Female Mediators and the Comprehensiveness of Peace Agreements

Transcription

1 Female Mediators and the Comprehensiveness of Peace Agreements Sarah Hillbert Master Thesis Spring 2017 Department of Peace and Conflict Research, Uppsala University Supervisor: Kristine Höglund

2 ABSTRACT The acknowledgement of women s important role as mediators in peace negotiations is emerging. However, previous research on the impact of a mediator s gender in mediation processes does not provide a conclusive answer. This study poses the research question: Is the comprehensiveness of a peace agreement influenced by the participation of female mediators, and if so, how? By drawing on theories suggesting mediators ability to influence peace negotiations and theories on the importance of women s presence to address different kinds of issues in peace negotiations, the study hypothesizes that a gender-mixed mediation team increases the chance of reaching a comprehensive peace agreement. A content analysis of the peace agreements met in Liberia 2003, Sudan 2005, Ivory Coast 2007 and DRC 2008 as well as an in-depth analysis of the mediation process of Sudan and Liberia presents evidence suggesting that female mediators do influence peace agreement s comprehensiveness. Still, the study also underlines the importance of taking other factors, such as civil society women s groups active participation, as well as the strategy used by mediators, into account. 2

3 LIST OF ABBREVIATIONS AU AFDL CNL CPA DDR EC ECOWAS EU FLNC DRC ICGLR IGAD LURD MARWOPNET MLC MODEL MPCI NGO NPA NPFL NPP NTGL OAU PDCI-RDA RCD SPLM/A UCDP UN UNSC UN SRSG WIPNET African Union Alliance des Forces Démocratiques pour la Libération du Congo Conseil National de Liberation Comprehensive Peace Agreement 1 Disarmament, Demobilization and Reintegration European Commission Economic Community of West African States European Union Front de Libération Natinale Congolais Democratic Republic of Congo International Conference on the Great Lakes Region Inter-governmental Authority on Development Liberians United for Reconciliation and Democracy The Mano River Women Peace Network Mouvement de Libération Congolais Movement for Democracy in Liberia The Patriotic Movement of Côte d'ivoire Non-governmental organization Norwegian People s Aid National Patriotic Front of Liberia National Patriotic Party National Transitional Government of Liberia Organisation of African Unity Parti Démocratique de la Côte d Ivoire-Rassemblement Démocratie African Rassemblement Congolais pour la Démocratic Sudan People s Liberation Movement and Army Uppsala Conflict Data Program United Nations United Nations Security Council United Nations Special Representative of the Secretary-General The Women in Peace Building Program 1 This is the name of the peace agreement met on Sudan 9 January

4 TABLE OF CONTENTS 1. INTRODUCTION PREVIOUS RESEARCH Women in peace processes Women and mediation Research gap THEORETICAL FRAMEWORK Mediators ability to affect negotiations Women as mediators Causal chain RESEARCH DESIGN AND METHOD Case selection Definition of comprehensive peace agreement Sources FEMALE MEDIATORS AND THE SCOPE OF PEACE AGREEMENTS Sudan DRC Liberia Ivory Coast The scope of peace agreements Conclusion: scope of peace agreements IN-DEPTH CASE STUDY: SUDAN Context prior and during mediation process Tracing the mechanism: agenda setting, interacting with parties and participation of women s groups Conclusions of case analysis: Sudan IN-DEPTH CASE STUDY: LIBERIA

5 7.1 Context prior and during mediation process Tracing the mechanism: agenda setting, interacting with parties and participation of women s groups Conclusions of case analysis: Liberia COMPARATIVE ANALYSIS Between-case comparison Reflections on the findings Alternative explanations and additional observations CONCLUSIONS BIBLIOGRAPHY

6 1. Introduction Bringing women to the peace table improves the quality of agreements reached, and increases the chances of successful implementation (Ban Ki-Moon 2009) Since the end of the Cold War, negotiations resulting in peace agreements is the most commonly used method to end armed conflict. Often, a third party having the role of a mediator is present during the negotiations. According to several scholars, mediation is the most promising, and in many ways, the most acceptable method to use when handling conflicts (Bercovitch & Fretter 2007). However, despite rigorous research on mediation, the effect of different types of mediators and their ability to influence the mediation process is still debated. In recent years, the importance of including women in peace processes has gained greater recognition in research as well as by world leaders. The former Secretary-Generals of the United Nations, Ban Ki-Moon and Kofi Annan, have both repeatedly stressed the importance of bringing women to the peace table. The United Nations Security Council Resolution 1325 on Women, Peace and Security, passed in 2000, also stresses the particular value women can bring to the negotiation table (UNSC Resolution 1325, 2000). As one response to this, several networks of women mediators have been formed around the world. These networks aim to be important tools for women to participate in peace negotiations and peace-making. The long-term goal of peace negotiations is to reach a durable peace and a first step in that direction is to reach a comprehensive peace agreement (Badran 2014), i.e. a peace agreement including an extensive set of issues. Mediators have a great influence in peace negotiations and might be able to promote the reaching of such an agreement (Bercovitch 2011). Still, despite the recognition of mediators ability to influence negotiations, and despite the increasing awareness of the importance of including women in peace-making efforts, systematic studies investigating if and how female mediators can have an impact on the outcome of negotiations is missing. The actual participation of women at decision-making level in peace negotiations also remain extremely low. This study aims to understand how women mediators can affect peace-making efforts. This will be done through an investigation of how female mediators might affect the content of a peace agreement reached through mediation. 6

7 The research question of this study is: Is the comprehensiveness of a peace agreement influenced by the participation of female mediators, and if so, how? The question is addressed by drawing on theories suggesting that mediators have an ability to influence peace negotiations through agenda setting and encouragement of discussion on certain issues (Gerami 2009). It also draws on theories on the importance of women s presence to address different kinds of issues in peace negotiations and women s ability to provide a link between the negotiations and civil society groups (Phillips 1995,1998; Anderlini 2007). The study hypothesizes that mediation teams that consists of both women and men will broaden the set of issues brought to the table, consequently leading to more comprehensive peace agreements, compared to peace agreements negotiated by a mediation team consisting of only men. Previous research on gender and mediation has mostly focused on the different styles and approaches women and men use as mediators (Herrman et al. 2003; Nelson, Zarankin & Ben- Ari 2010; Butts 2010), as well as female mediators effectiveness (Maxwell 1992; Stuhlmacher & Morrissett 2008; Moore 2003). Yet, the results of these studies do not provide a coherent answer. Furthermore, characteristics-related explanations for women s participation as mediators and arguments pointing to the need of just representation are stressed (Ellerby 2013; Porter 2003). Studies on women s contributions in civil society on grass root level and as negotiating delegates for warring parties provide evidence that women s participation can influence peace processes (see for example Sumie 2004 and Anderlini 2007), and also that women have a positive impact on peace (Stone 2004; Aggestam 2016) However, the interaction of both female and male mediators is still an area of limited research. In most previous mediation efforts where women have participated as mediators they have done so as part of a mediation team also consisting of men. Hence, it is of interest to empirically examine the concrete contribution women have brought as mediators in a mediation team. This study aims to both contribute to the understanding of women s particular contribution to mediation efforts, and specifically to the understanding of if, and how, a gender-mixed mediation team is more successful in reaching comprehensive peace agreements. 7

8 The hypothesized relationship between gender-mixed mediation teams and comprehensive peace agreements will be tested in two steps. Firstly, the study aims to investigate if there is a correlation between the gender-composition of mediation teams and the comprehensiveness of peace agreements. This is done by comparing four peace agreements that have been reached through mediation after Two of the agreements were mediated by a gendermixed mediation team; Sudan 2005 and Democratic Republic of the Congo (DRC) 2008, and two were mediated by all-male mediation teams; Liberia 2003 and Ivory Coast Secondly, the study tries to unfold the causal mechanism and investigate how the participation of women in mediation teams may affect the content of a peace agreement. This is done through an in-depth study of the mediation processes prior to the agreements met on the cases of Sudan and Liberia. Sources used is the peace agreements reached through the four mediation processes, information gathered from publications made by the participating mediators and negotiators, as well as sources of academic nature handling the two mediation processes. The evidence stemming from the analysis suggests limited support for the study s hypothesis. The content analysis of the four peace agreements does not provide a significant pattern clearly suggesting that gender-mixed mediation teams are more likely to reach comprehensive peace agreements. However, the peace agreement on Sudan, that was mediated by a gendermixed mediation team, is categorized as comprehensive. The in-depth comparative analysis of the cases of Liberia and Sudan provides evidence to assume that the mediation teams did have an impact in the forming of the agreements. Hence, there is still reason to assume that women mediators do have an impact on the comprehensiveness of peace agreements. However, even though previous research provides support for the causal mechanism, the study was, due to too big contextual differences in the cases studied, not able to draw any clear support for this. Furthermore, the analysis underlines the possible impact of the active participation of civil society women s groups, as well as the strategy used by mediators. The study will proceed as follows, firstly, previous research in the area of women in peacemaking and mediation, is presented to narrow down the research gap to which this study aims to contribute. Secondly, the theoretical framework, discussing mediators ability to influence negotiations and women s particular contribution as mediators, is presented. Following this, the study explains the design of the investigation that will be made and explains how the cases in the study are selected. Further, definitions of key concepts and a presentation of material 8

9 used for the study is made, before turning to the content analysis of four peace agreements. The study then continues to an in-depth analysis of the mediation processes leading to the peace agreements in Liberia and Sudan. Lastly, reflections on the findings and a discussion on alternative explanations and additional observations is provided before the study is concluded. 2. Previous research This section will provide insight of previous research which will provide a basis and point of departure for this study. First, it will provide an insight of previous research examining women s different roles and abilities to affect peace process in general, before continuing to previous research on the impact female mediators may have in peace negotiations. 2.1 Women in peace processes Research related to women in peace processes has previously focused on the question of why women s participation in these processes is important (see for example Anderlini 2007; O Reilly & Súilleabháin 2013; Ellerby 2013; Porter 2003). First, it is a question about equality. Since a peace agreement to a large extent determines a country s future both women and men will be affected by how and in which direction the post-conflict rebuild of the state will be structured (Ellerby 2013). Second, men and women are affected differently by conflict. For example, many women lose their husbands in war and must take care of the household and family on their own. Women s sexual integrity might also be undermined as rape is often used as a weapon of war. If women do not participate in the negotiations towards an agreement the issues women face will not be addressed and their subordination exacerbated (Porter 2003). However, women are generally excluded from higher levels of peacebuilding processes. Therefore, most of the literature examining women s role in peace processes focuses on the positive effect of women in civil society organisations at grass root level. These studies explain how women, through working in local active organisations, influence peace processes and promote the building of a stable and durable peace. For example, it is suggested that women can choose to identify themselves as mothers and in that way cut across international borders and ethnic enclaves, which will promote a more effective peace process. It is proposed that international security forces and diplomats should ally with these mobilized mothers since local women often have influence where international organisations do not (Hunt & Posa 2001). 9

10 Further, studies focusing on how gender equality affects UN peace operations show that UN peace operations are more likely to succeed in countries where women possess greater status in society (Gizelis 2009). This phenomenon is explained through the argument that the United Nations, in societies where women have a higher level of participation, can tap into greater social capital that can stimulate the peace process. By assisting women, the United Nations can target local communities and children and are able to address long term problems and concerns that may otherwise undermine peace-building efforts. In societies where women hold greater status it is easier for the United Nations to establish a durable peace since they have a more stable foundation of the society from the beginning (Gizelis 2009). When it comes to the study of women s influence on peace agreements, much focus has been to examine the implementation of UNSC Resolution 1325 (see for example Ellerby 2013 and Bell and O Rouke 2010). These studies are examining how, or if, peace agreements address women and women s issues. The results show that there is a marked rise in references to women in peace agreements reached after the adoption of UNSC Resolution However, it is discussed whether the reference to women can be ascribed to the adoption of Resolution 1325, and indirectly to the world s recognition of the importance to address women s specific need in peace processes, or if the processes examined in practice had created a more extreme situation for women, which might have made it appropriate to include references to women in the peace agreement (Bell & Rouke 2010). Ellerby (2013) compares the text of peace agreements with the different aspects of women s representation, incorporation, protection and recognition as stated in UNSC Resolution 1325, and finds that more agreements since 2005 seems to better include these aspects than in the previous 15 years, but very few cases include all forms of women s inclusion. However, the gender of the mediators assisting to reach these agreements has not been examined in any of these studies. Hence, it is interesting to examine this further and investigate whether the gender-composition of mediation teams might influence the specific topics and references included in peace agreements. 10

11 2.2 Women and mediation When it comes to the issue of mediation, there is no coherency in the literature regarding any difference between women and men in the role of mediators. One reason could be that data on outcome differences between mediation effort conducted by men or women is extremely limited (Stuhlmacher & Morrissett 2008). Yet, some studies have found variation in mediator style between the sexes. Picard (2002) examines how 90 mediators of both sexes in Canada, that practice in contexts ranging from labour-management to divorce and international trade disputes, describe their role as mediators. She finds that both male and female mediators describe themselves as facilitators, but that there is a difference of the meaning of this role between the sexes. Women emphasize the practice of facilitating both communication and process, while men tend to only speak about facilitating the process (Picard 2002). This difference is also supported by Tannen (1994) who demonstrates that there is a difference between women s and men s style of communication. She suggests that men seek to demonstrate knowledge and skill in conversations, whereas women tend to use conversation as a way of establishing connections and relationships. In a study examining 189 Israeli mediator s perception of their job, it was found that both female and male mediators had the same primary goal of helping the parties to settle dispute. Both female and male mediators were task oriented and endorsed the reaching of an agreement as important criteria for success. However, women ascribed more importance to relationship enhancement than men (Nelson, Zarankin & Ben-Ali 2010). In a study by Maxwell (1992) it is suggested that female mediators are slightly more effective than male mediators in bringing parties to a settlement, and that the settlements female mediators reach are more binding. However, contradictory results were found in a metaanalysis conducted by Stuhlmacher and Walters, 1999 (Stuhlmacher & Morrissett 2008). Still other studies find that there are no statistically different differences in negotiation outcomes and performance between men and women (Craver & Barnes 1999, cited in Moore 2003:57). The lack of coherence in research related to gender and mediation might be explained by the lack of studies of women s participation in peace-making (Stone 2004). Stone (2004) acknowledges that current literature often does not make a clear distinction of women s participation in peacebuilding at lower level and their participation in activities at higher-level of peace processes, such as officially participating in peace negotiations. Through a quantitative analysis she examines the effect of women s participation in peace-making and 11

12 the durability of peace. The study shows that increasing women s participation could increase the probability of violence ending within one year by 24, 9 %. Further, increasing women s inclusion as active participants in the peace plan makes it 25 % more likely that violence will end within five years. Yet, the results are puzzling since they show that an increase of language on women s human rights in the text of the peace agreement and if female third party participants are present, conflict is more likely to continue by 24 % and 30 % respectively (Stone 2004). However, the study is not comprised to the narrow focus of mediation, instead the dependent variable female peace-making is defined as the inclusion of women in negotiation, mediation, and peace agreement processes between conflicting parties. Hence, there is no distinction between the different roles women may have in the peace-making process. 2.3 Research gap Previous research on gender and mediation has mostly focused on the different styles and approaches women and men use as mediators (Herrman et al. 2003; Nelson, Zarakin & Ben- Ari 2010, Butts 2010). Furthermore, questions of how effective female mediators are in bringing parties to a settlement, and to reach more binding settlements have been addressed (Maxwell 1992; Stuhlmacher & Morrissett 2008; Moore 2003). Yet, the results of these studies do not provide a coherent answer. Characteristics-related explanations for why women should participate as mediators as well as justice arguments are brought forth (Ellerby 2013; Porter 2003). Moreover, studies on women s contributions in civil society on grass root level and as negotiating delegates for warring parties provide evidence that the participation of women can influence peace processes (see for example Sumie 2004 and Anderlini 2007). When examining peace negotiations where women has participated as negotiators, mediators, witnesses or signatories to peace agreements, it is evident that women have a positive impact on the duration of peace (Stone 2004). However, the interaction of both female and male mediators is still an area of limited research. In most of previous mediation efforts where women have participated as mediators they have done so as part of a mediation team also consisting of men. Hence, it is of interest to empirically examine the concrete contribution women can bring as mediators in a mediation team. Therefore, this study aims to both contribute to the understanding of women s particular contribution to mediation efforts in a mediation team, and specifically to the understanding of if and how a gender-mixed mediation team can have an impact on the reaching of a comprehensive peace agreement. 12

13 3. Theoretical framework In this section a theoretical framework will be presented. This framework will be used to develop the study s hypothesis and is built on theories on mediators ability to affect mediation outcome and theories on how women can influence peace-making. 3.1 Mediators ability to affect negotiations This study will investigate how the gender-composition of a mediation team might influence the comprehensiveness of peace agreements. To be able to do so, it must first clarify what a mediator is and how a mediator may affect negotiations. This study defines mediation as a reactive process of conflict management whereby parties seek the assistance of, or accept an offer of help from, an individual, group, state or organization to settle their conflict, or resolve their problem without resorting to physical force or invoking the authority of the law (Bercovitch & Houston 1996:13). 2 As the definition reveals, mediation efforts may be brought forth by different types of mediators, i.e. states, individuals, groups and organizations. It is also important to acknowledge that mediation efforts often are conducted by a mediation team and not one single mediator. Often, a chief mediator is assigned but he or she is usually accompanied by a team of mediators, all of who can bring great contributions to the process. This study will define a mediator as someone undertaking mediation activities, who is officially part of the mediation delegation, and is actively participating and contributing to the mediation process. A prerequisite for meaningful mediation is that the conflicting parties accept the mediator. Often, the mediator is invited by one, or all, of the warring parties. Sometimes the mediator will offer its support without an invitation. In any case, it is important to acknowledge that mediators often have an own agenda or intention with getting involved in mediation processes (Bercovitch 2011). Mediators often want to affect, change, resolve, modify or influence the conflict in some direction, and may both consciously and unconsciously bring ideas, knowledge, resources and interests of their own, of the group, or organization that they are representing (Bercovitch 2011). 2 By using this definition only mediation by non-violent means will be examined, even though some scholars state that violence might be needed to, for example, diminish spoilers. This study however, does not recognize such acts as mediation. 13

14 The goal of peace negotiations is often to reach a durable peace. A first step in that direction is to reach a comprehensive peace agreement (Badran 2014). Mediators may use different strategies and approaches to reach such an agreement. Some mediators focus on enhancing communication and foster constructive dialogue between the parties, so that the parties themselves may bring their concerns, needs and interest to the discussion. Others focus on reaching a settlement through offering their own suggestions. Mediators may also determine the order of speaking and reframe the parties statements (Gerami 2009). Mediators are often prohibited from judging the substance of an agreement, but in order to bring the parties hidden interests to the surface, mediators may evaluate the content of the agreement to make sure that all interests are represented (Gerami 2009). Several scholars describe mediators as powerful catalytic agents who in fact are able to introduce their own agendas and affect negotiations in their desired direction (see for example Brett et al 1986; Carnevale 1986; Shapiro et al. 1985; Young 1967, cited in Bercovitch 2011). Process decisions and decisions regarding the agenda are powerful tools for a mediator. The mediator may encourage discussion around certain issue, and leave other issues off the table. In this way, the mediator can steer the processes and the amount of time and resources dedicated to each topic, encouraging or discouraging the discussion of certain issues (Gerami 2009). The type of mediator may also have an impact on the mediator s ability to influence negotiations. Bercovitch (2011) argues that Head of states possess a higher chance to reach a successful outcome than individuals or non-governmental organizations (NGO:s). At the same time, individuals are argued to be able to better operate under secrecy, which may be an important aspect in some conflicts. Whether the mediator is passive, facilitative or highly active may also affect the mediator s ability to influence the mediation (Bercovitch 2011). This study emphasizes the power mediators have in negotiations, but does still acknowledge that mediation is a dynamic and interactive process. It is not just the mediator but also the parties to the conflict that may play an important role in mediation processes. The parties may be powerful both in the shaping of the process and in the eventual outcome (Gerami 2009). Situational factors, such as power balance between the parties, where the negotiations take place, the timing of the mediation, motivational factors for the parties and interactional factors such as previous history of the parties, nature of polity, nature of economy and adherence to established norms also play a part when evaluating the outcome of mediation (Bercovitch 2011). Still, a central assumption in this study is that mediators have an ability to influence 14

15 the mediation process and indirectly the content of peace agreements. Depending on who the mediator is, or how the mediation team is composed some outcomes are more likely than others. Mediators have the ability and power to form the processes, set the agenda and encourage or discourage parties to discuss certain issues. This study aims to examine whether there is a difference between employing all-male mediator teams, as has been the most common composition traditionally, or if mediation teams with women mediators may affect the outcome of mediation making the peace agreements more comprehensive. 3.2 Women as mediators As previously presented in this study, research on female mediators impact on peace negotiations is limited. A reason for this is most likely that there, unfortunately, are very few cases where women have been participating as mediators in peace negotiations. However, research on women in politics, women participating as negotiators in peace processes and women participating at grass root level in peacebuilding processes is more plenteous. This study will use findings from this research combined with the limited findings on female mediators to conduct a theoretical framework explaining how women s participation as mediators is expected to influence the comprehensiveness of peace agreements. When the United Nations Security Council unanimously adopted Resolution 1325 on Women, Peace and Security in 2000, the world s leaders recognized that women bring particular value to the negotiation table and that it is important to include women at the higher level of peacemaking (UNSC Resolution 1325). Further, prior research has highlighted reasons for including women in peace processes both in terms of equality and fair representation, as well as women s specific experiences in war (see for example Ellerby 2013; Porter 2003; Anderlini 2007). The gap between government leadership and grassroots activism is also expected to be bridged with women at the negotiation table (Anderlini 2007). Some scholars argue that fresh perspectives and approaches to peace negotiations are needed in the world today (O Reilly & Súilleabháin 2013). They see the participation of women as a tool for contributing to this and argue that women may bring different experiences to the talks. They further point to evidence suggesting that a more durable peace is reached when women participate in the peace process (O Reilly & Súilleabháin 2013). Peace negotiations 15

16 offer a widow of opportunity for being able to set an agenda for a sustainable peace and it is important that women participate in the forming of this (Ellerby 2013). Castillejo (2016) stress that women bring important issues to the peacebuilding agenda, that men seem to overlook. This follows the practice of the UN, regional organizations, donors, and civil society which are increasingly supporting conflict resolution training and informal dialogues for women, and brokering their participation in official negotiations to ensure that women and their common agenda are represented at the peace table (Sumie 2004). This concept, women s common agenda or women s interests is one of the most controversial concepts in feminist-oriented political science (Wängerud 2000). However, this concept is also commonly used to explain the importance of women s representation in politics. The theory of the politics of presence developed by Phillips (1995, 1998) argues that female politicians are necessary to create room on the political agenda for women's interests. The theory describes a distinctive group identity, based on shared common interests, that is argued to apply to all women. These issues are for example abortion, childcare or equal opportunities in education and the labour force (Phillips 1995, 1998). Phillips argues that women and men have different experiences and that women s particular needs, interests and concerns will be inadequately addressed if politics is dominated by men. Most researchers agree that gender has a certain impact on politics, but that the difference lays in how much significance that is ascribed to this factor (Wängerud 2000). Wängerud (2000) argues that women as a group have a common interest in increasing their autonomy. She sees the politicization of women's everyday life experiences as a first step in that process. In her study of members of the Swedish Parliament and gender equality policies she finds that women are better at representing women and equality issues than men, and that female politicians are more likely to be in contact with women organizations than men. Furthermore, a study of the British Parliament, where the Members of Parliament are grouped according to values, priorities, and legislative roles, found that women have more feminist and leftist values than men (Lovenduski & Norris 1996). Women also give higher priority to social policies (Wängerud 2000). As studies on women in politics suggests, empirics from prior peace processes also shows that where women have been party to official negotiations, as for example Burundi (2000), Afghanistan ( ) and the Inter-Congolese Dialogues ( ), women s groups have 16

17 proposed certain issues they wish incorporated in peace agreements. Their proposals often include: statutory guarantee of women s rights and equal treatment, quota for women in decision making process, special measures ensuring the safe return and re-integration of displaced women and women s rights to property ownership and inheritance (Sumie 2004). In Burundi, Guatemala and Northern Ireland women managed to incorporate some of these proposals into peace agreements, making each agreement more progressive and inclusive of political commitments towards women (Sumie 2004). These cases indicate that women can be able to affect the outcome of a peace negotiation by getting issues of their concern into the text of the peace agreement. Furthermore, when more women participate in peace negotiations, a broader spectrum of issues seems to be brought to the agenda (Aggestam 2016). Women s participation in peace processes do also contribute to the enlargement of the scope of peace agreements, specifically by including critical societal priorities, which carry an increased potential of securing a durable peace (Mbwadzawo & Ngwazi 2013). Moreover, more gender specific language seems to be included in peace agreements when women participate in peace negotiations (Aggestam 2016). Also, the broader discussions lay a better ground for a robust and equal peaceful development over time (Aggestam 2016). Furthermore, one of the strongest correlations that has been discovered in the field of peace and conflict research is a positive relationship between gender equality and peaceful societies (Melander 2016). Studies show that more women in the labour force, more women having higher education and women s equal rights stated in law, covariate with lower risk for civil war, torture and other threats against human rights. This relationship still holds when controlling for economic development, democracy and different types of religion (Melander 2016). This relationship is interesting since it can suggest that women biologically are more peaceful than men. Hence, when the society they live in becomes more gender equal and women gain more power the whole society also becomes more peaceful. Among warring parties and armies, we also see a high overrepresentation of men as combatants. According to social constructivist theories this is the result of power structures that have been embedded in institutions and that reinforce traditional norms through formal and informal rules in society. Since men have been dominating the area of war and negotiations for so long these structures are now favouring men, and that is why we see a lot of men, and very few women, mediating and negotiating (Beckwith 2005). However, these theories fail to explain why masculinity, and not femininity, is most associated with war in all 17

18 known cultures. If biological attributes do not differ between women and men it would be reasonable to expect some cultures constructing norms of war being associated with femininity, or that no difference between masculinity or femininity would be made in relation to war (Melander 2016). Instead, studies have found that women have a less positive approach towards war than men, and experimental studies show that when women handle conflicts they use methods that promote peace to a greater extent than men (Melander 2016). Empirics also show that women are disproportionately highly represented in groups that mobilize communities and societies to demand belligerents to lay down their weapons and negotiate a peace agreement (UN Women 2012). Furthermore, empirics show that in cases where women take part in the negotiations as delegates of parties or as gender advisors to the mediation team, their impact on the language of the agreement text and the inclusion of provisions specific to women is frequently significant (UN Women 2012). For example, the UN stresses that women s absence from the peace table is a major reason for why negotiations do not cover the issue of sexual violence (United Nations 2012). Mediators often feel that addressing sexual violence is not a priority for the negotiating parties, and mediators also underline that they have not received any clear orientation on this issue (UN Women 2012). Addressing conflict-related sexual violence and other gender-related issues in peace processes can be a matter of possessing the right expertise. UN Women withholds the argument that if women are not able to represent their concerns, no one else will do so, hence an important set of issues will remain neglected. 3.3 Causal chain This study draws on the literature previously presented and hypothesizes that where women participate in mediation teams, peace agreements will be more comprehensive compared to peace agreements met where only male mediators have been participating in the negotiations. A mediation team consist of at least two mediators, and a gender-mixed mediation team consists of at least one male mediator and one female mediator. The concept of gender is deeply complex, but for the purpose of this study gender is defined as the biological definition of being woman or man. 18

19 Figure 1: Causal chain explaining how a gender-mixed mediation team might influence the comprehensiveness of peace agreements. The causal chain for this hypothesis is illustrated in figure 1. With a gender-mixed mediation team, the aspects of both women and men are expected to be brought to the negotiation table. Women are affected by war in a different way than men, hence they are expected to bring up different issues. Further, women award higher priority to social policies and are expected to bring up provisions specific to women. Following this, and the assumption that mediators do have an ability to influence the content of a peace agreement, it seems reasonable to believe that if a team of both female and male mediators participate in peace negotiations the set of issues brought to the table will be broaden. This, in turn, will increase the chances of reaching a comprehensive peace agreement. Also, women tend to be more prone to reach out to civil society women s groups than men. This would be yet another reason for believing that including women in mediation efforts would broaden the set of issues brought to discussion. The notion of the mediators central role and possibility to affect negotiations draws the conclusion that if the mediators are successful in bringing more issues to discussion they would also be able to make the warring parties sign more comprehensive agreements dealing with these issues. 4. Research design and method The research design used in this study is a comparative case study. This design is particularly helpful when aiming to reach a deeper understanding of causal processes, and it is also preferable if the phenomenon studied is rare or difficult to observe (Johnson & Reynolds 2012). Since female mediators are a rare phenomenon and since this study aims to reach a deeper understanding about if and how female mediators can influence the content of peace agreements it is desirable to apply a comparative case study design. However, there are also critique against this design that must be taken into account. Firstly, the huge amount of detailed information that is conducted in this type of study heightens the 19

20 risk of bias in presenting evidence (Johnson & Reynolds 2012). However, this sort of bias is not limited to case studies and will be accounted for when drawing conclusions from the material gathered for this study. Secondly, some scholars have highlighted that conclusions drawn from case studies may lack in their possibility to be generalized. Since only a few cases are studied it might not be accurate to apply the findings to a broader set of cases. A case study might not represent a sample of a population and therefore the investigator s goal should not be to account for statistical generalization. However, case studies may be used to expand and generalize theoretical propositions (Johnson & Reynolds 2012), as is the aim for this study. The question of if and how female mediators can influence the scope of a peace agreement will be investigated in two steps. First, four peace agreements, of which two are mediated by a gender-mixed mediation team (Sudan 2005 and DRC 2008) and two are mediated by an allmale mediation team (Liberia 2003 and Ivory Coast 2007), will be compared with focus on the amount of provisions included in the agreement and the type of provisions they include. The study both count the provisions and analyse to what extent references to certain issues are made. The aim of this analysis is to examine if there is a relationship between the independent variable gender-mixed mediation team and the dependent variable comprehensive peace agreement. The comparison will be conducted through the method of content analysis. Content analysis is particularly helpful if aiming to systematically describe the content of a text and quantify something (Boréus & Bergström 2012). There are two branches within content analysis, qualitative and quantitative. Quantitative content analysis is used to, for example, count the mention of something in a text and suits material which does not need interpretation. Qualitative content analysis, on the other hand, aims at interpreting the text and is favourable when there is a need to, for example, take the context into account when analysing (Boréus & Bergström 2012). This study will use both a quantitative and qualitative content analysis. When references to different provisions in the agreements and the inclusion of certain provisions are counted a quantitative approach is used, whereas when analysing the extent to which the agreements refer to certain issues the qualitative approached is used. The use of a quantitative content analysis provides high reliability, as it is easy for others to replicate the work of counting references. The use of qualitative content analysis provides validity to the study, as the evidence is analysed with regard to context. However, the use of a qualitative 20

21 content analysis could lead to a biased interpretation of the evidence. To minimize this risk, the criteria for the different categories that agreements may fall under in this study s investigation was set before conducting the analysis of the agreement. Further, to be able to reveal the causal mechanism between a gender-mixed mediation team and comprehensive peace agreements the study will also include an in-depth comparative case study of the mediation efforts in Sudan and Liberia. The in-depth comparative case study will take the form of a process tracing aiming to establish the factors influencing the content of the final agreement. Careful description is the basis for process tracing as it focuses on the unfolding of events or situations over time. In that way it can contribute to the evaluation of causal claims (Collier 2011). However, a challenge with the use of process tracing is that there can be multiple paths to the same outcome (Bennet & Checkel 2015). This study will minimize the risk of drawing inaccurate conclusions by also thoroughly reflect upon alternative explanations that also can have an impact on the comprehensiveness of peace agreements. 4.1 Case selection A major limitation in the study of female mediators is the lack of cases. Even to simply examine actors of mediation, without any gender aspect, is difficult already. One reason for this is the challenge of limited information on actors actually participating in a mediation process. Often, the name of the country or organization that facilitated the negotiations are mentioned, but to find information on all the persons actually carrying out the mediation can in some cases be difficult. Often, one specific person is appointed as chief mediator, however, the mediation team supporting the chief mediator, and who could bring great contributions to the negotiations, may not be mentioned. Female mediators in peace processes are a quite rare phenomenon. However, the amount of cases one will find depends on the definition of mediator. For this study, I have thoroughly gone through different kinds of sources to try to map women who have participated in mediation efforts. For example, Margaret Anstee of the UK played an active role in facilitating negotiations for a peace settlement in the Republic of Angola in the early 1990s. As special adviser of the UN Secretary-General, New Zealander Ann Hercus conducted shuttle talks in the Republic of Cyprus in the late 1990s. Mary Robinsson, Karin Landgren, 21

22 Birgitta Anali-Holst are all recognized as making contributions to negotiations of varying kind (O Reilly & Súilleabháin 2013). Betty Bigombe mediated between the government and the rebel leader of the Lord s Resistance Army in Uganda late 1990s and early 2000 (Lamwaka 2016). Further, Garca Machel was one of three mediators in the post-election violence in Kenya 2008 (UN Women 2012). Still, it has been challenging to find cases for this study since the study also puts the prerequisite of examining cases with women mediators where the mediation processes resulted in a peace agreement. Further, when comparing cases and studying how the independent variable influence the dependent variable it is often also favourable to choose cases that are as similar as possible in all other aspects than the independent variable. This strategy is called Mills method of difference, or the most similar systems design. By choosing cases as similar as possible, except for the independent variable, one might be more confident in explaining any eventual difference in the dependent variable by the variation in the independent variable (Johnson & Reynolds 2012). With these criteria, the four cases chosen are the Accra Peace Agreement on Liberia (2003), the Comprehensive Peace Agreement 3 on Sudan (2005), the Acte d'engagement Sud Kivu on the Democratic Republic of Congo (2008), and the Ouagadougou Peace Agreement on Ivory Coast (2007). These cases are all located in Africa, their conflicts concern government incompatibility, they have had civil society women s groups working at grass root level, and all agreements were met after the adoption of UNSC Resolution 1325 in However, two of the agreements were mediated by a gender-mixed mediation team, while two were mediated by an all-male mediation team. In table 1 the cases and their characteristics are summarized. 3 This is the name of the agreement and has nothing to do with the evaluation of its comprehensiveness. 22

23 Table 1: Overview of cases specifying the different cases characteristics. That the cases are located in Africa is favourable for the study s comparison since it is assumed that cases from the same continent are more alike in the environmental context, compared to cases from two different continents. That civil society women s groups are participating in the peace process at grass root level reduces the risk of ascribing a gendermixed mediation team a certain outcome that in fact might have been influenced by women working on grass root level. However, the participation of the women s groups at grass root level may still differ from case to case and will be taken into consideration when evaluating the mediation outcome. That all peace agreements are reached after the adoption of UNSC Resolution 1325 in 2000 is argued to pose a stronger test for the study s hypothesis. Since the inclusion of women in peace-making was strongly recognized in world by this time, the over all conditions are more favourable for bringing up more issues and applying a gender perspective in peace agreements, and could have been so even if female mediators were not present. Although the four cases have many similarities, they do differ in length of mediation processes and amount of prior agreements met. This could be less favorable for the study when analyzing the impact of the independent variable. Yet, due to lack of material, this study still found these cases to be the most sufficient to include. The impact of the differing factors will also be taken into account when analyzing the cases and will reflected upon in relation the study s findings. Another challenge with choosing cases is that mediation processes today are often a multiparty efforts. This can make the evaluation of different parties impact on the mediation 23

24 effort challenging. A first step for selecting the cases was to examine the signatories to the peace agreements. However, what is important to notice is that there is a significant difference in signing an agreement and actually having an active role in the mediation effort. Therefore, the study further examined the actual role of the signatories in the mediation process. In the case of Sudan and DRC two women being part of the mediation processes and who held official status as mediators were found. In the case of Ivory Coast, no woman could be found in relation to the mediation process whereas in the case of Liberia one woman signed the peace agreement on behalf of the African Union. However, a deeper examination of the mediation efforts in Liberia did not reveal anything on her role or contribution to the mediation efforts. Hence, as this study defines a mediator as an active delegate of the mediation team and not solely a representative signing the agreement, the case of Liberia is in this study categorized as a mediation effort without official women mediators contributing to the mediation effort. Prior to analysing the content of the four peace agreements a brief conflict overview will be provided. The time period for this overview start with the countries gained independence, or as in the case of Liberia its foundation, and ends with the signing of the analysed peace agreement. This overview aims to give a contextual background and does therefore include such an extensive time period. When turning to the in-depth analysis of the mediation process in Liberia and Sudan the analysis starts at the time when the mediation team that mediated the final agreement first starts their initiative and ends with the signing of the analysed peace agreement. This is done since only the period when the mediation team facilitated the examined agreement is of interest for the study s hypothesis and causal chain. 4.2 Definition of comprehensive peace agreement The dependent variable of this study is comprehensive peace agreement. The Oxford dictionaries define comprehensive as dealing with all or nearly all elements or aspects of something (Oxford dictionaries). This study will draw on this definition and use the categorization of three different degrees of comprehensiveness of peace agreements: comprehensive, intermediate and limited. Although peace agreements are designed to address particular conflicts and are customtailored to parties demands, broad parameters that fit most conflict situations can be 24