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
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.

Fire and smoke billowing from Norwegian owned Front Altair tanker said to have been attacked in the waters of the Gulf of Oman, June 2019. ISNA / AFP

The Middle East between Collective Security and Collective Breakdown

For years, Gulf powers have mulled the notion of regional dialogue to calm existing crises and head off new ones. Today, with several active Middle Eastern conflicts, all sensitive to rising U.S.-Iran tensions, it is an idea whose time has come.

  • Share
  • Save
  • Print
  • Download PDF Full Report

What’s new?* Middle East tensions spiked in the past year following attacks on oil tankers and Saudi oil facilities, the U.S. killing of a senior Iranian commander and Iranian military retaliation. Some of Washington’s allies, losing confidence the U.S. will reliably extend military protection, have started making cautious diplomatic overtures to Iran.

Why does it matter? While these tentative steps toward de-escalation are welcome, they risk being inadequate, particularly in the absence of regular, high-level communication channels among potential conflict actors. Existing UN-led mechanisms for resolving individual conflicts, such as Yemen, are worthwhile but insufficient to lessen region-wide tensions.

What should be done? Diplomatic efforts are needed to both de-escalate tensions and make progress toward resolving regional conflicts. Gulf actors, supported by external stakeholders, should consider launching an inclusive sub-regional dialogue aimed at reducing the risk of inadvertent conflict by opening new communication channels.

* Crisis Group conducted the fieldwork for this Report before the COVID-19 pandemic. Some dynamics examined in this publication may have changed in the meantime. Moving forward, we will be factoring the impact of the pandemic into our research and recommendations, as well as offering dedicated coverage of how the outbreak is affecting conflicts around the world.

Executive Summary

The Middle East is, arguably, in as dangerous a condition as it has been in its modern history. A single incident could spark an escalation, which – uncontrolled – could set off a chain reaction of violent confrontations, involving local, regional and extra-regional powers. Established mechanisms for bringing individual conflicts, such as the wars in Syria and Yemen, to a peaceful resolution are making only halting, if any, progress. When a crisis of this magnitude crests, but before it erupts into full-blown war, the attention it attracts can create new opportunities for preventive action. The notion of a collective and inclusive security dialogue that aims to diminish tensions has been around for many years, focused on the Gulf sub-region. The time to launch one is overdue. The first step is to produce concrete ideas and international support for such a dialogue, which can open new channels of communication. To maximise chances of success, the effort should start modestly, possibly initiated by smaller Gulf states with the active diplomatic backing of a group of European and other governments.

Established mechanisms for bringing individual conflicts to a peaceful resolution are making only halting, if any, progress.

Underlying current tensions is the four-decade rivalry between the U.S. and Iran. The Trump administration’s withdrawal from the 2015 Iran nuclear deal and launch of a “maximum pressure” campaign of economic sanctions against Iran in 2018 already has given rise to several violent incidents. These include attacks on shipping in the Strait of Hormuz and Gulf of Oman in May and June of 2019; Iran’s 20 June downing of a U.S. surveillance drone that may or may not have entered Iranian airspace; the large-scale missile and rocket attack on Aramco installations in Saudi Arabia on 14 September; the 3 January 2020 killing, in a U.S. drone strike, of Qassem Soleimani, commander of Iran’s Qods Force – a special forces unit of the Islamic Revolutionary Guard Corps – and Abu Mahdi al-Muhandis, deputy head of Iran-backed paramilitary forces in Iraq; and a tandem of tit-for-tat attacks in Iraq between an Iran-backed paramilitary group and U.S. forces on 11 and 12 March.

Any of these incidents could have triggered commensurate retaliation, setting off a dangerous spiral. Reportedly, President Donald Trump was minutes away from responding to the drone incident in June 2019 by ordering an airstrike on Iranian military assets. Many in his administration also were pushing for retaliation after the Aramco attack. Iran retaliated for the Soleimani killing by firing missiles at Iraqi military bases on which U.S. troops were co-located, yet because no American was killed and because Tehran immediately communicated that this was the extent of its response, the U.S. did not escalate further.

This and other rivalries – between Iran and its allies on one hand and Israel on the other, and between Iran and Saudi Arabia – are colouring ongoing conflicts in the Middle East, turning civil wars into proxy wars. These wars increasingly intersect, if not on the ground, then certainly in the way outside actors view them. The wars in Syria and Yemen, in particular, illustrate this dynamic, but so, too, do less violent events in Iraq and Lebanon, where Iran has cast popular uprisings in pursuit of local objectives as instigated by the U.S., while Washington has characterised the protests as principally directed against Iran. Even the outbreak of COVID-19 triggered a spasm of mutual recriminations between Tehran and Washington – when humanitarian cooperation could provide a face-saving way to reach across the aisle and ease hostilities.

UN-led negotiations have at times made some progress in containing individual conflicts, notably in Yemen. But they have yet to resolve them. These conflicts’ perpetuation increases the odds that they metastasise, generate fresh grievances and embolden adversaries to engage in ever riskier games of brinkmanship, testing each other’s red lines. Worse, these red lines appear blurred in the absence of functioning communication channels. Poor communication in turn increases the likelihood of accidental conflict – a war the principal actors say they do not seek but into which they may sleepwalk.

A more effective process would approach the region more comprehensively and its individual conflicts as interlinked.

Today’s crises in the Middle East call for a new approach, one that addresses the weaknesses of the solutions presently on offer and takes into account the situation’s growing complexity. The region has undergone a dramatic shift since the 2011 Arab uprisings, and any attempt to tamp down its turbulence likely will require an analytical and operational paradigm shift. A more effective process would approach the region more comprehensively and its individual conflicts as interlinked; even as it addresses ways to reduce immediate dangers, it also would offer a vision for, and path toward, an eventual collective and inclusive regional security arrangement.

By mid-2019, governments in fact began floating initiatives to tackle the burgeoning region-wide crisis, responding to fears that the threat of a major war had become acute. But so far none has taken decisive steps toward kick-starting a collective and inclusive process aimed at reducing tensions. Whatever has happened so far has been ad hoc and uncoordinated – or, worse, based on the construction of hostile alliances that feed, not fight, regional escalation, such as the Middle East Strategic Alliance and the U.S.-led Middle East conference in Warsaw in February 2019. President Emmanuel Macron of France has tried to prevent the worst by nudging Iran and the U.S. back to the negotiating table, but his efforts have yet to bear fruit. France has also sought to lower tensions in the Strait of Hormuz, through which flows a good proportion of the world’s oil, through a maritime security initiative. Track 1.5 and Track 2 dialogues involving Saudi and Iranian participants have aimed to open new channels of communication. Yet the region continues to teeter on a knife edge.

Governments in the Gulf sub-region that are least involved in hostilities but could be harmed the most if fighting erupts, notably Kuwait and Oman, should consider jointly taking the initiative to bring their more powerful and more directly involved neighbours – Iran and Saudi Arabia – into an informal effort to lessen tensions. Such an exercise, in the form of an inclusive security dialogue, ought to be based on core principles all can accept; focus on a combination of hard- and soft-power concerns they all share; and start modestly. This beginning alone could further open communication channels whose limited nature thus far has raised the risk of conflict through miscalculation. If successful, it could also by extension serve to ease tensions between the primary adversaries: the U.S. and Iran.

Any such regional dialogue almost certainly will not come about without external support.

Any such regional dialogue almost certainly will not come about without external support, especially from the U.S. Yet as long as Washington remains intent on pursuing its “maximum pressure” campaign against Iran, it is unlikely to back any initiative that involves engaging Tehran. Governments in Europe and elsewhere, however, whose security interests would be harmed by a wider conflagration but whose policies in the region Gulf states might not perceive as threatening could take the lead in preparing the ground. They might begin by reaching out to Gulf states to explore their readiness for an inclusive security dialogue, drawing inspiration from the successful negotiations that led to the 1975 Helsinki Accords during the Cold War.

The time to lay the groundwork for such an initiative may well have come. To generate political will to act, the worst of times may offer the best opportunity, and conditions in the Gulf arguably have reached that point. Although negative developments may yet occur that seemingly undermine prospects for such a collective effort, they should not be reason to give up. Indeed, while the U.S. and Iran engaged in mudslinging over the pandemic, some Gulf states opted for a different course, providing assistance to an Iran battling an acute health crisis. Because of such unpredictable events, any inclusive process should be of long duration, be immune to the day’s breaking news, and continue to move toward a gradual lessening of tensions through dialogue and steadily widening communication channels regardless of potential – indeed, inevitable – setbacks.

 Riyadh/Abu Dhabi/Tehran/Muscat/Doha/Brussels, 27 April 2020

Averting War in the Gulf Needs Regional Dialogue


I. Introduction

No major player in the Middle East appears to want to start a war with its adversaries, but events in 2019 have showcased risks of an inadvertent conflict. The danger is commensurate with tensions arising from competition between the U.S. and Iran, heightened by the U.S. withdrawal from the 2015 nuclear deal and reimposition of economic sanctions, as well as between Israel and Iran and between Iran and Saudi Arabia. The 2019 attacks on shipping in the Strait of Hormuz and the Gulf of Oman, as well as on oil pipelines in Saudi Arabia; the shooting down of a U.S. drone by Iranian fire; the devastating strikes on Saudi Arabian oil and gas facilities at Abqaiq and Khurais in September; and the 3 January 2020 U.S. drone strike that killed Qassem Soleimani, a top Islamic Revolutionary Guard Corps commander, and Abu Mahdi al-Muhandis, a senior commander of the Iran-backed Hashd al-Shaabi paramilitary groups in Iraq – all these incidents brought simmering tensions to a hard boil.

The fallout from these events was contained, a major reason being President Donald Trump’s reluctance to fight another war and Iran’s reluctance to take on a more powerful adversary directly, but there is no guarantee that this won’t change. With so many actors and conflict drivers in play, the margin of volatility and unpredictability is wide.

With so many actors and conflict drivers in play, the margin of volatility and unpredictability is wide.

There are lessons to be learned from the past. At the height of the Cold War, the world found ways of lowering risks of confrontation. Differences with today’s Middle East are vast, but the idea that took hold then – that a collective effort, inclusive of all the major powers on both sides of the divide, could reduce tensions through dialogue and the establishment of confidence-building mechanisms – is relevant nonetheless. That effort, embodied in the Conference on Security and Cooperation in Europe (CSCE), came to fruition in the 1975 Helsinki Accords, which preserved the peace and gave rise to a collective mechanism that would ensure that it stayed that way, the Organization for Security and Co-operation in Europe (OSCE).

The memory of that success has remained alive through a number of exercises, often led by non-governmental organisations and think-tanks, with a focus on other unstable parts of the world, including the Middle East. One example of this last genre is a report published by the Stockholm International Peace Research Institute in 2011, outlining a collective process.[fn]“Toward a Regional Security Regime for the Middle East: Issues and Options”, Stockholm International Peace Research Institute, October 2011. Other examples include Christian-P. Hanelt and Christian Koch, “A Gulf CSC Could Bring Peace and Greater Security to the Middle East”, Bertelsmann Stiftung, July 2015; Frederic Wehrey and Richard Sokolsky, “Imagining a New Security Order in the Persian Gulf”, Carnegie Endowment for International Peace, 14 October 2015; and Dalia Dassa Kaye, “Can It Happen Here? Prospects for Regional Security Cooperation in the Middle East”, The Century Foundation, 18 January 2018. For current initiatives, see Christian Koch and Adnan Tabatabai, “Tafahum: An Ideational Fundament on Which to Build a Security Roadmap for West Asia and the Arabian Peninsula”, CARPO and the Gulf Research Center Foundation, 17 July 2019; and Dina Esfandiary, “Bridging the Divide between Iran and the Arabian Peninsula”, The Century Foundation, 11 March 2019.Hide Footnote In recent years, think-tanks have organised workshops about the need for a Middle Eastern security architecture, at times drawing in expertise from the Cold War era.

Prompted by tensions in the Gulf, in 2019 governments began to take an active interest in the need for, and possibility of, a similar collective exercise. In July, Russia floated its idea of a security concept for the Gulf region and organised a workshop in Moscow in September to flesh it out.[fn]Russia’s Security Concept for the Gulf Area”, Ministry of Foreign Affairs of the Russian Federation, 23 July 2019. The proposal envisions “a single counter-terrorism coalition” that would embark on “a long-term programme of action” in pursuit of “a security system in the Gulf area”. It is based on a number of sensible principles. But it then posits ambitiously that: “Progress toward the establishment of a security system should be achieved on a step-by-step basis starting with most relevant and urgent problems. This concerns, first and foremost, combat against international terrorism, the settlement of the Iraqi, Yemeni and Syrian crises, and the implementation of all agreements reached on the Iranian nuclear programme”.Hide Footnote Iran laid out similar ideas in its Hormuz Peace Endeavour, launched at the UN General Assembly in late September, and invited Arab Gulf states to join it in a collective exercise.[fn]“Rouhani’s UN address to focus on JCPOA, sanctions, regional security”, Mehr News, 25 September 2019. Foreign Minister Javad Zarif summarised the initiative in an article in the Arab media; the English translation is available on the Iranian foreign ministry’s website. Iran has long argued for a collective security dialogue in the Gulf, backed by the international community, basing itself on UN Security Council Resolution 598 (1987), which laid the groundwork for the ceasefire that would end the Iran-Iraq war almost a year later. Paragraph 8 states that the Security Council “[f]urther requests the Secretary-General to examine, in consultation with Iran and Iraq and with other States of the region, measures to enhance the security and stability of the region”. At the Manama Dialogue in 2004, Iran floated a plan for a “Persian Gulf collective security framework” (International Institute for Strategic Studies, A Decade of the IISS Manama Dialogue, 2014, p. 14), and at the 2007 World Economic Forum it offered a ten-point proposal for promoting cooperation, security and development in the Persian Gulf region (Kaveh L. Afrasiabi, “Iran unveils a Persian Gulf security plan”, World Bulletin, 3 September 2018).Hide Footnote Shortly thereafter, the Iraqi foreign minister proposed convening a regional security conference focused on Iraq, contacting his counterparts in Turkey, Iran, Saudi Arabia, Jordan and Kuwait to press the idea.[fn]Crisis Group interview, senior Iraqi official, February 2020.Hide Footnote Baghdad’s leadership made sense, as Iraq is the seam of the Middle East, straddling the Sunni-Shiite divide and forming a bridge between Arabs and Persians. While the idea received EU backing, it was stillborn – at this writing, Baghdad was still awaiting Riyadh’s response.[fn]“Remarks by High Representative/Vice-President Federica Mogherini at joint press event with the Foreign Minister of Iraq”, European External Action Service, 13 July 2019.Hide Footnote Moreover, it was soon overtaken by mass protests in Iraq that paralysed and eventually brought down a government that was dysfunctional even in more propitious times.[fn]Iraqi politicians were still trying to form a government in April 2020. See also Alissa J. Rubin, “Oil prices crash, virus hits, commerce stops: Iraq is in trouble”, The New York Times, 29 March 2020.Hide Footnote

Optimally, states on the Gulf littoral would help initiate a collective process inclusive of all stakeholders.

These initiatives underline both the need for and timeliness of a concerted effort to reduce tensions in the Gulf. Optimally, states on the Gulf littoral would help initiate a collective process inclusive of all stakeholders. The Middle East and its various parts arguably has become ripe for collective security exercises that aim to reduce tensions. This report focuses on the Gulf sub-region because, while the Middle East as a whole has become increasingly interconnected, chances of success likely would be higher if an initiative were to start small and test ideas ahead of launching a broader process.

This report began as a Crisis Group presentation to the UN Security Council on 21 March 2019 (an informal interactive dialogue hosted by France), followed by a 24 May conversation in Crisis Group’s New York office, in which diplomats from a range of Middle Eastern states participated, focused on the need for a collective security dialogue in the Middle East, starting in the Gulf sub-region. With feedback from those discussions, and building on earlier Crisis Group work on intersecting conflicts, we produced a White Paper in July that brought these ideas together and which we distributed to selected governments.[fn]The previous work is published as International Crisis Group, Tackling the MENA Region’s Intersecting Conflicts, 22 December 2017.Hide Footnote This paper triggered further meetings with a number of regional and other stakeholders to gauge receptivity to the idea and assess whether the timing was right to launch an initiative. The present report reflects these many conversations.

A timeline of events relating to security and cooperation in the Middle East since 1979. 

II. A Dual Problem in Search of a Single New Approach

In tackling the proliferation of conflicts, conflict drivers and actors in the Middle East, a divided international community faces two primary challenges. One is the situation’s growing complexity, as conflicts increasingly intersect; the other is the limited tools it has at its disposal for de-escalating and containing them.

A. An Increasingly Complex Tangle of Conflicts

The 2011 Arab uprisings channelled popular demands for greater political participation, better governance and social justice. When authorities failed to deliver on these demands, and in several cases responded ruthlessly, the stage was set for more intense upheavals. At the same time, these were externalised, leading to far greater levels of violence. The most dramatic example is Syria, where an embattled regime used extreme brutality to suppress popular protests. The regime’s response gave rise to a civil war, which attracted outside intervention, turning it into a regional proxy war.[fn]See, for example, the following Crisis Group Middle East Reports and Briefings: Briefing N°47, Russia’s Choice in Syria, 30 March 2016; Report N°155, Rigged Cars and Barrel Bombs: Aleppo and the State of the Syrian War, 9 September 2014; Report N°143, Syria’s Metastasising Conflicts, 27 June 2013; Briefing N°128, Syria’s Mutating Conflict, 1 August 2012; and Report N°109, The Syrian Regime’s Slow-Motion Suicide, 13 July 2011.Hide Footnote In Yemen, an autocratic regime collapsed, but so did a subsequent effort to establish a new government through dialogue, opening the way to civil war and external military intervention.[fn]See, for example, the following Crisis Group Middle East Reports and Briefings: Report N°203, Saving the Stockholm Agreement and Averting a Regional Conflagration in Yemen, 18 July 2019; Report N°167, Yemen: Is Peace Possible?, 9 February 2016; Briefing N°45, Yemen at War, 28 March 2015; Report N°154, The Huthis: From Saada to Sanaa, 10 June 2014; Report N°145, Yemen’s Southern Question: Avoiding a Breakdown, 25 September 2013; and Report N°102, Yemen between Reform and Revolution, 10 March 2011.Hide Footnote

What happened in Deraa or Deir al-Zour, Sanaa or Saada affected state interests in Iran and Saudi Arabia, Turkey and the United Arab Emirates, to name but a few. The responses formulated in regional capitals inflected these conflicts as they unfolded. Washington and Moscow were drawn in when they saw their interests come under threat or detected opportunities to advance them. The result has been a web of overlapping responses to an interlocking set of conflicts involving multiple actors and cross-cutting alliances, compounded by outside military intervention. The array of players includes states weakened to the point of looking like non-state actors, and several potent non-state actors, some of which have assumed the trappings of sovereignty within a given territory, acting as if they are states. None holds a monopoly on the means of violence within established borders.[fn]See Crisis Group, Tackling the MENA Region’s Intersecting Conflicts, op. cit.Hide Footnote The interplay among various conflicts, both on the ground and in stakeholders’ perceptions, makes individual conflicts harder to address, and heightens the risk that external diplomacy and assistance will have adverse unintended consequences.

B. A Breakdown in Workable Conflict Resolution Tools

Post-2011 UN-led initiatives in Syria, Yemen or Libya have fallen short.

Efforts to bring Middle East conflicts to a peaceful conclusion have tended to fizzle, notwithstanding intensive diplomacy in some cases. Decades of failed Israeli-Palestinian peacemaking are usually mentioned in this regard, but post-2011 UN-led initiatives in Syria, Yemen or Libya also have fallen short. They have proceeded in fits and starts, and whatever progress they have made has often been undone by renewed military escalation by one of multiple actors with a stake in the conflict’s outcome.

A number of factors have contributed to the growing ineffectiveness of international tools available for conflict prevention and resolution:

A great-power standoff. The rise of a multipolar world has produced a great-power standoff that has politicised multilateral institutions.

  • Dysfunction marks the UN Security Council, much as it did during long stretches of the Cold War.[fn]See Richard Gowan, “Navigating the Storms at the UN Security Council”, Crisis Group Commentary, 5 February 2020. On UN dynamics regarding the Middle East during the Cold War, see Bruce D. Jones, “The Security Council and the Arab-Israeli Wars: ‘Responsibility Without Power’”, in Vaughan Lowe, Adam Roberts, Jennifer Welsh and Dominik Zaum (eds.), The Security Council and War: The Evolution of Thought and Practice Since 1945 (Oxford, 2008).Hide Footnote As global and regional powers disagree fiercely about individual conflicts, they have not only not thrown their weight behind UN peace processes but the P5 (and others) also have used the Security Council for a decade as space to defend their allies in the Middle East and control or limit diplomacy rather than look for real solutions (Yemen being a partial exception). The result is Security Council inaction or watered-down resolutions. An individual superpower’s support for one side in a conflict, as in the case of the U.S. in Yemen and Russia in Syria, has given the beneficiary a sense of impunity and even leverage vis-à-vis its sponsor when the latter wants to moderate its proxy’s behaviour, because the sponsor cannot afford to let its proxy fail. UN mediation efforts in the region are largely meant to cover up for lack of P5 unity, and the UN’s role often is to buy everyone time and mitigate violence rather than resolve it.
  • The standoff among great powers has politicised and undermined previously functioning institutions and agencies. One example is the Joint Investigative Mechanism (JIM), a partnership between the Organisation for the Prohibition of Chemical Weapons and the UN established by UN Security Council Resolution 2235 (2015). Russia’s refusal to accept the JIM’s November 2017 conclusion that the Syrian regime was responsible for the 4 April 2017 chemical attack on the Syrian town of Khan Sheikhoun led it to use its Security Council veto to block the renewal of the JIM’s mandate in November 2017.[fn]See UN Security Council, “Security Council Fails to Renew Mandate of Joint Investigative Mechanism on Chemical Weapons Use in Syria, as Permanent Member Casts Veto”, 24 October 2017. (The vote was in October 2017, just ahead of the official release of the JIM’s report on Khan Sheikhoun in November. It took effect later in November when the JIM’s mandate expired.)Hide Footnote

The absence of a powerful arbiter. At this stage, no country with real clout is in a position to play the role of arbiter able and prepared to pressure conflict actors when negotiations reach a stalemate, provide security guarantees when they contemplate a ceasefire or peace agreement, and dedicate significant resources to resolving the conflict.

  • Arguably, the U.S. has played this part at times. But more recently, it has taken a step back with regards to diplomatic efforts, taken sides even more clearly than in the past (in the case of Iran with its withdrawal from the nuclear deal and imposition of “maximum pressure”, and in the case of the Israeli-Palestinian conflict by rewriting the terms of a potential settlement), or sent confusing messages about its goals (in Syria and Libya in particular). While it has somewhat stepped up its involvement in some conflicts, such as Yemen, it has not acted decisively to help end them.
Moscow has positioned itself as a player in the wider Middle East.
  • Russia has used its military power to shape diplomatic settlements, but has focused its efforts on shoring up one side against the other rather than seeking genuine compromise.[fn]Syria and Libya are the two most prominent examples. As illustrated most recently in Idlib, Russia appears intent on helping the regime restore its writ over the entirety of the country through a mix of violence and short-lived agreements. See, for example, Crisis Group Commentary, “Deadly Clashes in Syria’s Idlib Show Limits of Turkey’s Options”, 29 February 2020. In Libya in 2019-2020, operatives of the Russian private security company Wagner Group played a role in Field Marshal Khalifa Haftar’s military offensive against the government in Tripoli even as Moscow ostensibly participated in the UN-led process to end the war. See Crisis Group Statement, “Libya: Turning the Berlin Conference’s Words into Action”, 22 January 2020; and Crisis Group Commentary, “What Prospects for a Ceasefire in Libya?”, 18 January 2020. On the Wagner Group’s involvement in Libya: Crisis Group interviews, military officers, Benghazi, May and October 2019; government officials, Tripoli, September-December 2019. See also "Putin-linked mercenaries are fighting on Libya’s front lines”, Bloomberg, 25 September 2019; and David Kirkpatrick, “Russian snipers, missiles and warplanes try to tilt Libyan war”, The New York Times, 5 November 2019.Hide Footnote Moscow has positioned itself as a player in the wider Middle East, fostering good relations with an impressive array of parties, including those in conflict with one another – including Israel, Iran, Saudi Arabia, Syria, Turkey and Egypt. While its relations facilitate communication and talks (for example, on both Syria and Libya), Moscow so far has been either unable or unwilling to translate them into meaningful steps to sustainably end conflicts. In the Syrian case, its support for the regime is driven in part by its belief that Damascus regaining full control would produce the most stable outcome, deter future uprisings and mark a significant Russian success in the region.
  • The EU, the global actor that arguably has felt the impact of Middle East wars most directly, has been unable to insert itself diplomatically into the region in a manner that can help mediate an end to its conflicts. It is hobbled by internal divisions, for example on Libya, and its potential for mediation is complicated by member states selling weapons to one side, as in the Yemen war. It is also weakened in its resolve by internal political turbulence and the divorce with Britain.[fn]The EU’s high representative for foreign affairs and security policy, Josep Borrell, showed awareness of the problem. In a February 2020 commentary, he said Europeans “must deal with the world as it is, not as they wish it to be. And that means relearning the language of power and combining the European Union’s resources in a way that maximizes their geopolitical impact”. Josep Borrell, “Embracing Europe’s Power”, Project Syndicate, 8 February 2020.Hide Footnote
  • For now, China appears uninterested in playing a political role in the Middle East, seemingly content to see its global rivals weakened and expending resources through their own involvement. Its principal interests arguably are in developing economic and trade relations and ensuring safe access to Gulf oil; the U.S.’s strong presence in the area so far has sufficed to achieve the latter goal.[fn]The Chinese government published its first “Arab policy paper” in 2016. The paper shows a multi-dimensional approach to the region, heavily tilted toward development cooperation and trade (as part of Beijing’s Belt and Road Initiative). “Full Text of China’s Arab Policy Paper”, Xinhua, 13 January 2016. See also Andrew Scobell and Alireza Nader, “China in the Middle East: The Wary Dragon”, Rand Corporation, 2016, which argues that China’s involvement in the Middle East is “driven primarily by economic interests” and that “Beijing is very reluctant to expand its level of security cooperation with the United States or Middle East states because it fears being embroiled in regional tensions and controversies” (p. x).Hide Footnote

Weak regional frameworks. In the Arab and Islamic worlds, regional mechanisms also function far below par. The Arab League (which excludes key Middle Eastern actors such as Iran, Israel and Turkey) is so deeply politicised and polarised as to foreclose an effective role in resolving most conflicts. The Organisation of Islamic Cooperation (which is more broadly inclusive of Middle East states except for Israel) has not developed strong conflict mediation capabilities. Even regional alliances, such as the Gulf Cooperation Council, have found it difficult to fashion a common position on the region’s conflicts, being internally divided and prone to exporting their competition to external arenas.[fn]See Crisis Group Middle East Report N°206, Intra-Gulf Competition in Africa’s Horn: Lessening the Impact, 19 September 2019.Hide Footnote

Arms transfers undermining peace efforts. Arms-producing states selling weaponry to one or more of the warring sides help fuel wars they profess to want diplomacy to end. The Yemen war and the humanitarian catastrophe it has wrought stand as among the most striking examples.[fn]In Libya, too, states professing to be supporting a negotiated solution to the conflict have fanned the flames by supplying one or the other side with weapons and other forms of military support.Hide Footnote

Non-inclusive peace negotiations. Several peace efforts have been hobbled by the exclusion on political grounds of one or more important stakeholders. The January 2014 decision by UN Secretary-General Ban Ki-moon to withdraw his invitation to Iran to participate in the Syria peace talks in Geneva (“Geneva II”), despite its prominent role in the Syria war, almost certainly contributed to the ineffectiveness of those talks. Likewise, Turkish pressure led to the exclusion of the Kurdistan Workers’ Party (PKK)-affiliated People’s Protection Units (YPG) from the Geneva talks, a decision at odds with the militia’s important role in Syria’s north east. Excluded parties are likely to both hinder achievement of a peace deal and seek to spoil any eventual agreement.

The weakening of Arab states and proliferation of armed non-state actors is also complicating stabilisation and peacemaking efforts.

Political fragmentation. The weakening of Arab states and proliferation of armed non-state actors – both a result of and a contributor to political fragmentation – is also complicating stabilisation and peacemaking efforts. In Yemen, for example, the fracturing of the government into multiple competing groups has hobbled its ability to form a common front against their shared enemy, the Huthi movement (Ansar Allah), either to fight more effectively or to forge a durable peace. In Iraq, the rise of Iran-backed paramilitary groups is preventing a dysfunctional state from establishing monopoly over the means of violence and thereby stabilising a country traumatised by repeated upheavals. In Palestine, a deep rift between the two principal political movements, Fatah and Hamas, has undermined the Palestinians’ negotiating position vis-à-vis Israel and accentuated the geographic division between the West Bank and Gaza.

III. Steps toward a Collective and Inclusive Regional Security Dialogue

A. The Need for a Comprehensive Approach

The growing complexity of the Middle East’s intertwined conflicts, coupled with the inadequacy of the tools the international community has employed to address them, make a powerful case for a new approach. One possibility would be to shift toward initiatives by “coalitions of the willing/like-minded” or “groups of friends” that enjoy the support of larger institutions such as the UN, EU, Arab League and Gulf Cooperation Council (GCC). A key would be to view and deal with the Middle East’s conflicts as interconnected, the ignition of any part of which could produce a more general conflagration.

Such an approach would need to take into account how most of the region’s wars started locally and radiated outward, drawing in regional and even global actors in concentric circles; Syria and Yemen, in particular, followed this pattern. One way to end the wars would be to reverse the process. If Russia and the U.S. were to reach a shared understanding of how to resolve the Syrian conflict, they could more plausibly constrain the actions of local and regional actors; likewise, if Iran and Saudi Arabia were to stop fuelling the conflict in Yemen and emboldening their respective local allies, the odds of an intra-Yemeni settlement might get better. The end of negative outside interference would likely not be a sufficient condition for successful conflict resolution, but it is almost certainly a prerequisite. The reality of outside interference does not mean that parties ought to postpone efforts to resolve local conflicts until global and regional powers find ways to accommodate their competing interests. They can lay important groundwork in the meantime. Local actors could even shape the agenda toward a solution.

B. What a Region-wide De-escalation Process Could Look Like

1. Start of a dialogue in the Gulf sub-region

A sensible way forward would be to design a process that helps stakeholders minimise risks of an inadvertent outbreak of hostilities through a broadly inclusive dialogue aimed at de-escalating regional tensions. Such a process ought to be incremental, starting in the Gulf sub-region and broadening to include other Middle East actors as the dialogue progresses. States external to the Middle East should be invited from the outset to provide support.

An inclusive dialogue could begin in the Gulf sub-region, where tensions arguably are highest and pose the greatest risk of escalating to all-out war.

An inclusive dialogue could begin in a relatively limited fashion, namely in the Gulf sub-region, involving the six GCC members (Saudi Arabia, the United Arab Emirates (UAE), Qatar, Oman, Kuwait and Bahrain) along with Iran and Iraq – all principal actors in the Gulf, where tensions arguably are highest and pose the greatest risk of escalating to all-out war. While the main antagonists most likely to trigger a war are the U.S. and Iran, a successful dialogue between the two, however desirable, still seems far off. In the meantime, because these two powers maintain webs of alliances in the Middle East, lessening tensions among Gulf actors could reduce the likelihood of a U.S.-Iran conflict and start a process of broader de-escalation.

Iran has already floated a proposal for an inclusive Gulf dialogue – its Hormuz Peace Endeavour.[fn]See fn 3 above. The Iranian leadership sent a message to the six GCC governments in October 2019 inviting them to join its Hormuz Peace Endeavour; only Kuwait, Qatar and Oman replied positively or acknowledged receipt; Saudi Arabia, the UAE and Bahrain refrained from responding. “Rouhani sends letter on Iran’s Hormuz Peace Endeavor to Arab leaders”, Mehr News, 2 November 2019. The text of Iran’s initiative states that all sides “need to commit to respect each other’s sovereignty and territorial integrity, the inviolability of our international borders and the peaceful settlement of disputes” and “should categorically reject any threat or use of force or participation in coalitions against each other”. Unmentioned is Iran’s use of non-state proxies, which has fuelled its adversaries’ suspicion that it is seeking to establish regional hegemony.Hide Footnote Tehran is not the best placed to initiate or lead a Gulf dialogue, because it is a direct party to Middle East conflicts, not a neutral mediator. It would be better if smaller Gulf states with a relatively neutral profile – namely Kuwait and Oman – could take the initiative. They might issue invitations to a conference to discuss ways to increase mutual understanding on a broad range of issues of common concern, especially in the security field. There is historical precedent for this approach: in the CSCE process that culminated in the 1975 Helsinki Accords, the initial invitation came from Finland, even though organising a conference across the East-West divide was originally a Soviet idea, and most of the talks took place in Switzerland.[fn]See Michael Cotey Morgan, The Final Act: The Helsinki Accords and the Transformation of the Cold War (Princeton, 2018), pp. 75, 87 and 108-109.Hide Footnote

To varying degrees, GCC members have expressed alarm at developments in their neighbourhood and desire for mechanisms that would allay risks of war. Qatar has been explicit about it.[fn]The Qatari foreign minister told the European Parliament that his country favours a binding collective security agreement in the Middle East that “must encompass all countries in the region for peace to be effective”; it should also “be based on agreed principles of security, developed with rules of governance, dispute resolution and accountability, respectful of sovereignty and equality among its members, and a commitment of non-interference in [other states’] internal affairs”. “Remarks by Qatari Foreign Minister Mohammed bin Abdulrahman bin Jassim Al-Thani”, European Parliament, Committee on Foreign Affairs, Brussels, 19 February 2020. Hide Footnote The UAE, and possibly Saudi Arabia as well, also put out feelers to Iran in the second half of 2019.[fn]UAE security officials have visited Iran several times since the June 2019 escalation in the Gulf, and Iranian security officials have visited the UAE as well in response. Crisis Group interviews, Iranian and Gulf officials, August-December 2019. A report in The New York Times suggested that: “The Emiratis began their secret talks with Iran after concluding that they could play a unique role lowering temperatures and that they had little confidence in the Trump administration’s approach to Iran, according to American and other Western officials. … The Saudis also explored a diplomatic breakthrough with Iran using Iraqi and Pakistani intermediaries”. Mark Mazzetti, Ronen Bergman and Farnaz Fassihi, “How months of miscalculation led the U.S. and Iran to the brink of war”, The New York Times, 13 February 2020.Hide Footnote Some Gulf officials even hint that they want to develop a security policy less dependent on the U.S., realising that Washington may not always be willing to protect them.[fn]Crisis Group interview, Gulf academic, January 2020.Hide Footnote Yet officials from key Gulf Arab states also have made clear that their governments will not initiate or participate in such a collective exercise including Iran without Washington backing them. Fear of alienating the U.S. takes precedence in their strategic thinking, and they are wary of any initiative that does not enjoy a minimum level of U.S. support, as well as agreement among GCC members.[fn]Crisis Group interviews, Gulf officials, December 2019-March 2020.Hide Footnote Saudi Arabia has indicated that it will not engage in dialogue with Iran until Iran “changes its behaviour” in the region. A senior Saudi official put it most bluntly:

The Omanis and the Kuwaitis and Pakistanis tried to open channels between us and Iran, but we are not talking with the Iranians. Who should we talk to? Rouhani and Zarif have no authority. And the IRGC wants to kill you. … We do not need a back channel with Tehran. We need Iran to change its behaviour.[fn]Crisis Group interview, Riyadh, December 2019. As examples of how Iran could change its behaviour, the official mentioned a halt to providing weapons to the Huthis and Hizbollah, removing its operators from Yemen and encouraging a prisoner exchange between the Huthis and Saudi Arabia.Hide Footnote

Iran seeks to gain international recognition for its right to establish and lead a regional security regime.

The Gulf states certainly do not accept the notion of Iran taking the lead through, for example, its Hormuz Peace Endeavour. An Emirati official commented that even should GCC states agree to engage with Iran, they see the initiative as an unacceptable attempt by Tehran to isolate one part of the problem – regional security – from issues of global concern, such as Iran’s nuclear or ballistic missiles program. Instead, this official said, the challenge Iran poses should be addressed as a package.[fn]Crisis Group interview, Abu Dhabi, March 2020. Bahraini and Saudi officials expressed a similar sentiment. Crisis Group interviews, Bahraini official, Manama, March 2020; senior Saudi official, Riyadh, December 2019. The negotiations that led to the 2015 signing of the Joint Comprehensive Plan of Action – the Iran nuclear deal – covered only Iran’s nuclear program, not its ballistic missiles or regional power projection. The Trump administration’s 2018 unilateral withdrawal from the JCPOA and articulation of twelve demands on Iran to alter its regional and defence policies opened up the possibility, at least in theory, of new negotiations over a more-for-more deal. “After the Deal: A New Iran Strategy", U.S. State Department, 21 May 2018.Hide Footnote Ebtesam Al Ketbi, an Emirati academic with close ties to the UAE leadership, added:

What Iran really conceals in this proposal is the fact that it considers itself as the most powerful and pivotal element in such envisioned structure and the party that has the final say in it. In other words, Iran seeks to gain international recognition for its right to establish and lead a regional security regime, and to exercise veto power over it.[fn]Crisis Group email communication, Ebtesam Al Ketbi, president, Emirates Policy Centre, 17 February 2020.Hide Footnote

2. A U.S. role

U.S. tolerance, if not full support, of a dialogue process in the Gulf in which its allies would participate may well be a precondition for its successful launch. At the moment, this seems inconceivable. The Trump administration is engaged in anti-Iran alliance building and in coercive diplomacy in the form of financial and economic sanctions.[fn]In 2019, the U.S. created the Warsaw Process in collaboration with Poland in an undisguised attempt to muster international support for its effort to isolate, ostracise and sanction Iran. The Warsaw conference that kick-started the process took place on 13-14 February 2019. Whether it achieved its primary objective is a matter of opinion. See Katie Rogers, “At Trump forum, countries share a foe (Iran) and awkwardness (a lot of it)”, The New York Times, 14 February 2019.Hide Footnote To the extent that the potential for direct negotiations existed in 2019 – the two sides appeared to inch toward a meeting between their presidents in New York at the end of September – it vanished with the U.S. killing of Qassem Soleimani, commander of Iran’s Qods Force, in January.

During the remainder of 2020, Washington likely will continue to ramp up pressure on Iran in an attempt to topple its leadership or force it to the negotiating table on terms favourable to the U.S. Other, more dangerous scenarios are plausible, possibly triggered by further attacks by Iran and/or its allies in the region.[fn]A U.S. official said: “All our warnings are blinking red. We are convinced that Iran or its proxies will take action against us in the region in an effort to push the U.S. out of the Middle East”. Crisis Group interview, Washington, February 2020.Hide Footnote The next opportunity for diplomacy, in other words, likely will come after the November 2020 presidential election. A Democratic administration likely would directly re-engage Iran, but even a re-elected President Trump – bereft of a viable off-ramp and facing Tehran’s refusal to negotiate under pressure – could conceivably agree to preliminary dialogue between U.S. Gulf allies and Iran to test the possibility of regional de-escalation and pave the way for direct U.S.-Iran talks.

While the U.S. remains the most powerful actor in the Middle East and its role in any dialogue ultimately will be critical, it does not need to be fully involved from the outset. During much of the CSCE process, the U.S. belittled it as useless, while focusing on securing bilateral nuclear arms control deals with the Soviets. Yet it did not block its European allies from engaging in negotiations, and sent its own delegation to participate, which generally made a constructive contribution. Only toward the end of the process, when it showed signs of success, did Washington have a change of heart, seeing the exercise’s value in reducing tensions in Europe and thus shielding its allies from military confrontation. Once it did, its diplomatic push made the final accords a swift reality.[fn]In 1972, Henry Kissinger, in his capacity as U.S. national security adviser (1969-1975), warned President Richard Nixon of the “dangers of holding grandiose conferences”, instead favouring back-channel talks with Moscow to end the arms race and the export of communism. Morgan, op. cit., pp. 13, 60 and 88. Kissinger also served as U.S. secretary of state (1973-1977) during the latter part of the CSCE negotiations.Hide Footnote This example illustrates how an initiative could arise from the ground up – from regional states, including U.S. allies, backed by a group of outside powers.

3. Support of an external core group

The mutuality of interest is even clearer, and more pressing, with the outbreak of COVID-19.

Given the unlikelihood that Gulf Arab states will engage in an inclusive dialogue without U.S. backing or that a U.S. green light is forthcoming under current political conditions, alternatives are needed, at least until Washington’s policy shifts. Other external powers should use this period to start exploring ways in which they can assure Gulf states of international interest in, and support for, a collective exercise initiated by them and prepare the ground for the eventual launch of a dialogue. They could start by fashioning a group of like-minded states (henceforth referred to as the external “core group”).

Alarmed by developments in the Gulf in 2019, a number of European governments have already started discussions among themselves in an attempt to test the waters, building on prior work by non-governmental organisations and think-tanks.[fn]Crisis Group discussions with government officials in various European countries, July 2019-March 2020. Some of the NGO efforts are mentioned in fn 1 above.Hide Footnote They have a clear interest in helping lessen tensions in the Gulf through dialogue, not least because of the region’s importance to the global economy’s health. As Bahrain’s foreign minister noted in November 2019:

In securing the region, regional countries cannot go it alone. … [J]ust as the international community benefits from the region, it also shoulders its share of responsibilities towards it. Global powers are an integral part of the regional architecture and in an interconnected world, that will remain the case.[fn]Speech by Sheikh Khalid bin Ahmed Al Khalifa, Bahrain’s minister of foreign affairs, at the Manama Dialogue in Bahrain, 23 November 2019.Hide Footnote

The mutuality of interest is even clearer, and more pressing, with the outbreak of COVID-19, whose spread has left no country unaffected and threatens to dampen commercial activity, if not severely damage entire economies, including in the Middle East. Iran has been hardest hit, but Iraq and other countries – many plagued by an overstretched health care infrastructure, poor governance, corruption or other factors – are likely to follow. European governments, fearing a knock-on effect in the form of mass migration to their countries, have directed humanitarian assistance toward strengthening Middle Eastern states’ capacity to cope with the pandemic.[fn]Crisis Group interviews, March-April 2020. In late March, the EU pledged to provide Morocco with €150 million and Tunisia with €250 million to help them fight the pandemic and blunt its socio-economic impact. “Tunisia: EU supports national efforts to fight against Covid-19”, EU press release, 30 March 2020; and “Covid-19: EU mobilisation in support of Morocco’s efforts”, EU press release, 31 March 2020. The EU also provided €240 million in aid to Lebanon, Jordan and Iraq to strengthen these states’ capacity to deal with the virus’s spread among Syrian refugees whom they are hosting. “EU approves close to €240 million to strengthen resilience in neighbouring countries hosting Syrian refugees in light of the coronavirus pandemic”, European Commission, 31 March 2020.Hide Footnote

The EU and European governments are also troubled by a Trump administration whose policies in the Middle East they see as raising tensions, and some are seeking to develop a greater degree of foreign policy autonomy from the U.S. within the bounds of the transatlantic alliance. At a UN Security Council debate on peace and security in the Middle East in late August 2019, government representatives made emphatic reference to the need for a collective effort at reducing tensions in the Gulf.[fn]The UK ambassador to the UN said that, following events in the Gulf, the international community should find a way to address the region’s challenges collectively, including by launching a “serious, inclusive dialogue between regional and international actors” concerning the situation in the Strait of Hormuz. The German representative reported that Berlin, Paris and London were “looking into options of how to foster regional cooperation on maritime security” in the Gulf. And a representative for the UN Secretary-General declared that “[t]he first order of business must be preventing the most acute flashpoints in the region from boiling over. Keeping the channels of communication open needs to be priority number one, followed by confidence-building measures to move parties away from confrontation towards dialogue”. UN Security Council meeting record, 8600th meeting, “Maintenance of international peace and security: Challenges to peace and security in the Middle East”, UNSC S/PV.8600, 20 August 2019.Hide Footnote In an extraordinary EU Foreign Affairs Council meeting responding to the spike in U.S.-Iran violence in Iraq, convened on 10 January 2020, EU foreign ministers gave the EU’s high representative for foreign affairs and security policy, Josep Borrell, a mandate to “carry out diplomatic efforts with all parties in order to help secure a de-escalation of tensions in the region, to support political dialogue and to promote a political regional solution”.[fn]Council of the European Union, 5173/20, 10 January 2020.Hide Footnote Laying the groundwork for a process that could eventually earn U.S. backing would be a step in that direction.

The EU and European governments are troubled by a Trump administration whose policies in the Middle East they see as raising tensions.

For now, European governments are reluctant to be seen to be mounting a collective effort at persuading Gulf states of the need for an inclusive dialogue. For their part, Gulf officials express scepticism that Europe is well placed to do so.[fn]According to a Gulf official, “There is very little Europe can do on this issue if the U.S. is not on board. Just look at the issue of INSTEX”. Crisis Group interview, March 2020. INSTEX is the Instrument in Support of Trade Exchanges, a mechanism that enables non-U.S. dollar transactions, established by European countries in 2019 in response to U.S. sanctions on Iran.Hide Footnote Still, even as they agree that any dialogue should be locally owned and initiated by Gulf actors, European governments make clear that they believe such an effort is necessary and that Gulf states are unlikely to do so without the backing of external powers. European governments are therefore using the present time to start discussing preparatory issues of participation (who should join an external core group at this stage, including non-European states), approach (how to convince Gulf states to participate), timing and prerequisites for a successful launch.[fn]Crisis Group interviews, European capitals, 2019-2020.Hide Footnote

4. Principles guiding a dialogue process

A principal aim of a Gulf-based security dialogue should be to reduce tensions by establishing communication channels. Discussions could focus on shared principles governing their relations and cover each side’s motivations, core concerns and threat perceptions. They could then evolve toward decisions on concrete confidence-building measures.

Initially, relatively modest steps could include agreement on reducing inflammatory rhetoric; issuing unilateral statements in support of dialogue and joint statements outlining shared principles and interests; or opening multiple direct (even if confidential) communication channels, such as a de-confliction hotline among Gulf states and with outside actors whose military assets are deployed in the Gulf. The sides could also initiate technical discussions on matters of shared concern, such as cross-border adverse effects of climate change (extreme heat, droughts, water scarcity), deteriorating water quality, disaster preparedness, the spread of COVID-19, maritime security and religious tourism/pilgrimages. If and when initial discussions start to yield results, they could be scaled up to focus on ways to de-escalate tensions through shared security mechanisms (eg, prior notification of troop movements and military exercises; allowing adversaries to send military experts to observe such manoeuvres).

Eventually, Gulf parties could explore ways of fostering a durable cooperative regional security framework that includes all main stakeholders. In order to keep expectations in check, this objective might not be made explicit at the outset. By the same token, the goal at the initial stage ought not to be a “grand bargain”, whose achievement is desirable but highly unlikely under current circumstances. The purpose of a security dialogue, apart from whatever concrete progress can be reached on above-mentioned issues, should be to lower tensions by opening communication channels, and thus reduce risks of a deliberate or inadvertent major war.

One could draw positive and negative lessons from the Conference on Security and Cooperation in Europe that led to the 1975 Helsinki Accords.

Past experiences in negotiating international agreements may offer useful ideas, despite significant differences in geopolitical configurations and conditions. For example, one could draw positive and negative lessons from the Conference on Security and Cooperation in Europe (CSCE) that led to the 1975 Helsinki Accords or even the negotiations that yielded the 1648 Peace of Westphalia.[fn]See Morgan, op. cit., for a historical analysis of the process that brought about the 1975 Helsinki Accords; and Patrick Milton, Michael Axworthy and Brendan Simms, Towards a Westphalia for the Middle East (London, 2018), for lessons that can be drawn from the negotiations that helped end the Thirty Years’ War in Europe.Hide Footnote The use of relatively neutral powers to initiate a process is one such lesson; others are mentioned below.

A process aimed at de-escalating tensions and establishing mechanisms to reduce chances of stumbling into war could be guided by the following broad principles:

Dialogue as a process. Attempting regional dialogue ought not to be a one-off affair. The CSCE talks took almost three years from start to finish (not counting a long preparatory period), with negotiators hunkered down at a conference centre in Geneva throughout most of this time.[fn]The talks began in November 1972 and concluded in June 1975 with the signing of the final act. See “The Helsinki Process and the OSCE”.Hide Footnote

The importance of process versus outcome. Setting clear objectives at the outset about the desired outcome may be less important than the dialogue itself. In the process of talking, tensions may diminish as participants share information and perspectives and some converging views might emerge.[fn]Michael Cotey Morgan, author of a seminal history of the Conference on Security and Cooperation in Europe that produced the Helsinki Accords, said: “Was there a grand design at the outset of the CSCE? That is the crucial question. The answer is yes and no. There was a perception of a domestic and international crisis, and that some kind of broad comprehensive response was needed. The CSCE was a key component of this response. … The back and forth about how to go about a conference went on for years (ie, until 1972), and then the negotiations took nearly three years to produce the Final Act. At the start of the negotiations, none of the participating states could predict what would emerge from them. The result reflected a combination of grand strategic goals and improvised bargains”. Presentation at Crisis Group office, New York, 24 May 2019 (based on notes, approved by Michael Cotey Morgan).Hide Footnote It would be important to maintain a running draft non-paper on overlapping positions and near-agreements.

The region’s governments should want to own such a process even if they cannot muster the political will to initiate it by themselves. By contrast, outside imposition would breed resistance in the form of foot-dragging or outright rebellion. The CSCE process provides an example of a situation in which two sides in conflict found it advantageous to start talking.[fn]Morgan, op. cit., pp. 18-49.Hide Footnote The moment may come when Gulf actors reach the same point.

Over time, such a process should be as inclusive as possible. A principal goal should be to limit the risk that those excluded or unwilling to participate will become spoilers. A dialogue process should start in the Gulf and be extended to other regional players only once it gains momentum and a modicum of success.[fn]The CSCE process included all main stakeholders across the East-West divide, comprising 35 nations: Warsaw Pact members, all Western European states, the U.S. and Canada. On the Eastern Bloc side, Albania refused to participate; on the Western European side, Monaco joined only at a late stage, signing the final act. Morgan, op. cit., p. 108. Likewise, Milton, Axworthy and Simms say that negotiations that ended the Thirty Years’ War demonstrated that “a peace settlement reached at a congress at which all parties are represented … is less likely to be broken at the nearest opportunity than a set of bilateral or sectional arrangements”. Towards a Westphalia for the Middle East, op. cit., p. 111.Hide Footnote

The process should be predicated on consensus-based decision-making. The goal should be to maximise buy-in.[fn]The CSCE strictly observed the consensus rule. Morgan reports: “Granting every state a veto over every decision increased the risk that the proceedings would end up stalemated, but it also protected participants from unpalatable results and gave the resulting agreement a political weight that it would not have enjoyed if decisions had been taken by majority vote”. The Final Act, op. cit., pp. 108-109. Morgan also recounts that when the small Western European island nation of Malta – a country that “had fewer people than many European cities” – threatened to throw a last-minute spanner in the CSCE’s wheels by conditioning its signature to a final agreement on certain concessions in the Mediterranean from its Western allies, it got what it wanted. In Morgan’s words, however infuriating Malta’s “antics” were to other participants in effectively taking the conference hostage, the CSCE’s influence “depended in part on its universality. … Scrapping the rule of consensus would have undermined the CSCE’s implicit claim that international affairs in Europe were governed as much by rules … as by power”, ie, the implication that “some states mattered, but others did not”. The Final Act, op. cit., pp. 199-200.Hide Footnote

Keep proceedings out of the limelight. It would be difficult to shield any such large-scale process from international attention, but every effort should be made to keep proceedings closed to anyone but the stakeholders in order to increase chances of progress.[fn]Morgan recalls that the CSCE talks unfolded behind closed doors, and that: “Out of the media spotlight, the diplomats in Helsinki and Geneva created a culture of their own. … Through many months and thousands of meetings, they came to respect and even like one another”. While this familiarity did not eliminate the political conflicts, he writes, it “may have blunted [them]”. The Final Act, op. cit., p. 112.Hide Footnote Participants could establish a media committee to update the public daily and address media questions in order to achieve maximum transparency without hindering the process.

The UN remains the sole non-partisan actor with the requisite expertise to guide a conflict to a peaceful conclusion.

The UN should take a limited supporting role. It would be unrealistic to expect the UN to be able to assume a major role, at least as long as the Security Council remains as dysfunctional as it is at present, whereas working through the UN General Assembly and member states would recreate an unwieldy process. Yet it would be important to have the UN’s blessing. The UN Security Council could give its imprimatur under Resolution 598 (Paragraph 8) of 1987, which provides a legal basis for a formal UN role in a regional dialogue.[fn]See fn 3 above.Hide Footnote Iran, Kuwait and Russia, all of whom advocate a direct UN role, have referred to the resolution in the past. The UN has been a useful venue for early discussions on the topic, and UN officials may be able to nudge actors in the region in the right direction. The UN also remains the sole non-partisan actor with the requisite expertise to guide a conflict to a peaceful conclusion and assist in the post-conflict transition. It could play a useful auxiliary, mostly technical, role. The aim should not be a UN-led process but a process that the UN can support.

Do not start at the leadership level. To avoid politicisation during the initial exploratory stages, the level of participation could be held relatively low in the early stages, involving perhaps formal and informal political advisers to senior policymakers before graduating to leadership circles.

Be flexible. If, for example, representatives of adversaries refuse to meet face to face, proximity talks – in which they are located in separate venues with a mediator shuttling back and forth – or discussions through lower-level officials (whose participation may be less controversial) might be possible until conditions ripen.[fn]At Westphalia, the peace congress was convened in two separate cities for this reason. See Milton, Axworthy and Simms, op. cit., p. 112, who say that by such means, “the most unpalatable encounters could be avoided, if necessary”, while preserving inclusivity.Hide Footnote

Obtain “no objection” (nihil obstat) declarations from fence sitters. If certain external stakeholders are not prepared to back the envisioned collective exercise, participants should seek to secure their agreement not to block the process.

Parallel efforts. Because of the precarious nature of any collective effort and the multiple obstacles and challenges it faces, concerned actors should not hesitate to engage in parallel bilateral efforts to reduce tensions. Such efforts should be mutually supportive, however. Coordination is therefore critical. Of many experiments, one can hope that at least one will succeed.

IV. Favourable and Unfavourable Conditions

While a discussion regarding the optimal timing for a regional security dialogue is important, it should not distract the parties from the overriding concern that the crisis in the Gulf and wider region is acute and that all avenues for reducing tensions should be explored urgently.

A. Factors Conducive to Starting a Process

The political will to engage in risky diplomatic exercises tends to rise when dangers of military confrontation are highest.

With the U.S. placing “maximum pressure” on Iran, now may not be the time to launch a process aimed at de-escalating tensions. Yet it also arguably is precisely the right moment, as the political will to engage in risky diplomatic exercises tends to rise when dangers of military confrontation are highest. Conditions in the Middle East are at a potential tipping point. The situation lends itself to sudden and rapid escalation, possibly triggered by non-state actors or rogue elements with no stake in maintaining calm. Dialogue may help states desirous of avoiding confrontation use the sense of crisis to achieve progress in negotiations. The fact that the UAE opened a direct channel of communication with Iran in July 2019 is one such indication; the apparent eagerness of European governments to support a Gulf-based process is another.

Tensions in the Gulf are marked by significant polarisation and mistrust. Yet there are also factors that suggest the potential for dialogue.

A rough power balance. Iran made dramatic gains in the wake of state collapse in Iraq, Syria and Yemen, scaring Saudi Arabia and smaller Gulf states, and ringing alarm bells in Israel. Iran prides itself on its ability to provide its own security, while Saudi Arabia has a powerful international protector in the U.S. More recently, U.S. sanctions have weakened Iran’s economy and arguably drawn a circle around its regional reach even as U.S. inaction after the September 2019 Aramco attacks suggests that Saudi Arabia may be unable to count on consistent U.S. protection.[fn]Hossein Mousavian and Abdulaziz Sager argued as much in a joint op-ed. “It’s time for the leaders of Iran and Saudi Arabia to talk”, The New York Times, 14 May 2019.Hide Footnote The U.S. killing of Qassem Soleimani in early January and Iran’s somewhat limited military response may have restored a fragile mutual deterrence in their interaction. Combined, these various factors may enable dialogue on a more level playing field.

States favouring diplomacy. States not directly responsible for increasing tensions but highly vulnerable to fallout from war have been giving signals that they prefer diplomacy over belligerence. These include both regional states (notably the UAE, which is directly involved in the standoff and has backed U.S. sanctions on Iran) and European governments, which see threats to their stability if the Middle East descends into war. In Asia, Pakistan also has given clear signals of concern, and has offered to use its relationships with both Saudi Arabia and Iran to mediate between the two. European governments have funded, quietly, NGO-led dialogues to explore ways to lessen tensions; in addition, today, they exhibit a keenness to become directly involved.

Quiet dialogue. NGO-led Track 1.5 and Track 2 dialogue exercises have kept communication channels between adversaries open, even if in a limited way. These surrogate efforts have clarified misunderstandings and passed messages; they have also laid the foundation for a jump to Track 1 negotiations once conditions ripen.[fn]Ibid.Hide Footnote

The COVID-19 crisis. Humanitarian crises may offer opportunities for dialogue and even cooperation. The COVID-19 outbreak may provide such an opportunity. Iran confirmed its first cases of infection on 19 February and, so far, has become far and away the most badly affected country in the MENA region.[fn]Crisis Group Middle East Briefing N°76, Flattening the Curve of U.S.-Iran Tensions, 2 April 2020.Hide Footnote Some of the smaller Gulf states responded. Qatar, not surprisingly, but also the UAE and Kuwait dispatched medical aid to Iran in February and March.[fn]There are pre-COVID-19 examples as well, of course. Following the deadly crush in Mecca in 2015, in which over 2,000 pilgrims died, including more than 400 Iranians, Iran and Saudi Arabia engaged in technical talks to calm tensions and recommence peaceful pilgrimages. Following the December 2003 earthquake in the Iranian city of Bam, the U.S., which had designated Iran as a member of the “axis of evil” the year before, shipped medical aid to Iran by military cargo plane. “U.S. airlifts disaster aid into Iran”, CNN, 30 December 2003.Hide Footnote These gestures are politically significant, and the parties could build upon them to open further channels of communication. Progress toward dialogue is not a given, because efforts to stem COVID-19’s spread have prompted states to close their borders, while meeting in a common space will probably be impossible for a while. Yet the memory of good-will may linger and pay off politically down the road.

B. Obstacles and Challenges

While certain conditions are necessary and may be present for the launch of a collective security effort in the Gulf, it would face tremendous obstacles as well. Most importantly, even if a window is open at any given time, it could rapidly close again due to a new escalation.

The main reason why Saudi Arabia has refused to engage in direct talks with Iran so far is its sense that Tehran would be negotiating from a position of strength.

An Arab perception of power imbalance does not help. While objectively the power balance appears relatively even, as discussed, a main reason why Saudi Arabia has refused to engage in direct talks with Iran so far is its sense that Tehran would be negotiating from a position of strength. Saudi officials have long insisted that they need to roll back Iranian influence in Yemen, Iraq, Syria and Lebanon before engaging in any talks.[fn]Crisis Group interviews, senior Saudi officials, February 2020, December 2019, September 2019.Hide Footnote Washington’s unreliability also adds to Riyadh’s sense of vulnerability despite the pressure under which Tehran is suffering. The Saudis also are highly reluctant to act in any way contrary to the Trump administration’s desires at a time when they believe Tehran is trying to drive a wedge between Riyadh and Washington. For now, at least, the kingdom says it will follow Washington’s lead in backing “maximum pressure” rather than diplomacy.[fn]A senior Saudi official said: “We know that Iran wants to drive a wedge between us and the U.S. by initiating a dialogue. But we cannot afford there being any daylight with Washington and for that reason we will support maximum pressure and avoid any official engagement with Iran”. Crisis Group interview, February 2020.Hide Footnote Entering into talks with Iran in these circumstances therefore would require both a change of heart and a leap of faith. It remains an open question whether and when the Saudi leadership might undergo the former and be willing to make the latter.

V. Initial Steps

Outside actors can encourage but should not determine what happens next, as attempts at imposition or other forms of undue pressure likely will fail. Ideas ought to arise organically and be locally owned. The following proposed steps are therefore offered merely as preliminary and provisional ideas:

  • European and other states willing to take the first step could establish a core group to express their support for an inclusive dialogue among Gulf states.
  • Policymakers from states participating in this external core group could initiate discussions with Gulf states and the U.S. to explore the possibility of such a process being launched from within the Gulf sub-region.
  • If these discussions prove successful, Gulf states that are keenest to proceed down this path could start informal talks with governments in the Gulf sub-region – the six GCC states, Iran and Iraq – and, when they deem the time ripe, extend an invitation to them to launch preparatory talks in order to set the agenda and rules of procedure of an eventual conference, and discuss initial ideas for feedback and support. The external core group would provide diplomatic support for this effort.
  • Based on this feedback, the Gulf actors could work out a preliminary plan for ways to manage a conference, including by setting up mechanisms to regularly brief the media and outside governments about the proceedings.

VI. Conclusion

The main threat the region faces today is not so much a war of choice but an inadvertent one.

As the risk of a military confrontation in the Gulf grows, it is imperative to find ways to de-escalate tensions. The main threat the region faces today is not so much a war of choice but an inadvertent one that results from miscalculation, misinterpretation or lack of timely communication. Key actors in the Middle East have refined the game of brinkmanship to the point of playing it right up to the edge. The result has been the fraying of the thread dividing war from no war.

An inclusive, collective regional security dialogue aimed at lessening tensions may have only a small chance of success. But in the current circumstances it would be irresponsible not to give it a try.

Riyadh/Abu Dhabi/Tehran/Muscat/Doha/Brussels, 27 April 2020

Appendix A: Map of the Middle East Region

Map of the Middle East Region. UNITED NATIONS