icon caret Arrow Down Arrow Left Arrow Right Arrow Up Line Camera icon set icon set Ellipsis icon set Facebook Favorite Globe Hamburger List Mail Map Marker Map Microphone Minus PDF Play Print RSS Search Share Trash Crisiswatch Alerts and Trends Box - 1080/761 Copy Twitter Video Camera  copyview Whatsapp Youtube
Mali’s Algiers Peace Agreement, Five Years On: An Uneasy Calm
Mali’s Algiers Peace Agreement, Five Years On: An Uneasy Calm
Burundi's President Pierre Nkurunziza embraces his South African counterpart Jacob Zuma as he departs after an Africa Union-sponsored dialogue on 27 February 2016. REUTERS/Evrard Ngendakumana
Briefing 122 / Africa

The African Union and the Burundi Crisis: Ambition versus Reality

To reverse Burundi’s slide toward a devastating social and humanitarian emergency – as ethnically-charged rhetoric worsens and refugees flee to neighbouring countries – the African Union needs to overcome its internal divisions, fix a so far incoherent response and facilitate a negotiated settlement between the government and the opposition.

  • Share
  • Save
  • Print
  • Download PDF Full Report

I. Overview

African leaders’ January 2016 decision not to endorse deployment of a 5,000-strong African Prevention and Protection Mission in Burundi (MAPROBU), as the African Union (AU) Peace and Security Council (PSC) recommended, revealed a wide rift between member states and the AU Commission (AUC) on how to address the crisis. The dispute seriously damaged AU credibility and showed that its ambition to prevent and resolve conflict does not match its capabilities, in part due to uncertainty about the extent of the AUC’s role. It also exposed procedural flaws in the PSC’s decision-making process. The incoherent response illustrates the limits of AUC and PSC freedom to act without the full support of leaders and the lack of coordination between Addis Ababa and the African UN Security Council members (the A3). If the AU is to fulfil its aim of “silencing the guns” on the continent, the AUC and member states must resolve these issues.

Map of Burundi. CRISIS GROUP

The crisis was sparked by President Pierre Nkurunziza’s 2015 decision to seek a third term, which triggered mass protests, an attempted coup, armed opposition attacks and a brutal crackdown that has fuelled a cycle of violence in which as many as 1,115 have already died, according to the Armed Conflict Location and Event Data project. The confrontation has settled into low-intensity warfare characterised by targeted assassinations, disappearances and torture and the government’s increasing resort to ethnically-charged rhetoric reminiscent of that preceding the mass atrocities of the 1990s. More than 300,000 have fled to neighbouring countries, and a further 108,000 are estimated to be internally displaced. An estimated 4.6 million of the eleven million population need food aid. With both urban and rural economies slowing and imposition of an austerity budget (an 18 per cent decrease on 2015), Burundi is sliding toward a devastating social and humanitarian emergency.

The AU and international partners have failed to halt the crisis. Nkurunziza has exploited divisions within and between the AU, the UN and the East African Community (EAC), the sub-regional organisation charged to lead the continental response. This enables his government to rebuff lacklustre EAC attempts to bring it to negotiations and has stalled the deployment of AUC-authorised human rights and military observers, as well as UN Security Council-sanctioned police. Lack of a shared analysis of the crisis’s nature fuels disunity.

The AU itself is divided. A majority of member states favour a less confrontational approach than the interventionist-inclined AUC, which together with the PSC, to avoid further embarrassment, has now shifted focus from Burundi, silencing needed warning voices. The crisis is political at its core, and only a negotiated settlement between government and opposition can end it. That requires re-examination of the 2000 Arusha accord, the power-sharing peace agreement of which the AU is a guarantor that ended the twelve-year civil war. Unless the AU, EAC and wider international community act in concert, it is a distant prospect.

To engage the government and opposition in a genuine and inclusive dialogue:

  • The AU, EAC and UN should immediately form a contact group to align positions and inject new impetus into the EAC-led mediation, whose leader, President Yoweri Museveni (Uganda), must become more personally engaged, as requested by the facilitator, Benjamin Mkapa.
     
  • The AU should expedite deployment of the 200 authorised human rights and military observers and work with the UN to quickly agree on and disburse the financial, technical, logistical and other assistance needed for the deployment.
     
  • The AU should immediately implement the decision to impose targeted sanctions, per the 17 October 2015 PSC communiqué, against those blocking negotiations, inciting violence and/or propagating hate speech.
     
  • The AU and European Union (EU) should change financing for the African Union Mission in Somalia (AMISOM) to pay Burundian soldiers directly, rather than permitting the government to keep a portion.

To strengthen AU crisis response capabilities:

  • Member states should meet their financial obligations so the AUC can be staffed adequately to carry out its mandate.
     
  • The PSC should engage consistently in finding political solutions, meeting monthly on Burundi, for example, to evaluate the security situation and discuss observer reports. Member states should engage fully on communiqués and ensure they have ownership of decisions. This requires adequate staffing of Addis Ababa embassies and clear communication channels with them.
     
  • AUC, PSC and A3 should work more closely together – including PSC observer status for the A3 – to ensure common analysis and more coherent responses.

II. The AU in Principle and Practice

A. Member States’ Commitment

Since its 2002 inception, the AU has been increasingly active in preventing and resolving conflict in Africa, due in part to growing recognition the UN cannot manage crises alone and in part to the AU’s desire to lead.[fn]The 2015 UN High-level Independent Panel on Peace Operations report highlights the AU, urging intensified collaboration also with its sub-regional partners.Hide Footnote  The organisation’s Constitutive Act set promotion of peace, security and stability on the continent as a founding objective. To help meet it, the African Peace and Security Architecture (APSA) was created. It asserts AU primacy in peace and security on the continent and devolves jointly to the PSC and AUC chairperson power to “undertake peace-making and peace-building functions to resolve conflicts”.[fn]APSA consists of the PSC, Panel of the Wise, Continental Early Warning System, Africa Standby Force and Africa Peace Fund. Articles 7, 17 of the Protocol Relating to the Establishment of the PSC, and the AU and Regional Economic Communities/regional mechanisms Memorandum of Understanding (MoU).Hide Footnote

That ambition to “silence the guns” often outstrips action, however, is not surprising given resource constraints. The Peace and Security Department (PSD), charged with administering the fifteen-member PSC, the Continental Early Warning System (CEWS), the Panel of the Wise and other mediation activity, as well as four peace support operations, has just 64 regular staff positions, 26 of which were filled in 2015.[fn]The peace support operations are AMISOM, the Regional Cooperation Initiative for the Elimination of the Lord’s Resistance Army, the hybrid UN-AU Mission in Darfur (UNAMID) and the Multinational Joint Taskforce for the fight against Boko Haram. “APSA 2014 Assessment Study: Final Report”, AU, 27 April 2015, p. 14.Hide Footnote  Personnel constraints are due to perennial underfunding. Only $169,833,340 of its estimated $416.9 million 2016 budget derives from member-state contributions, the rest from foreign partners, a reliance that compromises AU ownership of its agenda and reflects member states’ lack of commitment.[fn]Decision no.: Assembly/AU/Dec. 577(XXV). The budget does not include peace support operations. Member-state contributions mostly fund operational costs; 92 per cent of the AU’s 2016 program activities are expected to be funded by donors. “APSA 2014”, op. cit., p. 14.Hide Footnote  At the end of 2015, they had paid just 68 per cent of assessed contributions; only nineteen fully met their obligations. In July 2015, leaders pledged to fund 25 per cent of AU peace and security operations. A year later they adopted a mechanism that in theory should generate $400 million per year to help pay for AU missions, but it remains to be seen if it will be honoured.[fn]“2017 Budget Overview Paper”, AUC, 5 March 2016. “Declaration on self-reliance”, Assembly/AU/Decl.5(XXV), 15 July 2015. “The African Union Adopts the AU Peace Fund”, AUC press release, 18 July 2016.Hide Footnote

The uninspiring choices to replace AUC Chairperson Nkosazana Dlamini-Zuma is but one sign members may not actually want a strong AU. Their reluctance to cede the chairperson and PSC influence, despite vesting significant formal authority in them at the AU’s founding, is consistent with the Constitutive Act’s emphasis on national sovereignty but severely limits conflict response capacity.[fn]Some member states were dissatisfied with the quality of choices, so no candidate was able to secure the required two-thirds majority. Article 4 sets out the AU’s guiding principles, including 4(g) “non-interference by any Member State in the internal affairs of another”.Hide Footnote

B. The AU’s Principles in the Burundi Context

Beyond its commitment to preventing and resolving conflict, the AU also aspires to promote constitutional democracy and stop mass atrocities. The Burundi crisis challenges both principles, exposing inconsistency in interpreting and enforcing the Constitutive Act. Prohibition of unconstitutional changes of government (Article 4(p)) has origins in the 1990s’ broad shift from one-party and military rule to multi-party democracy. It is the only founding principle backed by a specific penalty: suspension from participation in AU activities.[fn]Articles 4(h) and 4(p) of the Constitutive Act respectively refer to prevention of mass atrocities and prohibition of unconstitutional changes of government. Mulugeta Gebrehiwot and Alex de Waal, “African Politics, African Peace”, The World Peace Foundation, July 2016. Solomon Dersso, “Unconstitutional Changes of Government and Unconstitutional Practices in Africa”, The World Peace Foundation, June 2016.Hide Footnote  The AU usually has condemned and, if asked, provided military support, when faced with coups, as in Mali and Guinea Bissau (2012) and Central African Republic and Egypt (2013).

However, what is unconstitutional change of government is not well defined. It includes an incumbent’s refusal to relinquish power after a free and fair election, but it is not clear if that extends to amendment or manipulation of the constitution to prevent change in government, as arguably in Burundi. Member states, with their respect for national sovereignty and preference for incumbency, tend to err on the side of the status quo. The commission, as guardian of AU norms and principles, tends to lean toward democratic transitions, but its position is not consistent. Ambivalence was further compounded by uncertainty around the circumstances in Burundi: whether Nkurunziza was eligible for a third-term was not cut and dry; the constitutional court, one of whose justices claims was manipulated and intimidated, concluded his first term did not count because he was appointed, not elected, and ruled he could stand again.[fn]The AUC did not condemn Rwanda’s or Congo Brazzaville’s constitutional changes in 2015. Some donors believed Nkurunziza’s case had validity. Crisis Group interviews, Western diplomat, April 2015; Judge Sylvere Nimpagaritse, Brussels, May 2016.Hide Footnote

The AU also upholds Article 4(h), “right of the Union to intervene in a Member State … in respect of grave circumstances, namely: war crimes, genocide and crimes against humanity”. Member states are historically reluctant to approve military or other action against a government, even if mass atrocities are being committed, as in Darfur (2004-2005) or South Sudan (2013-2014). The unprecedented PSC decision to invoke 4(h) for MAPROBU can be seen as a test of the AU’s graduation from the principle of non-interference, the bedrock of the AU’s predecessor, the Organisation of African Unity, to that of non-indifference.[fn]Paul D. Williams, “The African Union’s Conflict Management Capabilities”, Council on Foreign Relations, October 2011. An AU official said the AUC chairperson viewed the deployment decision as such a test. Crisis Group interview, Addis Ababa, 31 January 2016.Hide Footnote

III. The Initial Response

Dissent turned to defiance in March 2015 when demonstrators took to the streets in Bujumbura, Burundi’s capital. In the balance was not just the future of President Nkurunziza, but also the survival of the 2000 Arusha peace agreement, which included detailed power-sharing provisions. Protests increased in scale and frequency in April and May following Nkurunziza’s decision to stand for re-election, leading to violent clashes with security forces.[fn]This briefing focuses on AU and wider international responses to the crisis. For a full evaluation of the internal politics, see Crisis Group Africa Report N°224, Elections in Burundi: Moment of Truth, 17 April 2015; and Briefing N°111, Burundi: Peace Sacrificed, 29 May 2015.Hide Footnote

A. Public Diplomacy

Some praised the AU for its early intervention and strong position as the crisis began, with Dlamini-Zuma setting the tone and direction.[fn]Crisis Group interviews, U.S. officials, Addis Ababa, 29 January 2016; EU officials, Brussels, 29 February 2016; Addis Ababa, 19 April 2016. One official even called the AU’s response up until the 17 December communiqué “a model of AU diplomacy”.Hide Footnote  As tensions rose in March 2015, the AU political affairs commissioner and then the chairperson visited Bujumbura and called on the president to adhere to the constitution and Arusha, as well as for dialogue to ensure a credible election process and resolve disagreements over the third term. At first, the PSC and PSD appeared not to adopt the same tough stance, but as the situation deteriorated, positions began to align, and both the commission and the PSC engaged actively and concertedly.[fn]Crisis Group interview, UN official, Addis Ababa, 23 March 2016; “AU Commission Chairperson concludes Burundi visit with cautious optimism”, press release, AUC, 27 March 2015. The PSC urged stakeholders to respect the pending constitutional court decision on Nkurunziza’s eligibility, while Zuma dismissed the court’s findings. Zuma, Dlamini (DlaminiZuma). “Other than the #Burundi Court, all interpretations of the constitution & #Arusha Agreements are clear that there shouldn't be a third term.” 7 May 2015, 5:43 p.m. Tweet. Crisis Group interview, UN official, Addis Ababa, 15 March 2016. See also AU PSC communiqué, PSC/PR/COMM.(DI), 28 April 2015. Zuma issued nineteen press statements between March and December 2015. The PSC discussed Burundi at least monthly from March 2015.Hide Footnote

The AUC unprecedentedly refused to send a monitoring team because conditions for free and fair elections did not exist.[fn]AU communiqué, 28 June 2015.Hide Footnote  Many usual preventative diplomacy tools – high-level delegations, a special envoy, human rights and military observers, sanctions and investigation into human rights violations – were utilised, but to little effect, in part because without member states’ full political support they were not wholeheartedly deployed. The government was defiant, making concessions but then stalling or reneging on implementation.[fn]The AU has been unable to conclude an MoU with the government for 200 human rights and military observers. Only 42 have been deployed, and they cannot operate freely. In its 17 October 2015 communiqué, the PSC threatened sanctions but the AUC has yet to draw up a list of targets, let alone attempt to enforce them. The government made commitments for inclusive dialogue with the opposition only to go back on them or set unrealistic conditions, including refusing to deal with the opposition Conseil national pour le respect de l’accord d’Arusha pour la paix et la réconciliation au Burundi et de l’Etat de droit (CNARED) coalition. Yolande Bouka, Nanjala Nyabola, “The Crisis in Burundi and the Apathy of International Politics”, Heinrich-Böll-Stiftung, April 2016.Hide Footnote

In accordance with the principle of subsidiarity that generally governs its cooperation with regional economic communities, the AU ceded primary responsibility for the crisis to the EAC. Initially, they worked closely: Zuma attended EAC summits, and the PSC endorsed EAC decisions, including President Museveni’s appointment as chief negotiator in July 2016.[fn]EAC communiqué, 6 July 2016.Hide Footnote  Engaged in his own contentious election, he delegated responsibility to Defence Minister Crispus Kiyonga. As the crisis dragged on, it became clear the EAC-led dialogue was making no headway. But without consent of Museveni or the other regional leaders, the AU was unable to insert itself into the mediation process and had to settle for expressing concern about the slow pace and endorsing the EAC’s lacklustre efforts.[fn]Crisis Group interviews, civil society actor, Addis Ababa, 11 December 2015; UN official, Nairobi, 23 March 2016. Also see PSC communiqué, PSC/PR/COMM.DLI, 17 October 2015.Hide Footnote

B. Private Diplomacy

One reason the AUC’s and PSC’s aggressive public postures have had limited success is that they have not been backed by persistent private diplomacy. Critics suggest they should have both engaged earlier at the lower levels and mobilised a high-level dialogue with interlocutors Nkurunziza trusted.[fn]Crisis Group interviews, AU and UN officials, Addis Ababa, 27-31 January 2016.Hide Footnote  Zuma could have personally engaged more with Nkurunziza and others. Foreign minister when South Africa helped broker Arusha, she arguably has a great stake in the AU’s resolution of the crisis. An official suggested she believed the intervention would demonstrate the organisation’s transition from a position of non-interference to one of non-indifference. Having failed in March 2015 to persuade Nkurunziza against a third term, however, she did not really try again.[fn]Crisis Group interviews, AU, EU officials, Addis Ababa, 9 December 2015, 31 January 2016. Once President Nkurunziza committed to standing, preventative diplomacy options narrowed.Hide Footnote

The AU also lacked consistent representation in Bujumbura, relying instead on ad hoc deployment of special envoys, high-level delegations and commission figures, including the chairperson. Following the recall of its outspoken special representative for Burundi and the Great Lakes, Boubacar Diarra, in April 2015, it had no emissary for three critical months as the crisis rapidly escalated.[fn]Nkurunziza requested Diarra’s recall because he opposed his third term bid. According to an AU official, the chairperson acquiesced because she hoped it would win favour with the president. Crisis Group interview, AU official, Addis Ababa, 31 January 2016.Hide Footnote

Diarra’s replacement, Ibrahima Fall, a former UN assistant Secretary-General for Human Rights and for Political Affairs, as well as special representative of the Secretary-General (SRSG) for the Great Lakes region, has had a more discrete presence in Bujumbura. While keeping lines of communication open with the government, however, he has been unable to conclude the MoU for the human rights and military observers and has not proactively engaged with influential actors from the country or region. More dynamic representation, backed by international consensus, might have been better able to advance AU decisions.[fn]Crisis Group interviews, AU and UN officials, Addis Ababa, 29-31 January 2016; Western diplomat, Addis Ababa, 15 March 2016.

IV. The 17 December Communiqué and its Aftermath

On 17 December 2015, the PSC issued a communiqué authorising a 5,000-strong African Prevention and Protection Mission (MAPROBU) to prevent deterioration of security, protect civilians and help create conditions needed for a credible inter-Burundian dialogue. It gave the government 96 hours to accept. If rebuffed, the PSC agreed, it would recommend that the Assembly of Heads of State and Government (the AU’s highest decision-making body) invoke Article 4(h) of the Constitutive Act, which allows intervention in cases of war crimes, genocide and crimes against humanity.[fn]AU PSC communiqué, PSC/PR/COMM.(DLVX), 17 December 2015.Hide Footnote  This bold decision broke new ground in two respects: first, an ultimatum to a sitting president; secondly, invocation of Article 4(h). The Nkurunziza government quickly refused to admit foreign troops.[fn]“Burundi: We will not allow foreign troops to enter”, Al Jazeera, 21 December 2015.Hide Footnote

A. The Context for the Decision

Six days before the PSC communiqué, violence in Bujumbura intensified significantly. Reportedly, at least 87 were killed, some summarily executed, in intense fighting sparked by coordinated grenade attacks by armed opposition groups on military installations on 11 December. Observations from an African Commission on Human and Peoples’ Rights (ACHPR) fact-finding mission, AU human rights observers’ reports and social and traditional media accounts directly influenced PSC decisions.[fn]“Burundi: 87 killed in worst violence since April coup attempt”, The Guardian, 12 December 2015. “Report of the Fact Finding Mission of the African Commission on Human and Peoples’ Rights to Burundi”, ACHPR, May 2016. Crisis Group interviews, Addis Ababa 31 January; Brussels, 29 February 2016.Hide Footnote  There was strong belief in Addis Ababa, driven by the AUC but supported by a few key PSC members, that rapidly deteriorating security and grave human rights abuse warranted decisive action. With the 1994 Rwandan genocide in mind, the PSD hoped to demonstrate the AU would not stand idly by. “The communiqué made it clear that it considered the situation grave; it put Burundi on notice and sent a message to the world that the AU had done what it could”, an AU official said.[fn]Nigeria, Ethiopia, Algeria and to a lesser extent Uganda were strongly in favour of the PSD drafted communiqué. Crisis Group interviews, Addis Ababa, 25-31 January, 14-18 March 2016.Hide Footnote

B. Miscalculations and Missteps

The PSC intended the communiqué to freeze the crisis and force the government to negotiate.[fn]Crisis Group interview, 8 February 2016.Hide Footnote  While it arguably focused international attention, helping to curb the worst security force excesses and spurring Museveni to kick-start the stalled EAC mediation, it failed to engage Nkurunziza in an inclusive political dialogue with the opposition.[fn]Crisis Group interviews, AU officials, Addis Ababa, 27 January 2016; UN officials, New York, 9 February 2016.Hide Footnote  The government dismissed MAPROBU as an “invasion and occupation force”, shocking some in the AUC, which had been convinced it would grudgingly accept the mission. Burundi watchers said the AU lacked credible situational analysis and misread Nkurunziza’s character.[fn]“Burundi rejects African Union peacekeepers as ‘invasion force’”, Agence France-Presse, 20 December 2016. Crisis Group interviews, AU, UN officials, Western diplomats, Addis Ababa, 31 January, 16 March, 19 April 2016.Hide Footnote  AU, UN and Western officials called the PSC’s ultimatum a mistake and an insult.[fn]An official described it as an “extraordinary thing to do to a sovereign government”. The AUC chairperson informed the UN Secretary-General before Nkurunziza. Crisis Group interviews, Western diplomats, New York, 8 February 2016; AU, UN officials, Addis Ababa, 16 March 2016.Hide Footnote  It was, however, just one in a series of AUC and PSC missteps and miscalculations.

Article 4(h) touched a nerve with those member states whose democratic credentials and human rights records were criticised and feared it might one day be applied to them. Invoking that article, which sets a high bar – verification of war crimes, genocide or crimes against humanity – raised the stakes considerably, and the PSC left itself no other tool with which to ratchet up pressure on the government. Some PSC delegations did not understand the ramifications of its use or consult sufficiently with capitals.[fn]Crisis Group interviews, AU, UN, EU officials, Western diplomats, Addis Ababa, 25-31 January 2016.Hide Footnote  Shortcomings of PSC working practices were exposed. Unlike the UN Security Council, whose members draft resolutions, with one leading and consulting in varying degrees, the AUC mostly writes PSC communiqués, typically with little or no input from members. Meetings frequently leave scant time to discuss content; often there is no chance to work the outcomes of deliberations into the final text. As a result, the PSC tends not to buy sufficiently into its own resolutions. The 17 December communiqué was no exception.

The heavy PSC workload stretches often under-staffed Addis Ababa missions. Most ambassadors did not use the 24-hour silence period to clear the communiqué at home. Blindsided, some governments were unwilling to give wholehearted support. The AUC also neglected to confirm that Council members had briefed their governments in advance.[fn]An official said only about a third of member states sent the communiqué to capitals in the silence period. Crisis Group interview, Addis Ababa, 17 March 2016. South Africa’s international relations and cooperation department issued a statement on Burundi with no MAPROBU mention. Asked to clarify the stance, spokesperson Clayson Monyela said, “we support the AU position yes”. “South Africa expresses concern over the situation in Burundi”, press release, 18 December 2015; Monyela, Clayson (ClaysonMonyela). "@geoffreyyork We support the AU position yes.", 20 Dec 2015, 18:29 UTC. Tweet, @Clayson Monyela, 10.29am 20 December 2015. After dismissing use of force, Tanzanian Foreign Minister Augustine Mahiga announced MAPROBU support only in January. “Mahiga calls for talks to end the bloodshed”, The Citizen, 20 December 2015. “Consultations between [AUC] and Tanzania on the situation in Burundi”, AU press release, 8 January 2016. Crisis Group interviews, AU officials, Addis Ababa, 31 January, 16-17 March 2016.Hide Footnote  It had hoped the communiqué would spur bolder action, but the response was cautious. A sense of urgency and misplaced belief that others, including the UN, would follow the AU lead led to failure to give advance word to those called upon to endorse and assist MAPROBU. The absence of planning – no concept of operations or commitments from troop contributing countries – and Burundi’s rejection of an AU force, made it hard for the Security Council to give the anticipated backing.[fn]“We [the AU] can’t always back it up with action, but we can enable others”. Crisis Group interview, AU official, 9 December 2016. The Security Council merely “took note with interest” of the communiqué. Press statement, 19 December 2015. An AU official said the AUC subsequently believed the Council was the obstacle to deployment. Crisis Group interview, 29 January 2016; interviews, UN officials and Western diplomats, New York, 8-9 February 2016.Hide Footnote

C. Force Preparedness

The AUC did not intend to deploy MAPROBU immediately. Regardless of official statements, the East African Standby Force (EASF), expected to provide forces, was not ready, and without Burundi’s consent, the AU needed Security Council approval.[fn]In January 2015, the EASF was officially declared able to deploy within fourteen days with 90 days of supplies, but a Western diplomat estimated it would take at least six months to set up operation systems for a Burundi mission. Crisis Group interview, Addis Ababa, 26 January 2016. Use of force against a sovereign state is permitted only in self-defence or UN Charter Chapter VII authorisation. The AU communiqué requested such a Security Council resolution for MAPROBU. Paul D Williams, “The African Union’s Coercive Diplomacy in Burundi”, International Peace Institute Special Report, 18 December 2015.Hide Footnote  Finally, the AUC acknowledged that MAPROBU required considerable foreign financial and logistical support. It had been conceived primarily as a threat with which to push Nkurunziza into a genuine dialogue, rather than a tangible rapid reaction force, and he called the AU’s bluff.[fn]PSC Communiqué, op. cit. The AUC may also have wanted to use MAPROBU to push automatic use of UN assessed contributions to finance AU-led peace support missions. Crisis Group interviews, AU, UN, EU officials, Addis Ababa, 26-29 January 2016; New York, 8 February 2016.Hide Footnote

D. The AU Summit and the Decision Not to Deploy MAPROBU

The government’s refusal to accept MAPROBU meant the future of the stabilisation force rested with African heads of states due to meet at the biannual AU summit in Addis Ababa, 30-31 January 2016. Invoking Article 4(h) left the door open for unprecedented forcible deployment, if two thirds agreed the security situation was sufficiently grave. By then, violence had receded, but the AUC pressed on with a bruising approval process, exposing deep fault lines with member states concerned that the mission would violate Burundi’s sovereignty and impede dialogue. The AUC might have persuaded them to pressure Burundi to accept a more limited intervention force, but understaffed and underfunded, it lacked capacity for the political work needed to build consensus around deployment; some were unaware that was needed or naively confident a pared-down version would be endorsed. Nkurunziza’s emissaries travelled the continent, arguing the crisis was a case of relatively mild post-electoral violence now under control.[fn]The government and opposition dialled back confrontation prior to the summit. Crisis Group interviews, AU, UN officials, Addis Ababa, 16 March 2016. Solomon Dersso, “To intervene or not to intervene? An inside view of the AU’s decision-making on Article 4(h) and Burundi”, World Peace Foundation, February 2016. Crisis Group interviews, AU, EU officials, Western diplomat, Addis Ababa, 30-31 January, 17 March 2016. “Burundi reaffirms dialogue commitment”, Daily News, 16 January 2016; “SA presence in Burundi talks welcomed”, African News Agency, 22 January 2016.Hide Footnote

E. The AU High-level Delegation

Trying to keep MAPROBU alive, Peace and Security Commissioner Smail Chergui recommended the assembly send a senior delegation to Bujumbura to consult. This again laid bare divisions between member states and the AUC, which insisted deliberations focus solely on an inclusive intra-Burundian dialogue.[fn]Dersso, “To intervene?”, op. cit. The high-level delegation – presidents of Senegal, Gabon, Mauritania and South Africa, prime minister of Ethiopia – was in Burundi 25-26 February. The 29 January 2016 PSC communiqué limited the mission’s discussions to dialogue, PSC/AHG/COMM.3(DLXXI). A subsequent AUC statement widened the mandate to include MAPROBU. “The African Union appoints High-Level Delegation to Burundi”, 4 February 2016.Hide Footnote  It also revealed member-state disagreements. South African President Jacob Zuma, the delegation head, dominated proceedings and set the agenda. Pretoria was pivotal in mediation and peacekeeping during the civil war, and close ties between the African National Congress and Burundian National Council for the Defence of Democracy-Forces for the Defence of Democracy (CNDD-FDD), meant the president was sympathetic to Nkurunziza. Separate, contradictory communiqués, the first unilaterally released by Zuma, the second by the AUC with delegation-member support, reinforced an appearance of disarray.[fn]Crisis Group interview, African diplomat, Addis Ababa, 20 April 2016; “AU High Level Delegation to Burundi conclusion statement”, South African presidency, 27 February 2016; “Communiqué of the visit of the [AU] high level delegation to Burundi”, AU, 29 February 2016.Hide Footnote  The AU lost a major opportunity to insert itself into the mediation process.

F. The Fallout

Leaders’ failure to endorse MAPROBU severely damaged AU credibility, revealing the gap between ambition and capabilities. The 17 December communiqué was called “unrealistic” and “un-strategic”, and the military force row distracted from dialogue.[fn]Crisis Group interviews, Addis Ababa, 25-31 January; New York, 8-9 February; Nairobi 23-24 March 2016. “The Burundi Intervention that Wasn’t”, Foreign Policy (online), 2 February 2016; “Intervention that Never Was”, Africa Confidential, 5 February 2016. “African Union diplomacy fails to take off in Burundi. Literally and metaphorically”, UN Dispatch, 29 February 2016; African Union goes backwards on Burundi”, Daily Maverick, 31 January 2016.
Hide Footnote
 Invoking Article 4(h) divided member states. Some were adamant sovereignty trumps human rights; others did not share AUC analysis that Burundi was near to catastrophic violence. The AUC was seen to have over-stepped: one of its senior officials said, “we have embarrassed the continent”. Some member states, notably Egypt, dislike AUC influence over PSC decision-making, and many predict a push for member states to draft future measures. The AU lost any authority in Bujumbura and is marginalised in efforts to resolve the crisis.[fn]Gambia’s President Yahya Jammeh, for example, was clearly opposed to the use of force without consent; Tanzanian Foreign Minister Mahiga did not believe military intervention was warranted. Derrso, “To intervene or not?”, op. cit. Crisis Group interviews, AU, EU officials, African diplomats, Addis Ababa, 25-31 January, 17 March 2016; New York, 8 February 2016.Hide Footnote  Embarrassed by the failure of MAPROBU and other diplomatic attempts and faced with member-state indifference, the AUC and PSC appear to have lost impetus. Failing to act decisively, the AU sacrificed its moral authority to speak out about incumbents manipulating or eliminating constitutional term limits.

V. The Wider International Response

The AU response has been disappointing, but neither regional powers nor the UN fared better. Domestic considerations, power politics and historic allegiances, as well as antagonisms, have shaped Burundi’s neighbours’ hesitant response. At the UN, the divergent policies of Security Council members and divisions within and between the Council and Secretariat have thwarted attempts to find a solution. Institutional rivalries, coupled with the absence of a shared analysis, prevented a coordinated approach. The government exploited this disunity, playing the main actors – the AU, EAC and UN – off against one another.

A. Regional Dynamics

At the beginning of the crisis, a key concern was the legitimacy of Nkurunziza’s re-election attempt. Two allies, Tanzania’s President Jakaya Kikwete and South Africa’s President Zuma, cautioned him against standing. However, the EAC did not condemn the third-term bid – it would have been difficult for it to do so following the failed May 2015 coup attempt.[fn]Nina Wilén, “The rationales behind the EAC members’ response to the Burundi Crisis”, Georgetown Journal of International Affairs, Volume 17, Number 1, Winter/Spring 2016. “Tanzania’s Kikwete says Burundi should heed president term limit; risks regional tensions”, Mail & Guardian, 20 March 2016; “Zuma calls for Nkurunziza not to stand for a third term”, video, SABC Digital News, May 8 2015. A 31 May 2015 EAC communiqué stopped short of calling for Nkurunziza not to stand, insisting instead on postponing elections.Hide Footnote  For others, it was never a consideration. Uganda’s Museveni, in power for 29 years and having amended presidential term limits in 2005, was seeking a fifth mandate. President José Eduardo dos Santos has ruled Angola since 1979. Next door in the Democratic Republic of Congo (DRC), Joseph Kabila has been preparing his own attempt to extend his stay in office.

Zuma Calls for Nkurunziza Not to Stand for Third Term

YouTube/SABC Digital News

Once Nkurunziza won the election, the third term became moot in the region and much of Africa. For many African leaders, especially those who fought long liberation struggles, incumbency is paramount. Factor in the special relationships of Nkurunziza and his party with many of them and their parties, and it is easy to see why he has not come under serious pressure.[fn]After the July 2015 summit, and aside from a half-hearted attempt to kick-start the inter-Burundian dialogue in December. EAC leaders did little and did not meet until March. For the U.S. and other Western countries, the third term remains a concern (see also Section V.B.). Sheltered by Tanzania during the civil war, the ruling CNDD-FDD has links to its security services. Zuma and Museveni assisted the settlement that ended the war and brought Nkurunziza to power. Crisis Group interviews, UN officials, Addis Ababa, 29 January, 15 March 2016.Hide Footnote  Rwanda is the exception, consistently calling for him to step down. It argues not that term extensions are inherently bad – President Kagame pushed through a constitutional change that allows him to keep office until 2034 – but that Burundi has not prospered under his rule. It is also alarmed by the rise of anti-Tutsi hate speech, similar to that used prior to the 1994 Rwandan genocide, and fears the Democratic Forces for the Liberation of Rwanda could gain a foothold. The UN and U.S., among others, have accused Rwanda of supporting Burundian armed opposition groups to weaken Nkurunziza. Distrust of Rwanda’s perceived destabilisation of the region, and its alleged assistance to the armed insurrection, have shaped the response of other neighbours and contribute to EAC paralysis.[fn]Kagame told a University College (London) economist the crisis was about performance, not third terms. “Rwanda’s President Kagame tell Burundi’s Nkurunziza to step down”, Kenya Today, 11 May 2015. “Rwanda aids Burundi rebels, North Korea arms Congo – UN experts”, Reuters, 12 May 2016; US accuses Rwanda of stoking violence in Burundi”, Agence France-Presse, 11 February 2016. Yolande Bouka, “Missing the Target: The African Union’s Mediating Efforts in Burundi”, Egmont Royal Institute of International Relations, June 2016.Hide Footnote

Former Tanzanian President Benjamin Mkapa’s appointment as facilitator in March 2016 signalled increased Tanzanian engagement and revived optimism in the EAC process. However, negotiation delays, interference from the new EAC secretary general, Libérat Mfumukeko (an Nkurunziza loyalist), and the government’s refusal to sit down with the Conseil national pour le respect de l’accord d’Arusha pour la paix et la réconciliation au Burundi et de l’Etat de droit (CNARED), the opposition coalition, during the first round of talks on 21 May, sparked criticism of him. His decision to meet with exiled opponents separately in Brussels somewhat allayed fears, but the discussions collapsed in July, amid government walkouts and opposition boycotts.[fn]“Violence increases as Burundi talks delayed”, Voice of America, 2 May 2016; and “Burundi peace talks open in Tanzania with opposition criticism”, Reuters, 21 May 2016. “Burundi: pour le Cnared, la rencontre avec Mkapa est déjà une ‘victoire’”, Radio France Internationale (RFI), 10 June 2016; “Burundi: peace talks collapse in Arusha”, The East African, 16 July 2016.Hide Footnote

B. The UN’s Role

The UN has been actively involved. There have been two Security Council visits and another from the Secretary-General (February 2016), plus Council meetings resulting in three resolutions, two presidential statements and a series of press releases. However, these have borne little fruit. Two reasons stand out. First, disagreements over how to mitigate and resolve the crisis stymied decisive response. Secondly, the Council tried to follow the African lead, but the continent has rarely spoken with a single voice. The Council’s slow response, despite warnings by the Secretariat, also illustrates its declining influence in Burundi. This was particularly evident in the withdrawal, under pressure, of the UN’s special political mission and its replacement in January 2015 by an electoral observation mission, with a significantly narrower mandate.[fn]For example, in January 2015, the Secretariat warned that the elections could “either consolidate or unravel the peace consolidation efforts undertaken since the Arusha accords” and pointed to the eroding spirit of the accords, shrinking of political space, intimidation of opposition supporters and limitations on free assembly and expression. “Report on the United Nations Office in Burundi”, UNSC S/2015/36, 19 January 2015. “Burundi briefing and consultations”, What’s in Blue (www.whatsinblue.org), 27 January 2014. See also, Crisis Group Africa Report N°192, Burundi: Bye-bye Arusha?, 25 October 2012.Hide Footnote

As the 2015 elections approached, some Western governments focused statements on opposition to Nkurunziza’s third term. A number of non-permanent Security Council members and Secretariat officials argue that emphasising this, rather than the risk of violence and importance of dialogue, complicated consensus efforts in a Council whose divisions over other crises were starting to seep into its response. Its first visit to Burundi in March 2015 was of limited value, as members had not agreed on a clear message or possible leverage.[fn]Crisis Group interviews, New York, December 2015, February 2016. Whether a different focus would have made consensus easier is unclear, given already acrimonious relations over Ukraine and Syria. Some also argue that the insistence Nkurunziza stand down also made the opposition less likely to compromise. Crisis Group interviews, UN officials, New York, July 2015.Hide Footnote

An idea that gained traction after the second, January 2016 visit, was police deployment. Some Western permanent members initially wanted an armed international presence to protect civilians in the event of widespread violence following increasingly sharp anti-Tutsi rhetoric.[fn]Crisis Group interviews, UN and member state officials, New York, February 2016.Hide Footnote  Subsequently, as the Council perceived an immediate genocide threat subsiding, proponents argued police could increase monitoring and so deter continuing human rights abuse by authorities.

Related Content

In April 2016, the Secretariat presented the Council three deployment options: light (twenty-50 unarmed police to work with Burundian forces); middle (228 unarmed police); and a 3,000-strong force.[fn]Angola, China, Egypt, Russia and Venezuela favoured the light option, largely toeing Burundi’s line. France, the penholder, considered the middle course most viable given Council dynamics. The U.S. initially backed the most robust option, with formed police units to patrol visibly and report regularly. Crisis Group interviews, member-state officials, New York, April-August 2016. “Burundi: briefing on options for police deployment”, What’s in Blue, 26 April 2016.Hide Footnote  After months of intermittent discussion, the Council authorised deployment of up to 228 to monitor the security situation and support the office of the high commissioner for human rights in monitoring human rights abuses. Russia, which initially wanted the lighter option, voted in favour, but abstentions from Angola, China, Egypt, and Venezuela, all citing Bujumbura’s opposition to a large police presence, dampened any sense of Council unity. The Burundian government swiftly rejected the resolution.[fn]Resolution 2303, which authorised the police, also asked the Secretariat to report on how the UN could facilitate deployment of AU observers and for proposals on how the UN police component could cooperate with them. It did not specify whether police would be armed, but Council members generally assumed they would not be, per the Secretariat’s second option. UNSC S/RES/2303, 29 July 2016. Crisis Group interviews, New York. “Government reaction following the adoption by the UN Security Council of the Resolution 2303 (2016)”, 2 August 2016.Hide Footnote  Some Council members have privately urged the UN Secretariat to liaise with Nkurunziza to soften his position on the police, and some have engaged bilaterally with Burundi.[fn]Crisis Group interviews, New York, September 2016.Hide Footnote  But the Council has taken no further collective action on its resolution – and has not even formally discussed the Burundi situation since July – further illustrating the degree to which divisions undermine its effectiveness.

As Russia and China tend to take their cue on regional crises from African members, it further complicates the Council’s role that the A3 have not spoken with one voice or followed AU policy. During discussions in November 2015, Angola counselled against using language directly from earlier PSC communiqués. Russia and China supported its position until the AUC persuaded them to withdraw their objections. Since joining the Council in January 2016, Egypt has taken a strong pro-government, anti-interventionist stand on many crises, joining Angola, China, Russia and Venezuela to resist robust UN action in Burundi.[fn]Crisis Group interviews, AU and UN officials, Addis Ababa, 9 December 2015, 29 January 2016. Cairo wants to uphold the principle of non-interference, support a fellow Nile Basin country and exact payback for the AUC’s suspension of its participation following the 2013 ousting of President Mohamed Morsi. Egypt also became a member of the AU PSC in April 2016. Crisis Group interviews, AU official, Western diplomats, Addis Ababa, 25 January, 16 March 2016.Hide Footnote

C. International Discord and Disunity

Institutional rivalries, coupled with the absence of shared analysis, have prevented coordinated approaches.[fn]While strategic coordination has been limited, cooperation between the AU and UN at a working-level in Burundi has been more effective.Hide Footnote  Early attempts at collective action, such as initial cooperation between the AU and EAC and formation of the Joint International Facilitation Team (representing the EAC, AU, International Conference on the Great Lakes Region and UN), were short lived, leaving each appearing to defend its own role or trying to carve out space at the expense of a competitor.[fn]The joint team, created at a June 2015 PSC summit, apparently met just once, after which it called for election postponement to 30 July. “Statement by the Joint International Facilitation Team on Burundi”, 26 June 2015.Hide Footnote

Though its mediation stalled, the EAC was unwilling to allow the AU to engage further. The AUC has bristled at the UN’s enhanced role after MAPROBU’s failure, fearing that the Security Council’s proposed police contingent would steal the mandate and draw financial support away from its human rights and military observers. Nkurunziza has been quick to exploit UN-AUC tension, at times courting the former while giving the cold shoulder to the latter and lobbying Council members not to support the AU’s observer mission.[fn]Crisis Group interviews, UN official, 15 March 2016; AU officials, Addis Ababa, 31 January, 22 March 2016; UN officials and African diplomats, New York, 25-26 May 2016.Hide Footnote  Mkapa, like his predecessor, has been slow to accept UN help, relying instead on the EAC secretariat, headed by a Burundian diplomat, despite its limited mediation experience.

The international community missed a number of other opportunities to demonstrate joint commitment to resolving the crisis, such as could have been done if the PSC and AUC chairperson had joined UN counterparts’ Burundi visits, assuming positions and messaging were aligned in advance and then relayed clearly to Nkurunziza.[fn]Crisis Group interviews, AU and UN officials, Addis Ababa, 29 January 2016.Hide Footnote  They should take up further chances for collaboration.

VI. Moving Forward

The crisis is at an impasse, and genuine inclusive dialogue between government and opposition seems a remote possibility. Positions are entrenched, and both are playing for time as the toll rises.[fn]Crisis Group Africa Report N°235, Burundi: A Dangerous Third Term, 20 May 2016; commentary, “Insights from the Burundian Crisis (III): Back to Arusha and the Politics of Dialogue”, 7 July 2016.Hide Footnote  Despite widespread censure, the government has proven it can resist international pressure. The EAC-led mediation under Museveni has made little progress but is currently the only mechanism for dialogue, so the AU, UN and others should provide logistical, financial and political support. More crucially, the AU, UN and EAC need to build consensus and coordinate better. A contact group would be an important step toward aligning positions and could inject new life into negotiations. As requested by Mkapa, Museveni (and other EAC leaders) must become more personally engaged. Having agreed to mediate, he should shoulder the responsibilities and, as a minimum, set out his vision for the way forward.

Two key leverage points have not yet been fully utilised: targeted sanctions and denying Burundi’s financially advantageous participation in AU and UN peacekeeping operations. Implementation of limited sanctions has been haphazard and half-hearted.[fn]For example, in October 2015, the EU sanctioned four government members; two months later the U.S. sanctioned a slightly different group.  But with government revenue falling and further GDP decline forecast, the government is vulnerable to economic pressure. UN sanctions are unlikely given Security Council divisions, but the AU should implement the PSC’s October 2015 decision to impose targeted sanctions. Despite the difficulty of ensuring full adherence in a region generally supportive of Nkurunziza, that would send an important signal and add further pressure on the government. The EU and the U.S., whose special Great Lakes region envoy has already condemned the use of inflammatory rhetoric, should expand their existing sanctions to include those propagating hate speech. Removal of sanctions should be clearly benchmarked to reducing violence, ending impunity and starting dialogue.[fn]“Burundi braces for more violence”, Deutsche Welle, 7 November 2015. Crisis Group Statement, “Burundi: Time for Tough Messages”, 24 February 2016.

Funds from the army’s contribution to the AU mission in Somalia (AMISOM) and the UN mission in the Central African Republic are a vital lifeline for the government. Removal could force it to change its dangerous trajectory. The EU and AU should finalise changes so that AMISOM money is paid directly to the soldiers. The AU and its partners should also solicit other troop contributors to replace Burundian soldiers within AMISOM eventually, so as to pressure the authorities to begin an open, genuine dialogue.[fn] “EU takes aim where it hurts Burundi – peacekeeper funding”, Reuters, 29 March 2016. Crisis Group Report, Burundi: A Dangerous Third Term, op. cit. A UN Human Rights Council-mandated investigation that found widespread, systematic human rights violations in Burundi has urged the UN and AU to “phase out the use of Burundian troops in peacekeeping operations while the crisis continues”. “Report of the UN Independent Investigation on Burundi (UNIIB) established pursuant to Human Rights Council Resolution S-24/1*”, UN HRC A/HRC/33/37, 20 September 2016, p. 23. The Burundian government rejected the findings as “politically exaggerated”. Foreign Minister Alain Nyamitwe, UN General Assembly speech, 24 September 2016. 

VII. Conclusion

Attention and interest has shifted away from the Burundi crisis. The government appears to have realised that keeping casualties to a minimum limits scrutiny and is forging ahead with plans to change the constitution and abolish presidential term limits. Discussions about this could begin at October’s parliamentary session. Though the 2020 election cycle seems far off, international actors should press harder for a political settlement. Postponing firmer, more unified action would leave the country at best in a permanent state of low intensity violence. Despite internal divisions, the AU should not disengage but rather insist on deployment of its human rights and military observers. Institutional rivalries between the AU, EAC and UN must not block the concerted international action needed to secure a negotiated solution and prevent a deeper decent into civil war.

Addis Ababa/Nairobi/Brussels, 28 September 2016

VIII. Glossary

A3 The collective name for the three rotating African members of the UN Security Council, currently Angola, Egypt and Senegal.

ACHPR The African Commission on Human and People’s Rights oversees and interprets the African Charter on Human and People’s Rights, an international instrument designed to promote and protect human rights and basic freedoms on the continent.

AMISO African Union Mission in Somalia.

APSA The African Peace and Security Architecture, the umbrella term for the AU’s mechanisms for promoting peace, security and stability in Africa. It consists of the Peace and Security Council, the Panel of the Wise, the Continental Early Warning System, the Africa Standby Force and the Peace Fund.

AU African Union.

AUC The African Union Commission is the AU’s secretariat. Headquartered in Addis Ababa, it is led by the Chairperson (currently Nkosazana Dlamini-Zuma) and supported by a deputy chairperson and eight commissioners.

CEWS The Continental Early Warning System gathers information about potential conflicts or threats to the security of AU member states. It is housed in the Peace and Security Department’s Conflict Prevention and Early Warning Division.

CNARED The National Council for the Restoration of the Arusha Agreement and the Rule of Law is a coalition of the main exiled Burundian opposition movements, as well as two former presidents, members of civil society and CNDD-FDD dissidents.

CNDD-FDD The National Council for the Defence of Democracy-Forces for the Defence of Democracy is Burundi’s ruling party. During the civil war (1993-2005) it was a significant rebel group.

DRC Democratic Republic of Congo.

EAC The East African Community is an inter-governmental organisation with six members: Burundi, Kenya, Rwanda, South Sudan, Tanzania and Uganda.

EASF The Eastern Africa Standby Force is one of the five regional multidimensional forces that make up the African Standby Force. It members are Burundi, Comoros, Djibouti, Ethiopia, Kenya, Rwanda, Seychelles, Somalia, Sudan and Uganda.

EU European Union.

MAPROBU African Prevention and Protection Mission in Burundi.

PSC The Peace and Security Council, the AU’s decision-making organ for the prevention, management and resolution of conflict, is a standing committee of fifteen members elected according to regional representation and rotation.

PSD The AUC’s Peace and Security Department supports the PSC and Commission in activities related to the promotion of peace, security and stability in Africa.

MoU Memorandum of understanding.

RECs The Regional Economic Communities are groupings of African states formed to facilitate economic integration between member states. They are increasingly involved in coordinating AU member states’ interests in areas such as peace and security, development and governance. The AU recognises eight RECs.

RMs The regional coordinating mechanisms of the regional standby forces of Eastern and Northern Africa.

Le président du Mali, Ibrahim Boubacar Keita (G), donne l'accolade à Mahamadou Djery Maiga (D), vice-président et porte-parole du Conseil de transition de l'Etat de l'Azawad, le 20 juin 2015 à Bamako. AFP/Habibou Kouyate
Q&A / Africa

Mali’s Algiers Peace Agreement, Five Years On: An Uneasy Calm

La mise en œuvre de l’accord de paix au Mali demeure incomplète et laborieuse cinq ans après sa signature. Mathieu Pellerin analyse la situation actuelle et explique pourquoi il faut accélérer les efforts pour instaurer les réformes de fond prévues par l’accord de 2015. 

Five years after it was signed in June 2015, what has happened to the Agreement on Peace and Reconciliation in Mali?

In June 2015, the Malian government, a coalition of pro-government armed groups from northern Mali called the Platform and the Coordination of Azawad Movements (Coordination des mouvements de l’Azawad, CMA) – an alliance of rebel groups – convened in Bamako and signed an agreement to restore peace in the country. The signatories were under great pressure from an international mediation team to accept the final text, which was drafted after less than a year of often indirect negotiations. The mediation team was led by Algeria and included the UN Stabilisation Mission in Mali (MINUSMA), the Economic Community of West African States (ECOWAS), the African Union (AU) and the European Union, as well as the United States and France, who were initially designated “friends of the mediation”.

The agreement seeks to restore peace in Mali principally through a process of decentralisation or regionalisation, reconstituting a national army from the members of the former armed groups that were signatories, and boosting the economy (particularly in the north), based on dialogue, justice and national reconciliation.

None of the agreement’s five pillars have been satisfactorily applied.

The parties claim to support the agreement five years after signing it in June 2015, but its implementation has proved to be extremely difficult. The Carter Center – appointed as the Independent Observer in Mali in late 2017 – reports virtually no progress on this front: in 2017, 22 per cent of the agreement’s provisions had been put into effect, compared to 23 per cent three years later. None of the agreement’s five pillars have been satisfactorily applied.

The parties have not carried out the substantive political and institutional reforms defined in Section II of the agreement (the first section lays out the agreement’s general principles), starting with regionalisation. So far, the measures have been temporary or too limited to make any real impact on the ground. It took months of negotiation between the signatories and international partners of the Peace Agreement Monitoring Committee (Comité de suivi de l’accord, CSA) to appoint interim authorities in the northern regions, and with few tangible results. Three years on, these authorities have insufficient financial and human resources, and lack the training, to manage the regions effectively. The two new regions (Ménaka and Taoudenit) created in northern Mali, based on commitments made by President Amadou Toumani Touré in 2011, also lack resources. Voters in these regions could not choose deputies in the April 2020 legislative elections because the electoral districts had not yet been delineated.

On matters of defence and security (Section III), the process of Disarmament, Demobilisation and Reintegration (DDR) initiated by the state and backed by MINUSMA has weakened. Despite the deployment of a reconstituted Malian army battalion in February 2020 in Kidal, a hotbed of rebellion and CMA’s centre of operations in Mali’s far north, this force has never patrolled the town, and the CMA – chafing at its exclusion from a command role – has now “assigned” the battalion’s third company to Gao. The leaders of the movements and the Malian state’s chiefs of staff have not discussed the framework for a lasting means of integrating former armed groups’ members in the national army and its chain of command.

On the fifth anniversary of the agreement, this DDR process involves only 1,840 combatants from the signatory groups in an “accelerated DDR” phase, and they are not even the ones who fully reintegrated. UN Security Council Resolution 2480 (2019) set the goal of reintegrating 3,000 fighters by 2020, but it remains distant, and the next phase is uncertain. With nearly 85,000 combatants registered by the signatory groups, DDR remains incomplete and a sensitive issue. The mixed units of the Operational Coordination Mechanism (Mécanisme Opérationnel de Coordination, MOC), consisting of Malian soldiers and combatants from the signatory armed groups and partly assigned to the reconstituted army, were supposed to provide security in large towns in northern Mali. They are rarely seen on patrol, however, and have been targeted for attack, especially the 2017 Gao bombing of their camp. Some former fighters belonging to the MOC or to the reconstituted army have been involved in banditry and trafficking.

The joint administration of a long-term development fund by the Malian authorities and armed groups remains a challenge.

The parts of the agreement on development (Section IV) and reconciliation (Section V) remain largely overlooked. Nothing points to the possibility of genuine economic growth supported by the state or donors. A long-term development fund designed to support initiatives in northern Mali has been set up, but its joint administration by the Malian authorities and armed groups remains a challenge. Mali’s truth, justice and reconciliation commission, established in 2014, has continued its role as defined in the 2015 agreement, and it began holding public hearings in December 2019, but it generates hardly any interest.

Why the standstill?

The delayed implementation is symptomatic primarily of a lack of will among the signatories. Neither the Malian government nor the other parties were enthusiastic about the agreement’s text in 2015; international duress, particularly from Algeria, France and the U.S., pushed them to sign it. Civil society organisations in both northern and southern Mali that were supposed to represent local populations were effectively excluded from the process. While the Malian state and the signatory armed groups feel that outsiders foisted reconciliation upon them, southern Malians remain strongly distrustful of the former rebels and an agreement that was largely opaque to them. Many from the south think that the agreement is the first step toward an eventual partition of the country. According to the Mali-Mètre opinion survey (March 2020), “the vast majority of citizens interviewed (80.1 per cent) stated that they had ‘no’ knowledge (61 per cent) or ‘hardly any’ knowledge (19.1 per cent) of the peace agreement”.

Apart from the lack of will, the Malian state and the CMA are also keen to preserve the status quo: the CMA enjoys considerable de facto autonomy in its areas of influence in northern Mali, while many of its members have paid employment in the bodies set up by the agreements, such as the CSA and the interim authorities. In parallel, this state of affairs allows the Malian state to delay implementation of the 2015 agreement’s more sensitive provisions, particularly those implying constitutional reform. In August 2017, pressure from the public – mobilised in part against the agreement’s implementation – forced the government to postpone a draft constitutional referendum. By maintaining the status quo, the government prevents social unrest while still honouring its commitment to the international community to continue implementing the agreement.

We are not going to lay down our arms before getting what we took them up for in the first place.

The main parties to the agreement are therefore in a deadlock: the lack of political and institutional progress is leading the signatory armed groups to reject defence and security commitments. In an interview, one CMA official summed up the situation as follows: “We are not going to lay down our arms before getting what we took them up for in the first place”.

The international mediation team that pushed for the signing of the agreement has failed in its commitment to act as “the guarantor of [its] scrupulous implementation”, as specified in its text. The CSA has not exerted enough pressure on the parties to ensure the agreement’s proper implementation, in particular with regard to its key political and security provisions. International actors seem content with the status quo that allows them to focus on the jihadist threat, particularly in central Mali.

If the parties have not clashed since the agreement was signed, why does the impasse pose a problem?

The current stability is significant, and represents a source of satisfaction for some. But it is deceptive. The peace agreement may be partly responsible for the calm, but it owes more to a combination of factors that may turn out to be short-lived.

If they have failed to secure genuine implementation of the agreement, the international forces present in Mali have succeeded at deterring the signatories from resorting to the use of force. Their presence, however, will not be permanent. With instability spreading in central Mali, and across its borders, international actors such as Barkhane (a French anti-terrorism operation in the Sahel) and MINUSMA are increasingly turning their eyes elsewhere, such as Burkina Faso and Niger. In this vast region, the limited military forces (5,100 Barkhane and 13,000 MINUSMA soldiers) cannot be present everywhere.

Moreover, the stability in northern Mali is paradoxically linked to the CMA’s position of strength. Since 2015, violations of the peace agreement have pitted the armed groups of two coalitions against each other (rather than against the Malian state) due to political rivalries between the strongmen of different Touareg tribes or clashes between traffickers. The Platform – the coalition of pro-Bamako armed movements – has steadily weakened since 2017, and many of the factions have split off to join its rival, the CMA. Skirmishes are now rare again in northern Mali. Although the three parties signed the agreement in 2015, the Platform’s dwindling power has now left the CMA facing off against the government. In the longer term, the non-implementation of the agreement could give a pretext for the CMA, now in a strong position in the north of the country, to revive its quest for autonomy.

The non-implementation of the agreement could give a pretext for the CMA to revive its quest for autonomy.

The stability is also linked to the discovery of gold in the Kidal and Gourma regions. Panning for gold has effectively enabled a type of spontaneous yet temporary demobilisation of combatants from armed movements, especially the CMA. But the gold deposits will eventually run out. The current phase of artisanal mining will either come to an end or – more likely – yield to a phase of semi-mechanised mining that requires fewer workers. At that point, taking up arms could become more appealing.  

The current situation is therefore based on a precarious balance and cannot be described as a lasting solution; a flare-up of violence in the medium term cannot be ruled out. The peace process must deliver considerable progress in order to avoid becoming an empty shell that the signatories will end up abandoning in order to resume their hawkish positions.

Could improving the agreement’s implementation help solve the problem of jihadist insurrections spreading across other parts of northern Mali?

Some international actors and the Malian state consider that the reconstituted army, which must bring together Malian soldiers and combatants from armed groups, should engage in the fight against terrorism. It is risky, however, to connect the struggle against jihadist groups to the peace agreement’s implementation.

First, this idea gives the illusion that the signatory armed groups are capable of tackling jihadists. Many members of these signatory groups have been killed in the jihadists’ suicide bombings and other attacks; they are often forced to negotiate unofficial non-aggression pacts with the militant groups. Moreover, the “anti-terrorist” alliance created by Barkhane with two armed groups belonging to the Platform between 2017 and 2019 in the Liptako-Gourma region has proved unable to stem the jihadist expansion. Worse, it has exacerbated the situation by heightening local intercommunal tensions (see Crisis Group’s most recent report on Niger). The armed groups see no advantage in weakening their position in the anti-jihadist fight while the Malian state continues to raise the spectre of revising the peace agreement. Furthermore, most armed groups from the north have combatants in their ranks who were former members of jihadist groups before the French intervention, or else have family or tribal links to jihadist elements.

The issue of territorial and political autonomy is the core motivation for taking up arms in this region.

Although fighting terrorism attracts international attention, it is only one of the problems facing northern Mali today. Even if international and national forces were to succeed in eliminating or sidelining the jihadists, the signatory parties would still demand a satisfactory response to their demands for territorial autonomy in the north, which would almost certainly derail the Malian peace process. The issue of territorial and political autonomy – arising for the fourth time since 1963 – is the core motivation for taking up arms in this region. This is reflected in the agreement’s provisions on the implementation of effective regionalisation. In Niger, the state has allowed elites from the north, including former members of armed groups, to participate fully in running local administration. These elites have thus become better integrated into political and institutional affairs at a national level. Mali could follow this example that resolves a fundamental issue: how to dissuade people from joining armed groups and encourage military actors to take part in political and economic matters; even though it would be naïve to suppose that weapons and trafficking would disappear overnight. The most pressing goal is to ensure that these realities do not play into the hands of those with hawkish agendas.

How can the peace process move forward without jeopardising progress toward stability?

Expectations must be realistic. No one should feel satisfied with the current situation. At the same time, no one should exert pressure that may rekindle violence, for example by organising an unsuccessful referendum or redeploying the reconstituted army, which the signatory groups would judge as heavy-handed. The parties must take careful steps toward more effective implementation of the agreement. Given the various parties’ reluctance to apply the agreement in full, there is no magical solution for the problem. There are, however, two main areas where the peace process could gain new impetus: trust in the peace process, and political will to see it through.

Southern Malians’ opposition to the agreement has prevented progress toward its implementation. Since 2017, the government has postponed the deadline for the referendum on constitutional reform now scheduled for late 2020. This reform seeks to bring Mali’s constitution into line with the agreement’s terms, particularly by setting up a senate and regional assemblies whose presidents would be elected through direct universal suffrage. Opposition to the agreement, compounded by widespread discontent with a state weakened by seven years of crisis and recent disputed legislative elections, makes a positive outcome in such a referendum unlikely this year. Southern and central Malians account for almost 90 per cent of the electorate, and their mistrust of an agreement they do not properly understand would most likely lead them to reject the planned constitutional reform.

It is vital for southern Malians to give more support to the process.

It is therefore vital for southern Malians to give more support to the process through the political elites and civil society organisations supposed to represent them. They played no part in the discussions that led to the signing of the agreement in 2015, and many reject a text negotiated without their input. The 2015 text gave the Malian government the job of providing information and raising public awareness about the agreement’s content, but as the Carter Center observed, the government did little in this regard. There are now more public campaigns protesting against the peace agreement than in support of it. Awareness-raising initiatives have focused on northern populations, disregarding the fact that the agreement also applies to southern Mali, particularly through the regionalisation reform and the creation of a senate.

Five years after the signing of the agreement, it remains essential to address this shortcoming. Without the support of the population of southern Mali, many of its local interest groups will continue agitating to put the agreement on hold and to renegotiate its terms. Renegotiation is not in the interest of either the international community or the CMA, and over time could even lead to a resumption of belligerent discourse. The denunciation of the peace agreement is one of the grievances voiced by the organisers of the Movement of 5 June - Rally of Patriotic Forces (M5-RFP), a protest movement calling for the resignation of President Ibrahim Boubacar Keita that gathered tens of thousands of demonstrators on 5 and 19 June 2020, mostly in Bamako. Some of the movement’s leaders, such as filmmaker and former Minister Cheick Oumar Sissoko, have publicly called for the agreement to be revised, a position the M5-RFP has so far not adopted officially. The agreement remains a secondary issue for the movement, with other grievances aimed directly at President Keita taking precedence.

The CMA therefore needs to engage with southern Malians to explain that the agreement does not threaten to split up the country, and that regionalisation is a national reform and not limited to the north. The southern regions have everything to gain from a regionalisation process that would guarantee them a transfer of powers and resources unprecedented in Mali’s history. This awareness-raising could continue the work started with the inclusive national dialogue of 2019, namely the initiation of talks between the CMA and civil society organisations from southern Mali. Local elected representatives and traditional authorities from the north should be involved in these information campaigns in the southern regions. International partners sitting on the CSA monitoring committee, in particular MINUSMA, could help organise this work. Without guaranteeing the success of the referendum, such a move could still help relieve the pressure on the government exerted by southern elites that is holding up the agreement’s implementation.

The political authority in charge of implementing the agreement needs to be invested with greater power. The country’s president or, failing that, the prime minister, should become directly involved and support this authority, since these figures are the only ones able to give orders to the technical ministries and to resolve any disputes. The creation in 2016 of the president’s high representative to implement the agreement was a step in the right direction, but the person chosen for this job never had the necessary political clout or support to impose his views on a government that often remains unwilling to implement the agreement. The ministry of social cohesion, peace and national reconciliation, currently the government body in charge of this portfolio, has had no more success.

The key to implementation lies with the signatories themselves.

The top-level authorities of the signatory groups should be a more regular presence in Bamako, especially during the CSA’s most important sessions, since these constitute the main dialogue framework among the signatories. Otherwise, second-tier actors represent the groups, and their decisions fail to influence the other movements.

The international community must also continue to monitor progress, and to press for more, even though the current situation reveals the clear limitations of an externally imposed peace. The key to implementation lies with the signatories themselves.

The reality, however, is that Mali’s president must commit himself decisively and publicly to support the most sensitive provisions of the agreement – particularly the transfer of resources and power in terms of regionalisation and a reconstituted army. As long as he does not do so, the parties’ lack of will to implement the agreement will prove an insurmountable barrier.