Downloads: 2 available

Available in


Mediating Djibouti

Download Article
Meredith Preston McGhie describes the UN mediation in the 2008 peace talks in Djibouti between the Transitional Federal Government and the Alliance for the Re-liberation of Somalia.

Meredith Preston McGhie describes the tactics employed by the UN as mediators in the 2008 peace talks in Djibouti between the Transitional Federal Government (TFG) and the Alliance for the Re-liberation of Somalia (ARS). She concludes that the most significant lessons from Djibouti are the importance of flexibility, responsiveness and the ability to react to both the needs of the parties and the political situation. 

It is especially important in this longer political process to support mechanisms for continued political dialogue in order to cement the steps that were made in Djibouti.

Meredith Preston McGhie


The Djibouti mediation process

The conclusion of the Mbagathi peace process in 2004 ushered in a Transitional Federal Government (TFG) in Somalia and a renewed hope that through a political arrangement in the country would be on the path toward stability. This hope was short lived.

In 2006 opposition to the TFG in the form of the Islamic Courts Union (ICU) took control of Mogadishu. Within six months, on invitation by the TFG, the Ethiopian military moved in and pushed the ICU out. Further destabilisation ensued. Attempts at national reconciliation failed to draw in any opponents of the TFG and little progress was made.
As the situation on the ground worsened, so did the political deadlock. A coalition of disaffected MPs, Somali diaspora and the ICU leadership came together under the umbrella of the Alliance for the Re-liberation of Somalia (ARS) and established themselves in Asmara. Fighting escalated in Mogadishu between the ARS and the Ethiopian backed TFG (which was also protected by an African Union peacekeeping force from May 2007), and the security situation continued to deteriorate.
Amid increasing calls for a political solution to the crisis a number of external actors began to make discreet moves to try to build a constituency for political dialogue between these two groups. This developed formally into the Djibouti mediation process that lasted, in its official form, from June 2008 to February 2009, and resulted in a new TFG that brought together the ARS and former TFG within one transitional government.
This paper looks briefly, from the perspective of a participant in the talks, at the process and structure of the Djibouti peace negotiations between the TFG and the ARS and led by the UN Special Representative of the UN Secretary-General (SRSG), Ambassador Ould Abdallah. The paper reviews the tactics employed in running the talks in Djibouti, including how certain political events impacted on those tactics and how these were managed. 

Beginning the Djibouti process

While the ongoing insecurity and lack of political progress in Somalia in early 2008 made it clear that dialogue was required, bringing the two parties together proved challenging.

The first months of 2008 represented a pre-mediation phase, with quiet shuttle diplomacy and outreach to the ARS leadership in Asmara through various channels in order to secure agreement, in principle, to dialogue. There was also behind the scenes work by the SRSG and other key actors to secure a similar commitment from the TFG. There was a common understanding that discreet steps needed to be taken, as the situation was not yet ripe for a larger and more open process.
The official Djibouti process began in earnest in June 2008 with the convening of a meeting between the TFG, under the leadership of Prime Minister Nur Adde, and the ARS, led by Sheikh Sharif Sheikh Ahmed. In the early stages of the talks confidence was low, with the parties not yet ready to meet directly. The SRSG again used shuttle diplomacy to secure agreement on a basic agenda. When they were first brought to Djibouti the parties refused to stay in the same hotel or to meet directly.
Choosing to meet in Djibouti was a strategic decision, which highlights an interesting dilemma in mediation: using actors and locations which may not afford complete neutrality, but conversely can apply positive pressure on the process when engaged constructively.
The process needed to take place within the region and consideration was given to Nairobi. However Nairobi’s various distractions made this unfeasible for the kind of talks that were needed, and other locations in the region were also viewed as unworkable. There needed to be a sense of safe space and a retreat-like atmosphere, particularly one in which it would be possible to control the media and other actors who would be attracted to the process.
Djibouti was an important player, not only as the host, but also as a key supporter of the talks. Pressure on the parties by President Ismail Oumar Guelleh at key junctures helped advance the process. Djibouti facilitated the convening of the Somali Parliament, offering the use of the Djiboutian Parliamentary premises as well as the provision of security and accommodation. This was of particular importance given the time pressures that developed during the final phases of the process.

Basic principles

The June 2008 Djibouti Agreement (formally signed on 18 August) put in place basic principles between the parties, including the structures that would then facilitate advancing the process. These included agreement on a cessation of hostilities (although this had limited immediate impact on the ground) and the withdrawal of Ethiopian forces from Somalia. The agreement also established a High Level Political Committee and a Joint Security Committee that would to take the talks forward.

The agreement on the withdrawal of Ethiopian forces was important during this early period. Although this took time to implement (the forces did not fully withdraw until early 2009), it was a significant step in building confidence and keeping the ARS engaged in the process. That said, the slow nature of the withdrawal was also a challenge to the dynamics of the process.
This first agreement was used by Ambassador Ould Abdullah and the parties as a platform for the subsequent agreements on security and political cooperation reached in October and November 2008 and January 2009. Although it did not provide much detail, this early deal was deliberately used as a broad platform to move forward other agreements in the coming months.

Structuring Djibouti

The fundamental approach of Djibouti may not have appeared structured from the outside. In fact, as is often the case, much of the process was dictated by events and responses to them, with structure being sacrificed as a result. Flexibility was therefore central, yet with clear objectives and direction. While this flexibility was deliberate, the process was forced to become even more responsive to events in its final stages. Arguably the overall sustainability of the agreements has been more of a challenge as a result.

While there was no formalised strategy for the mediation, Ambassador Ould Abdullah had a vision of where he wanted to direct the talks. From a tactical perspective, four key principles threaded through the process and dictated how it played out: 1) the constructive use of deadlines to push the process along (although control over the deadlines, crucially, was lost in January 2009); 2) building mechanisms that would allow Somali leadership and ownership; 3) managing regional and international actors; and 4) ensuring flexibility to respond and adapt to the changing situation.
The evolution of the structure of Djibouti also developed as a response to the needs of the parties and therefore relied less on more standardised tools or mechanisms for structuring peace talks.
During early sessions at Djibouti the parties agreed to the establishment of two committees. These broke discussions down into political and security elements, under the High Level Political Committee and the Joint Security Committee, respectively. These committees were made up of 15 representatives from each of the two parties, each having their own chair.
The UN Political Office for Somalia (UNPOS) was the official chair of the meetings, but much of the time these sessions were managed by the two parties themselves, often without external actors present. The committees had minimal terms of reference, which allowed them to be used and adapted as required. The committees were also used for capacity building through brief workshops during each Djibouti session on different substantive issues, and they allowed space for the parties to become more comfortable in discussions.
The ability of these committees to build capacity for mediation of the delegations highlights a broader challenge in peace processes. While there is often a need for more guidance on mediation techniques, parties may resist this. Or the pace of events may leave insufficient time for this to take place – the priority will be addressing issues of substance that will move the process along.
In the Djibouti case, therefore, brief sessions on substantive issues such as transitional justice exposed the parties to international experts in these areas (including UN Special Rapporteurs and members of the UN Mediation Standby Team of Experts, for example), while also building confidence between the two delegations.

Generating momentum

Previous mediation processes in Somalia before the Djibouti talks had been drawn out affairs, with months spent in session. In contrast Djibouti moved quickly – often surprisingly so.

In order to generate momentum and avoid stagnation, the Djibouti process was initially arranged so that the parties would meet for a limited period once a month to advance the agenda. The remaining time was spent addressing political issues, building consensus and preparing the ground for the next round of talks. This gave the SRSG better control over the speed of the process. It allowed him to bring the parties together at a pace that enabled both consultation and time for agreements to percolate internally within the constituencies of the two parties, while still ensuring that momentum was not lost.
This approach meant that the parties met for two to three days each month, at which point agreements were cemented. However much of the groundwork was being done during the intervening periods, during which time ownership of the process by those not directly involved was also being developed.

Managing Djibouti

A critical element in Djibouti was the need to maintain the pace and momentum of the talks. It was necessary to demonstrate regular progress in order to convince detractors of the process of the benefits of ‘coming on board’. This required both successes in the monthly rounds of talks as well as sustained external pressure to highlight the negative consequences of hindering them.

In practice, this meant carefully managing the statements and communiqués that resulted from each round of talks so that they would demonstrate success, but not giving too much out publicly before the parties had consolidated their own constituencies’ support behind agreements.
For example, following agreement in principle in November 2008 to political cooperation, the delegations of the two parties needed to return to their constituencies and talk through what this would mean and to get their support for the details that would then be agreed in the next round of discussions.
However the ability of the mediator to control the pace of the talks changed with the political developments at the end of 2008, and especially the unexpected resignation of President Abdullahi Yusuf on 29 December. This set in motion a new set of deadlines under the Transitional Federal Charter whereby a new President needed to be elected within 30 days.
Rather than managing and controlling the deadlines, therefore, the mediation now had to compress the political cooperation dialogue into this revised and curtailed timeframe, and to contend with the other new political realities that the President’s resignation brought with it.
Although this was a difficult shift for the mediation, the momentum it produced – for both the parties and the regional and international players – made it possible to push this aspect of the talks to a rapid conclusion. This was by no means uncontroversial, with considerable debate among external and Somali actors around the broader impact of speeding up the process.
Bringing the existing TFG MPs and the prospective ARS MPs to Djibouti was an example of the importance of momentum in peace talks. Convincing MPs to come to Djibouti indicated that the balance of support among Somalis was shifting towards the Djibouti process. Equally importantly, momentum was generated by pressure from Somali actors in Djibouti, as well as through pressure from regional and international actors.
It is important to note, however, that there were political interests at play. It was these rather than a pure commitment to reconciliation, or the need for further dialogue, which helped make progress. The shift can be interpreted as a recognition that the political balance had changed, and different actors calculated that their political interests lay in being part of the process rather than remaining outside it.

External actors

The influence of regional actors – negative or positive – cannot be overestimated in the Somali context. In October and November 2008, as discussions in Djibouti became more focused on a ceasefire and an agreement on political cooperation, coordinated pressure from the international and regional community proved vital.

UN Security Council Resolutions, in particular resolution 1844 of 20 November 2008, in concert with strong declarations from the Inter-Governmental Authority on Development (IGAD) and the AU, asserted a zero tolerance policy for people obstructing the peace process. This helped to maintain pressure on the momentum of the talks. While this was an important point of leverage in the Djibouti process, sanctions – targeted or more general – are still a challenging issue for regional organisations to grapple with. Questions remain about whether this limits the role of these institutions as mediators once sanctions are applied.
A huge amount of the time and energy of Ambassador Ould Abdullah and his advisors was taken up with regional travel, to ensure that key external actors were on board and regularly briefed, and that they understood the unspoken messages that were coming out of the talks. This was a delicate task that often only core mediation team members were able to undertake.

Taking the Djibouti process forward

The first phase of the Djibouti talks has resulted in agreement on the formation of a newly formed Transitional Federal Government, established in early 2009. This has included the expansion of the Parliament from 275 to 550 members, to bring in ARS MPs and an expanded Cabinet.

What was not agreed in Djibouti at this time, however, was how this new coalition would actually function in terms of its day to day operations. Many power sharing agreements end at the water’s edge of the ‘big picture’ – the drawing together of two opposing groups. Perhaps the most challenging piece of such a power sharing arrangement is how these groups manage their day-to-day affairs – how to ensure the parties remain in consultation, not only on large scale decisions, but also on other aspects of governance, including key appointments such as ambassadors.
The Djibouti process was one step in the broader process of political settlement and peacemaking in Somalia, which entered a new phase with the formal conclusion of the Djibouti process in February 2009. The subsequent period has been exceptionally difficult for the continuation of political dialogue – both within the TFG to cement the gains that have been made, and to draw others into a constructive dialogue and out of a cycle of violence.
Balancing a continued structure for dialogue with the flexibility to allow the new government to address the continually changing circumstances will be critical to embedding the successes of Djibouti.

Looking back on Djibouti, one year on

Mediation processes are the ‘art of the possible’. Many of the decisions taken in a given political setting take on a different hue with the benefit of hindsight. Looking back on Djibouti almost one year later, some insights emerge.

Perhaps the most significant lessons from Djibouti are the importance of flexibility, responsiveness and the ability to react to both the needs of the parties and the political situation. This was a critical feature of the process and is one of the reasons for its success in securing the political agreement. However it remains to be seen to what extent Djibouti has succeeded in terms of cementing structures for the parties to manage the next more challenging period of reconciliation in Somalia.
The question that has subsequently emerged is how to continue the momentum of those months. It can be argued that Djibouti moved too fast. But what it offered was a new mechanism for organising consultations and developing political cooperation through the monthly sessions.
In any mediation process, when formally structured mediation ends it is always a challenge to identify other clear channels of focus. In the post-settlement implementation period, many priorities emerge and the parties’ attention therefore risks becoming more diffuse.
For Somalia, it is especially important in this longer political process to support mechanisms for continued political dialogue in order to cement the steps that were made in Djibouti.