tribunal for the Former Yugoslavia.” Human Rights Quarterly 17 (4): 649–90.

53.

Nincic

D. J.

. 2002. “Race, Gender, and War.” Journal of Peace Research 39 (5): 547–68.

54.

Paxton

Green

Hughes

. 2008. “Women in Parliament, 1945-2003: Cross-national Dataset.” Inter-university Consortium for Political and Social Research 24340, University of Michigan, Ann Arbor.

55.

Pearl

2000. Causality: Models, Reasoning, and Inference. New York: Cambridge University Press.

56.

Pearson

F. S.

Baumann

R. A.

. 1993. “International Military Intervention, 1946-1988.” Inter-university Consortium for Political and Social Research, Data Collection 6035, University of Michigan, Ann Arbor.

57.

Pearson

Dancey

. 2011. “Speaking for the Underrepresented in the House of Representatives: Voicing Women’s Interests in a Partisan Era.” Politics & Gender 7 (4): 493–519.

58.

Peksen

2011. “Foreign Military Intervention and Women’s Rights.” Journal of Peace Research 48 (4): 455–68.

59.

Phillips

2006. “Food Security and Women’s Health: A Feminist Perspective for International Social Work.” International Social Work 52 (4): 485–98.

60.

Plumper

Neumayer

. 2006. “The Unequal Burden of War: The Effect of Armed Conflict on the Gender Gap in Life Expectancy.” International Organization 60 (3): 723–54.

61.

Powley

2007. “Rwanda: The Impact of Women Legislators on Policy Outcomes Affecting Children and Families.” UNICEF: The State of the World’s Children. Accessed September 10, 2015. http://www.unicef.org/sowc07/docs/powley.pdf

62.

Rainey

2014. “Arguing for a Negligible Effect.” American Journal of Political Science 58 (4): 1083–91.

63.

Regan

P. M.

Paskeviciute

. 2003. “Women’s Access to Politics and Peaceful States.” Journal of Peace Research 40 (3): 287–302.

64.

Reiter

2015. “The Positivist Study of Gender and International Relations.” Journal of Conflict Resolution 59 (7): 1301–1326.

65.

Schrodt

P. A.

2014. “Seven Deadly Sins of Contemporary Quantitative Political Analysis.” Journal of Peace Research 51 (2): 287–300.

66.

Schultz

2001. Democracy and Coercive Diplomacy. Cambridge, UK: Cambridge University Press.

67.

Seager

2005. “Noticing Gender (or Not) in Disasters.” Social Policy 36:29–30.

68.

Senese

P. D.

Vasquez

J. A.

. 2008. The Steps to War: An Empirical Study. Princeton, NJ: Princeton University Press.

69.

Shapiro

R. Y.

Mahajan

. 1986. “Gender Differences in Policy Preferences: A Summary of Trends from the 1960s to the 1980s.” The Public Opinion Quarterly 50 (1): 42–61.

70.

Shea

Teo

T. K.

Levy

J. S.

. 2014. “Opposition Politics and International Crises: A Formal Model.” International Studies Quarterly 58 (4): 741–51.

71.

Smith

T. W.

1984. “The Polls: Gender and Attitudes Toward Violence.” Public Opinion Quarterly 48 (3): 384–96.

72.

Snyder

1991. Myths of Empire: Domestic Politics and International Ambition. Ithaca, NY: Cornell University Press.

73.

Svolik

M. W.

2012. The Politics of Authoritarian Rule. New York: Cambridge University Press.

74.

Swers

M. L.

1998. “Are Women More Likely to Vote for Women’s Issue Bills than Their Male Colleagues?” Legislative Studies Quarterly 23 (3): 435–48.

75.

Swers

M. L.

2002. The Difference Women Make, the Policy Impact of Women in Congress. Chicago: University of Chicago Press.

76.

Swers

M. L.

2006. “Providing for the Common Defense: An Analysis of Gender Differences in Defense Policy Participation in the US Senate.” Legislative Studies Quarterly 31 (1): 136.

77.

Swers

M. L.

2007. “Building a Reputation on National Security: The Impact of Stereotypes Related to Gender and Military Experience.” Legislative Studies Quarterly 32 (4): 559–95.

78.

Swers

M. L.

2013. Women in the club: Gender and policy making in the Senate. Chicago: University of Chicago Press.

79.

Tessler

Nachtwey

Grant

. 1999. “Further Tests of the Women and Peace Hypothesis: Evidence from Cross-national Survey Research in the Middle East.” International Studies Quarterly 43 (3): 519–31.

80.

Tessler

Warriner

. 1997. “Gender, Feminism, and Attitudes Toward International Conflict: Exploring Relationships with Survey Data from the Middle East.” World Politics 49 (2): 250–81.

81.

Thomas

1991. “The Impact of Women on State Legislative Policies.” The Journal of Politics 53 (4): 958–76.

82.

Thomas

1994. How Women Legislate. New York: Oxford University Press.

83.

Togeby

1994. “The Gender Gap in Foreign Policy Attitudes.” Journal of Peace Research 31 (4): 375–92.

84.

Weiss

T. G.

2005. Military-civilian Interactions: Humanitarian Crises and the Responsibility to Protect. Oxford, UK: Rowman & Littlefield.

85.

Weldon

S. L.

2002. “Beyond Bodies: Institutional Sources of Representation for Women in Democratic Policymaking.” The Journal of Politics 64 (4): 1153–74.

86.

Wilcox

Hewitt

Allsop

. 1996. “The Gender Gap in Attitudes toward the Gulf War: A Cross-national Perspective.” Journal of Peace Research 33 (1): 67–82.

87.

Wiseman

Volden

Wittmer

. 2013. “When Are Women More Effective Lawmakers Than Men?” American Journal of Political Science 57 (2): 326–41.

88.

Wooldridge

J. M.

2001. Econometric Analysis of Cross Section and Panel Data. Cambridge, MA: MIT Press.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

3.70 MB

0.00 MB

1.94 MB