Inclusion and Civil Resistance: Similarities and Differences

Could there be some merit in marrying together the two separate fields of civil society inclusion and nonviolent mass action? These two areas focus on very similar phenomena: how non-warring actors can influence the dynamics and outcomes of political or violent conflict, without using force or working through established and institutionalized democratic procedures (such as elections). Nevertheless, the two fields have largely been working in different silos. This is unfortunate, we think, as there are several benefits from connecting these literatures.

The two areas of research, albeit separately, have generated important insights into the shape and dynamics of civil society in conflictual contexts. We would like to point to two insights that they share. First, the two fields identify a plethora of ways through which non-warring actors seek to influence their situation. Mass action is one of several modes of civil society inclusion that exist (Paffenholz 2010, 2014). Sharp (1973) listed 198 techniques of civil resistance and, although mass-based demonstrations have dominated the attention of existing literature, there is recognition of the great variety in which civil society engagement may manifest itself. In fact, previous research on civil resistance has shown that relying on different types of techniques, being able to diligently sequence between them, innovating new techniques, and adapting to the context are all key explanations for why nonviolent movements become successful (Ackerman & Kruegler 1994; Schock 2005).

Second, the two fields have aptly demonstrated the potential and power that lies in the hands of civil society actors. Nonviolent mobilization can contribute to achieving fundamental social change, and may be a more effective strategy for challenging the status quo than relying on violent means (Ackerman & Kruegler 1994; Chenoweth & Stephan 2011; Nepstad 2011; Vinthagen 2015). Previous research has also shown that popular-based nonviolent means are more likely to lead to democracy in the longer run (Ackerman & Karatnycky 2005; Celestino & Gleditsch 2013; Teorell 2010).

Yet, there are also differences in regard to the focus adopted. First, the civil society inclusion literature has tended to study the question of impact: what is the effect of civil society inclusion on the content and outcome of peace negotiations (Nilsson 2012, 2018; Wanis-St. John & Kew 2008). This is, of course, also a central question for the field of nonviolent mass action. Still, scholars within nonviolent mass action have, in addition, taken up interest in exploring the question of why and under what conditions civil resistance occurs (Butcher & Svensson 2016; Chenoweth & Ulfelder 2017; Schaftenaar 2017), which is something that has not been a focal point of interest for inclusion scholars.

Second, there is another important difference in emphasis within the two scholarly fields. Whereas research on nonviolent mass action has tended to depart from a movement-based analysis, there is a broader recognition in the civil society inclusion literature that civil society is made up of a diverse set of actors. For example, this area of research differentiates between women’s organizations, youth groups, religious actors and non-governmental organizations (Paffenholz 2010; Pouligny 2005; Wanis-St. John & Kew 2008).

Civil Society Inclusion: Conceptual Framework

Civil society is a contested concept and can be defined in different ways. Theoretically, we may think of civil society in conflicts as the non-state, non-warring, actors occupying a space between the realms of official politics, on the one hand, and the realm of family-based communities, on the other. In line with Spurk (2010) and Belloni (2008), civil society is here seen as “separate from the state and political parties and consists of the wide range of voluntary organizations in society such as religious associations, women’s organizations, human rights groups, and trade unions” (Nilsson 2012: 246). In addition to civil society actors, we also study the involvement of political parties since these, at least in some cases, may have similar roles as the civil society, but we treat the former as separate from the civil society sphere. Recent works highlight the need to not only study the inclusion of civil society actors, but also more carefully consider the involvement of political parties in peace processes (Nilsson 2012; Ross 2019). Hence, in our analysis, we seek to capture the involvement from a wide range of actors including political parties and various civil society organizations such as women’s organizations, youth organizations, and religious actors.⁸ Occasionally, we use the term ‘non-warring actors’ as an overarching concept to capture the involvement from either civil society actors, political parties, or both.

Notably, we do not include in our definition of political parties those organizations which merely serve as the political branch of a military organization, but only those political parties that are engaging through non-violent means. We recognize that these are not always easy distinctions to be made, and may vary over time, yet, the political branch of a militant organization is often identified. It is important to point out that a preference for one or the other side does not disqualify an actor from being categorized as a civil society actor or a political party. It is not necessarily uncommon that civil society actors or political parties are aligned with one of the belligerent sides. In our empirical analysis, we seek to highlight and problematize such linkages.

Mass action is defined as “campaigns, demonstration, street action, protests, and petitions” in relation to the negotiations, using the definition from Paffenholz (2014: 77). The terminology in previous research is not consistent and alternative terms seeking to capture this set of actions include ‘civil resistance’, ‘strategic nonviolence’, ‘nonviolent action’, ‘people power’, ‘contentious politics’ and ‘nonviolent struggle’ (Schock 2003). We use the term ‘mass action’ to differentiate it from elite-based approaches, which lends itself to inclusion by a selected few. While mass action sometimes is organized by existing civil society organizations, which also invites participation from the broader population, other times it may be less organized. It is important to point out that we only focus on mass action relating to the negotiations, and not demonstrations and campaigns in general. We do not include a numerical threshold but recognize that too small actions will generally not be reported in mass media. The civil resistance literature commonly focuses on sizable instances of mass action, where civilians organize in larger gatherings and through popular campaigns with more than a thousand participants, and where they raise maximalist political demands, including regime change and secession (Chenoweth & Stephan 2011). In line with Paffenholz (2014), we use a different conceptualization of mass action and are interested in various types of campaigns, demonstrations, or protests, which may also include smaller gatherings. Notably, in terms of the stated goals of those involved, we focus on events of mass action, which concern political negotiations.

Coordination in Peace Processes

A body of scholarly knowledge exists on the dynamics of coordination in third-party intervention (Bhattarai 2016; Crocker, Hampson & Aall 2001; Fisher 2006; Kriesberg 1996; Nan & Strimling 2006; Vuković 2012), and the coordination literature is helpful in creating an understanding of how the interaction between the table (elite-based negotiation) and the street (mass action) can unfold. We draw three general insights from this literature. First, it is widely recognized in this field of research that in a situation of transformation of social conflict in which a plethora of different actors and organizations are active, coordination is a key aspect in order to understand how the situation develops. It is therefore important to scrutinize not only which actors are active, but also how they interact with each other. Previous research has shown that coordination helps to address a set of problems that can occur in a multiparty conflict. This includes using different resources to work on various aspects of a conflict, increasing the bargaining leverage of the coordinating actors, and creating new entry-points through a more diverse network. Coordination can also increase the efficiency by facilitating the division of labor, as well as enhance the legitimacy of the actors by broadening the number of participants involved. The literature also pinpoints that there are several problems associated with multiparty mediation and that failure to coordinate can create obstacles for effective peace processes. These include passing the buck of responsibility, enhancing the risk of ‘forum shopping’ between different negotiation tracks, creating misunderstandings, and increasing the risk that different actors work to undermine, rather than strengthen, each other’s’ efforts. A lack of coordination between actors can lead to mixed messaging, which in turn can increase the gaps in communication and lead to misunderstanding between the parties (Crocker, Hampson & Aall 2001). In line with these concerns, Kriesberg (1996) identifies how the multiplicity of intermediaries can create obstacles for peacemaking processes such as leading to divergent expectations, the undermining of the policies, and draining on the main parties’ attention.

A second important insight that we draw from the coordination literature is that coordination can vary. Nan & Strimling (2006: 2) define coordination broadly to include “… information sharing, collaborative analysis and strategizing, resource sharing, formal partnerships, and other means of synchronizing and/or integrating activities.” Strimling (2006) has developed a useful framework for analyzing this variation by identifying a spectrum of different forms of interactions between third-party actors, ranging from lower to higher levels of intensity. Drawing on Nan & Strimling (2006) and Strimling (2006: 94), we focus on communication, which is the least ambitious form, largely restricted to information sharing; resource sharing between actors; synchronization of various types of activities and actions;⁹ collaboration which implies that the parties jointly decide and implement common actions and entails a partnership that can be either ad-hoc or sustained over time; and lastly integration, which is the most ambitious form of interaction in which various aspects of the included actors are combined into one. Obviously, at one end of the spectrum, the units involved in this interaction cease to exist as separate entities and are fully integrated into one.

A third, relevant insight from the coordination literature is differentiating between the efforts that occur simultaneously or sequentially. According to Fisher, “coordination is directed toward increasing complementarity, either simultaneously or sequentially” (2006: 68), italics added. Thus, coordination is about different actors seeking to complement each other’s resources in order to achieve needs or goals that they may not achieve individually. The temporal scope of this coordination is important to consider: how coordination may unfold sequentially over time, with one actor’s engagement leading to another’s, or in parallel processes.

The above three insights from previous research on coordination by peacemaking actors and intermediaries serve as the framework for our analysis of the interplay between elite-based negotiation attempts, on the one hand, and mass action, on the other. Where applicable, we will also highlight some instances where coordination occurs between civil society actors primarily involved at the elite-level.

Liberia

The peace process in Liberia in 2003 saw several examples of mass-based engagement from civil society actors, as well as elite-based participation of both civil society actors and political parties. The Liberian civil war first broke out in 1989 and came to last until 1996 when a peace accord was signed. Former warlord Charles Taylor won the presidential elections of 1997. After several years of repressive rule under Taylor, civil war broke out again in 1999 when the rebel group LURD made their first incursions from neighboring Guinea. Another rebel group, MODEL, later joined the armed struggle and, in the spring of 2003, the two rebel organizations were making significant advancements on the battlefield (Nilsson 2009; Nilsson & Söderberg Kovacs 2005). Meanwhile, peacemaking efforts occurred to try to end the brutal civil war.

Peace talks between the government and the LURD rebels were initially scheduled to take place in March 2003, but these were postponed. This announcement triggered a series of mass-based actions carried out by different civil society actors, most notably the Women in Peace Building Network (WIPNET). This women’s group mobilized a diverse range of participants, and made use of channels such as the radio to summon people to the streets. In particular, they convened the women in the capital by broadcasting, “If you want peace, make it your duty to come to the Monrovia city hall at 8 A.M. Wear white” (Gbowee 2011: 135). Around a thousand women responded to this call, going into the streets to demand a cessation of hostilities and for negotiations to begin between the warring sides (Gbowee 2011: 135; Agence France- Presse 2003a). In another protest carried out by WIPNET, 100 women marched silently through Monrovia. Their banners read, “We are tired of war” and “Our children must live in peace” (All Africa 2003a). On April 17, WIPNET conducted a sit-in at the gates of the parliament demanding a halt of violence in the country (Gbowee 2011: 138; All Africa 2003a). These actions contributed to build pressure on Taylor’s government, who would offer them the chance to speak to him directly and express their demands (Gbowee 2011: 140). Additional actions carried out by WIPNET included a sex strike announced by radio. The intention was to put pressure on the men and was justified by claiming that, as long as fighting continued, no one was innocent, and that the inaction to stop it would make one guilty (Gbowee 2011: 147). The sex strike lasted for a few months.

Besides WIPNET, other non-combatant actors were mobilizing before the beginning of negotiations. The Inter-Religious Council of Liberia (IRCL) met with the 18 Liberian political parties in an event which brought consensus on calling for immediate negotiations and a ceasefire, as written in their common resolution (All Africa 2003b). In May, another form of collaboration took place as a forum of women leaders was organized by the Mano River Women Peace Network (MARWOPNET) with the objective of building consensus on the way to restore peace in Liberia (All Africa 2003c). In line with this approach, Archbishop Michael K. Francis stressed the need for civil society to seek a common agenda regarding the peace negotiations (All Africa 2003c). In this sense, there was a sequential collaboration between the different non-warring actors.

In June, negotiations between the government, LURD, and MODEL were finally taking place. Simultaneously, different mass-based actions continued to take place in relation to the talks. However, a more elite-driven form of participation was introduced as civil society groups, church representatives and the 18 political parties were present at the table together with the warring actors (Agence France Presse 2003b). In particular, organizations such as MARWOPNET, the Inter-Religious Council of Liberia, and the Liberian Bar Association were granted observer status at the negotiations (Paffenholz 2014; IPTI 2018).

As the negotiations were ongoing, WIPNET organized prayers for peace in Liberia and carried out a series of demonstrations outside of the negotiation room, chanting, “We want peace” (Gbowee 2011: 155; All Africa 2003d). Notably, representatives of the 50,000 Liberian refugees in Ghana demonstrated in front of Taylor’s hotel, expressing their tiredness of the conflict (All Africa 2003e). In another event the same month, 100 WIPNET activists, dressed in white, gathered in front of the US embassy in Monrovia and demanded their intervention to put an end to the Liberian conflict (All Africa 2003f). Furthermore, mass-based demonstrations were also carried out by civilians in support of the signing of a ceasefire (Agence France Presse 2003c), and by Liberian civil society groups attending the talks, this time protesting against the collapse of a ceasefire agreed by the belligerents (All Africa 2003g). In July, women from WIPNET blocked the entrance of the conference room where negotiations were taking place. The women refused to let the representatives out until they reached a settlement. Although an agreement was not reached that day, this action increased the pressure to move forward with the talks (BBC Monitoring Africa 2003a; 2003b; Gbowee 2011: 161–163; Nilsson 2009; Zanker 2014).

As part of the negotiations, representatives of the political parties and civil society organizations met to draft an agenda for political reform (All Africa 2003e). In particular, this elite-based participation at the talks was focused on the organization of presidential elections (All Africa 2003e), and they played a direct role in nominating candidates for a transitional government (IPTI 2018). The civil society actors also shared statements on thematic issues with the combatants, the mediator, and sometimes the press (IPTI 2018). Apart from the formal negotiations, another example of more elite-driven initiatives were the talks between the community of Sant’Egidio and LURD leaders in early August, resulting in the rebels vowing to respect a recently agreed upon ceasefire and to withdraw from Monrovia as the peacekeeping troops arrived (BBC Monitoring Africa 2003c; Agence France Presse 2003d).

Finally, a comprehensive peace agreement was signed in Accra, Ghana on 18 August. This resulted in a significant reduction of the violence in the country over the following years. The agreement included provisions demanded by civil society actors, such as the inclusion of a Ministry of Gender in the transitional government (Accra Agreement 2003; IPTI 2018).

There was a high degree of coordination between the mass-based and the elite-driven initiatives to influence the direction of the peace talks. Illustrating a form of resource sharing, WIPNET declined an offer to participate as observers at the negotiations in order to avoid the impression that they were competing with MARWOPNET or that they mistrusted the latter (Gbowee 2011: 156). This decision can be seen as a form of parallel synchronization since the two women’s groups kept separate their involvement, from within and from outside the conference room, respectively. These women’s groups agreed to share information on what MARWOPNET heard at the talks, and to collaborate on a daily press release expressing the position of women with a unified voice (Gbowee 2011: 156). Importantly, both groups aimed to address the persistent gender inequalities in Liberia (IPTI 2018), thus sharing a common agenda. In another example of information sharing between the groups, a Women’s Forum was organized at the conclusion of the talks to reflect on the negotiations (IPTI 2018).

This coordination from within and outside of the negotiating table took place despite some existing differences between WIPNET and MARWOPNET, which, however, did not prevent them from coordinating their actions in various ways. The members of the latter consisted of an educated elite of women (Moran and Pitcher 2004) which contrasted with the more diverse but mostly non-wealthy, indigenous members of WIPNET (Gbowee 2011: 144; IPTI 2018). Leymah Gbowee, coordinator of WIPNET, admitted that the relationship with MARWOPNET was not easy prior to their decision to coordinate with each other during the peace talks, and that women from MARWOPNET never attended the actions called by WIPNET despite always being invited (Gbowee 2011: 144).

In addition, it seems that the actors from civil society organizations participating at the elite level occasionally collaborated in their actions. Archbishop Michael K. Francis stressed the need for civil society to seek a common agenda even before the peace negotiations had started (All Africa 2003c). As for the political parties, they also met together with the IRCL before the peace talks to call for negotiations and a ceasefire (All Africa 2003f), showing a common stance and a close collaboration with the inter-religious organization, synchronizing their demands for an end to violence and for peace talks to begin. It is more unclear the extent to which Sant’Egidio coordinated with the other civil society actors. They held a separate meeting with the LURD rebels in Rome demanding respect for a ceasefire (BBC Monitoring Africa 2003d), and previously convened LURD and the government to halt the former’s advance on Monrovia for humanitarian reasons (Agence France Presse 2003d). In both events there is no indication that other civilian organizations or political parties participated.

In short, the Liberian case illustrates various forms of coordination between the mass and elite-based participation of civil society actors, particularly by women’s groups, as well as occasional collaboration among the participants at the elite level. At least to some degree, WIPNET and MARWOPNET shared information and synchronized their actions from within and from outside of the negotiating table, as well as collaborated on a daily press release. At the elite level, the IRCL together with political parties and other civil society organizations appear to have collaborated sequentially: they met before the launch of the peace talks to develop a common agenda that they would later advocate at the table.

The Central African Republic

The Central African Republic has seen many conflicts and coups over the years since its independence from France in 1960, until 2003 when the forces of Francois Bozizé overthrew the government of CAR and took over power. As a response to that, in 2006 a coalition of smaller rebel groups united under the name of UFDR (Union of Democratic Forces for Unity) to contest the legitimacy of the government, given their shared understanding of the political and economic marginalization under the new regime. Their biggest claims were that Bozizé carried out exclusionist policies centered only on his own political and ethnic base, while failing to share the wealth with the rest of the country (UCDP 2019).

In the case of the Central African Republic, the interplay between the civil society groups and political parties at different levels is observable through mass actions in the streets and participation of groups at the elite-level in formal negotiations. The population of CAR, having lived through many coups and many rebel groups active in different parts of the country, started taking to the streets in response to the escalation of violence between the UFDR and the government. In April 2006, almost immediately after rebel violence started in the north of the country, a women’s group started to mobilize for peace. At least 2,000 women marched in the streets of Bangui to protest the increased rebellion in the northwest of the country. Women representing civil servants, women’s associations, and women’s religious organizations marched to the Presidential Palace and handed over a memorandum to President François Bozizé, in which they “asked the authorities and the rebels to meet and negotiate for peace” (All Africa 2006).

Refusing to listen to the clamor for peace from the streets, the government sought France’s support and an increase in their military deployment in the country to fight off the rebels. However, the rebel insurgence carried on, and the protest organized by the women’s group was followed by several other protests of different nature and composition, but always mobilizing civil society around calls for an end to violence. President Bozizé started utilizing the platform of civil society discontent towards the violence to promote his own government as a victim of rebel violence. In November, rebels occupied a series of towns in the north of the country and, in response to these events, Bozizé mobilized civil society for protests demanding peace. His goal was to integrate the agenda of civil society and the government around peace and against rebel violence. However, the civil society groups present at the protests were known to have ties with the government, and government officials were also spotted marching with the crowd. Many protesters were seen carrying banners and signs that read “No to war!” Despite these efforts, Bozizé was not successful in his attempt to integrate his agenda with that of civil society, and people soon began to call for the government to take up on a rebel offer to negotiate a settlement to the conflict (Agence France Presse 2006; US Fed News 2006).

The pressure from the streets, at the mass-level, was coupled with pressure from different groups from civil society and political parties at the elite-level in a sequential manner. The protests in the streets played an important role in calling attention to the desire of the population for peace and pressuring the involved parties into negotiations. As a response to that, a Council of Elders led by Pastor Isaac Zokoe, started meeting with the rebels to arrange negotiations. The Council, which consisted of 30 community leaders and human rights campaigners, was formed in 2003 to “negotiate disputes between government and dissidents.” They were working on gaining trust from the rebels to facilitate negotiations with the government, signaling to President Bozizé their intention to seek a peaceful resolution to the conflict (Associated Press Newswires 2007a). Following Bozizé’s rejection of the rebel group’s proposal for a national dialogue, and charged by the public mobilization in the streets, the opposition political parties started meeting to discuss how they could design a new proposal and pressure the government for a national dialogue towards peace (Voice of America 2006). Soon enough, after the spread-out mobilization of civil society and political parties across different levels, the government organized a conference on peace and security and invited civil society actors – labor unions, NGOs, and religious organizations. Bozizé’s intention was to be in charge of the conversation and frame the conflict as a regional problem coming from Sudan and Chad instead of addressing the domestic grievances from the rebels. The political parties played an important role in forming a strong and democratic opposition to pressure the government into holding talks with the insurgents (Feniou 2006).

Negotiations eventually started in 2007 under much secrecy, and three peace agreements were reached by December 2008, in which the parties agreed to provisions on ending the hostilities, military power-sharing, DDR, the rebel group’s transition into a political party, a government of national unity, and new elections (Associated Press Newswires 2007b; Reuters 2007). In 2008, during the national dialogue mediated by the Center for Humanitarian Dialogue, the government and the UFDR were joined not only by political parties and civil society representatives from unions, religious groups, and other types of organizations, but also by the other rebel groups active in other parts of the country (BBC Monitoring 2008; International Crisis Group 2008; Agence France Presse 2008). The aim of the national dialogue was to achieve a great level of inclusivity and to be able to discuss and agree on a comprehensive peace deal between all levels of society, something never seen before in the country (Xinhua News Agency 2008). Even though many organizations had ties to the government and initially Bozizé saw this as an opportunity to dictate the whole process, the political parties played a significant role in steering the talks and especially making sure that they would result in formal agreements of a binding nature, while the presidential groups were pressuring for an unbinding dialogue (Sguaitamatti 2008).

After signing the peace agreement in 2008 and maintaining his seat in the presidency in the 2010 elections, many groups started contesting the legitimacy of the elections. Discontent was exacerbated by the lack of implementation of the provisions agreed upon in the peace negotiations. In that context, other rebel groups arose to overthrow the government. In 2012, remnants of the UFDR renamed themselves Séléka, and launched a major offensive against the government (UCDP 2019). They stormed the country and rapidly took over four regional cities, and soon closed in on the capital. Given the rapid escalation of violence, people took to the streets to protest the rebel offensive and denounce violence.

Once again, civil society mobilized rapidly in pressuring both the government and the rebels for peace initiatives to begin. Two main protests gathered politicians from the opposition and students from the National Association of the Central African Studies (ANECA) in the capital city of Bangui to denounce violence and demand peace (BBC Monitoring Africa 2012). Other protests, this time by government supporters, marched through Bangui from the US embassy chanting for peace. However, things took a violent turn once frustrated demonstrators arrived at the French embassy to ask the former colonial power to help to stop the rebel advance. During this incident windows were broken and the French flag was taken down (Agence France Presse 2012a; Reuters 2012). Women’s organizations joined in on the attempt by civil society to mobilize support for peace through mass action by organizing another protest in December 2012. About 300 women marched through the streets of Bangui to urge the rebels from Séléka to stop fighting. The women carried banners and chanted “we want peace in CAR” as they handed a petition to Prime Minister Faustin Archange Touadera (Agence France Presse 2012b).

Following the mobilization of civil society at the mass-level, the opposition political parties quickly organized themselves at the elite-level. They used the political leverage created by the protests to align their agenda towards peace, but they also mobilized themselves around the rebels’ demands for Bozizé to step down (Al Jazeera 2013; Reuters 2013). During the negotiations, the political parties were responsible for pressuring both warring parties into an agreement on political changes in the government as well as an end to hostilities. They were also responsible for naming the new Prime Minister that would take over the new government (Agence France Presse 2013). The parties signed an agreement in January 2013 under the auspices of the ECOWAS in Gabon, where they agreed to end hostilities and promote a series of political reforms. They also agreed on provisions concerning security guarantees, reconciliation, political prisoners, DDR, and a commitment to use peaceful means (Peacemaker 2019).

The civil society and political party involvement in the peace process in CAR is a vibrant example of a sequential coordination of the actors at different levels. At first, civil society mobilizes at the mass-based level, rallies attention, and calls the parties into action by setting the tone of their demands around peace negotiations. The government tries to capitalize on the commotion generated by civil society by attempting to integrate both movements. It does so by mobilizing its own protests together with civil society organizations known to hold ties with the government. However, the government fails to capitalize on these efforts. Once the protesters started demanding that President Bozizé accept peace negotiations with the rebels, they made him accountable for the situation, portraying him as responsible for peace, instead of a victim. On the other hand, the political parties had more success in coordinating their actions at the elite-level with those by civil society at the mass-level. They used the commotion and the momentum created by the protests in the streets to increase their political leverage and pressure the government into peace negotiations as well as substantial governmental reforms.

Burundi

The Burundian peace process leading up to the Arusha Accords saw a plethora of civil society actors taking action both for and against the peace process. A wide array of such actors, both domestic and international, involved themselves in the process. These included religious actors, youth organizations, women’s groups, as well as political parties. Actions taken by these actors were not limited to partaking in official negotiations, but also hosting workshops and parallel forums, consulting with the mediators or the warring parties, and taking to the streets to demonstrate. Although generally considered an example of inclusive negotiations, the civil society involvement was largely limited to elite-level actors, with few examples of mass-based mobilization or avenues through which outside actors could affect the negotiations.

The Burundian civil war fought between 1994 and 2005 pitched several Hutu-oriented rebel groups against the Tutsi-dominated government. In the run up to the conflict, Burundi had embarked on a democratization process. After the assassination of the Hutu president conducted by the Tutsi-dominated army, the country descended into chaos with three rebel groups challenging the government for control of the country: Palipehutu-FNL, Frolina, and CNDD (later also CNDD-FDD).

The years preceding official negotiations were marked with high levels of violence during which domestic non-combatant parties held little influence over the warring actors. International civil society actors however managed to set up initial contacts between the government and CNDD. The initiative led to a series of meetings under the auspices of the Sant’Egidio Community in Rome between September 1996 and April 1997, which paved the way for the later, more inclusive process (Barltrop 2008; International Alert 2012b).

A more comprehensive series of negotiations, later to be known as the Arusha talks, were launched in 1998 and included both CNDD and Frolina. In the early rounds of negotiations, participation of non-warring parties was most notable in the presence of representatives from 17 political parties (International Alert 2012b). These political parties made up the bulk of the 126 participants that partook in the first round of talks in 1998, whereas civil society actors were initially not included in the negotiations (International Alert 2012a). The inclusion of these political parties – who later became signatories of the final accord – is an example of elite-level inclusion as it gave a small number of non-warring actors extensive opportunities to influence the negotiations, while the peace process simultaneously remained highly exclusive to the average Burundian. It is worth noting that these parties were often also ethnic based, with some having explicit or implicit ties to one of the warring sides.

In addition to political parties, religious actors and women’s groups would later come to join the negotiations. Both international and local religious actors partook in the talks. Matteo Zuppi of the Italian Community of Sant’Egidio participated in the formal negotiations after being named the head of one of the negotiating committees in October 1998 (Xinhua News Agency 1998). Most actors however were limited to less formal channels of influence. Local religious actors, while sporadically included in the formal negotiations, held two separate consultations with lead mediator Nelson Mandela. In one of these consultations, religious actors were part of a ‘civil society delegation’, which also brought together youth organizations and women’s groups in a collaborative action to influence the process away from the negotiation table. These consultative meetings also offered a larger segment of civil society the opportunity to voice their opinions compared to the formal inclusion that was mainly offered political parties (BBC Monitoring Africa 2000c; BBC Monitoring Africa 2000d). Hence, in this case the collaborative action occurred between civil society actors which mainly engaged in elite-based approaches, rather than between the elite and mass-based level.

Women’s formal participation increased as the Arusha process progressed. Initially, a few women were offered temporary observer status after women demanded female representation in mid-1998 (International Alert 2012; Xinhua News Agency 1998). Additional women were included in the negotiating parties’ delegations, but these female participants did not advocate women’s rights or issues specifically, as they followed the agenda of their respective political parties (Daley 2007). Women’s formal participation would eventually be expanded and made permanent in January 2000, when seven women were given observer roles at the table (Anderson 2016). A substantial opportunity to influence the negotiations came towards the end of the Arusha process when an All-Party Burundi’s Women’s Peace Conference was initiated parallel to the ongoing negotiations. The parallel conference brought together 50 women of various backgrounds, including the political parties and armed movements, and allowed them to formulate proposals for integrating a gender dimension in the final peace accord, again expanding the segment of civil society that could affect the negotiations (International Alert 2012a; 2012b). Despite the conference opening up the peace talks to more women, it still fell short of providing a channel for the great mass of Burundian women to affect the Arusha talks. Many of these writings were eventually included in the final document signed in August 2002 (Daley 2007).

A final way in which Burundian civil society actors tried to influence the negotiations were the protests hosted in Bujumbura by a small group of civil society actors, headed by a splinter group of the UPRONA party under the leadership of Charles Mukasi. The protests were dismissive of the ongoing peace process and opposed to the government negotiating with some of the warring factions. At least three of these protests were held in the year 2000 (Associated Press Newswires 2000; BBC Monitoring Africa 2000a, 2000b). Although the number of individuals involved in these protests was limited, they constitute an example of collaboration between different types of civil society actors, and the most clearly mass-based civil society events recorded during the peace process.

The Arusha process thus offers several examples of how different civil society actors mobilized to influence the negotiations. At the formal talks, political parties collaborated by forming coalitions roughly following the ethnic line that permeated the conflict (BBC Monitoring Africa 1999). The consultative meetings in which various branches of civil society participated can be seen as a coordinated effort trying to influence the negotiations. Women’s groups, in particular, utilized several different channels to shape the peace process (Inclusive Peace 2018). The sequence of these actions illustrates an increasingly strong position from where women’s groups could seek to influence the peace process: from meeting separately with the mediator to observing the talks and developing independent suggestions in parallel forums. The All-Party Burundi’s Women’s Peace Conference displays an integrated level of coordinated action, as women coordinated under a body separate from their previous alignments (BBC Monitoring Africa 2000c; Anderson 2016; International Alert 2012a). Still, the coordination of women’s groups was seemingly only present within the elite level of civil society, with no independent mass action having been recorded. The bellicose mass actions in Bujumbura however contradicted the ambitions of other civil society actors. It showcased both the diversity of opinions that existed in the Burundian society at the time, and a possible disunity between the actors involved on the elite and the mass level. Hence, to the extent that coordination occurred during the Burundian peace process, this was largely taking place at the elite-level, and did not involve mass- based actions.