Arrow Down Arrow Left Arrow Right Arrow Up 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 Twitter Video Camera Youtube
Myanmar Foreign Minister and State Counselor Aung San Suu Kyi attended the opening ceremony of the Union Peace Conference at Naypyidaw, Myanmar’s capital city, on 12 January 2016. AFP/Ye Aung Thu
Briefing 149 / Asia

Myanmar’s Peace Process: Getting to a Political Dialogue

After almost 70 years of armed conflict, Myanmar has a rare but fading opportunity to finalise a broad-based, federal settlement. The government must adopt a more flexible approach that allays opposition concerns, and armed groups need to go beyond preliminaries and engage in meaningful discussions.

Also available in: Burmese [PDF]

I. Overview

The current government term may be the best chance for a negotiated political settlement to almost 70 years of armed conflict that has devastated the lives of minority communities and held back Myanmar as a whole. Aung San Suu Kyi and her administration have made the peace process a top priority. While the previous government did the same, she has a number of advantages, such as her domestic political stature, huge election mandate and strong international backing, including qualified support on the issue from China. These contributed to participation by nearly all armed groups – something the former government had been unable to achieve – in the Panglong-21 peace conference that commenced on 31 August. But if real progress is to be made, both the government and armed groups need to adjust their approach so they can start a substantive political dialogue as soon as possible.

Pangalong-21 was important for its broad inclusion of armed groups, not for its content, and the challenges going forward should not be underestimated. Many groups attended not out of support for the process, but because they considered they had no alternative. Many felt that they were treated poorly and the conference was badly organised. The largest opposition armed group, the United Wa State Party (UWSP), sent only a junior delegation that walked out on the second day. An escalation of fighting in recent months, including use of air power and long-range artillery by the Myanmar military, has further eroded trust.

Such issues are not unexpected; what matters is the resilience of the process to deal with them. The announced scheduling of further Panglong-21 conferences every six months (the next for February 2017) imposes an artificially rigid timeframe that limits the flexibility required to overcome obstacles. Weak capacity in the government’s peace secretariat, the National Reconciliation and Peace Centre (NRPC), is another challenge. It will take difficult negotiations to convince most groups to sign the Nationwide Ceasefire Agreement (NCA), a sine qua non for participation in the upcoming political dialogue process – future Panglong-21 conferences and the discussions feeding into them – that has been clearly articulated by both the government and military. This will be even harder if the military continues its forceful posture on the ground.

Eight groups signed the NCA in October 2015, but at least ten other armed groups have reservations. Some, like the UWSP, have better de facto self-governance arrange­ments already and worry their status would be undermined by signing. Others are concerned that the new government has a more unilateral approach to the peace process and that if they sign, political solutions are more likely to be imposed than negotiated. Three groups without bilateral ceasefires are resisting government demands to issue statements renouncing armed struggle in principle.

The government should consider adopting a more flexible timeframe for the peace conferences and reassure armed groups by demonstrating a less unilateral approach to the process in general. It needs to ensure that civil society, women and youth have a stronger voice in the process. It should also take steps to ensure that it has the necessary support capacity in place at the NRPC.

Armed groups need to recognise that though they have legitimate concerns about the process, they are unlikely to get a better chance to achieve a negotiated political settlement. Aung San Suu Kyi has expressed firm support for a federal, democratic solution and has unparalleled political authority to deliver it, particularly with the Burman majority. Now is the time to start discussing the contours of that deal, rather than continuing to focus on preliminaries.

The alternative is not attractive. Time is not on the side of the armed groups. Unless both sides grasp the current opportunity, the prospect of a negotiated solution will recede, likely to be replaced by a messy, drawn-out endgame that fails to address the underlying grievances of the minority communities, including their demands for a federal system and greater equality. This would be to the detriment of peace and stability in the borderlands and to Myanmar’s future as a prosperous, tolerant and democratic country.

II. Peace Legacy from the Previous Government

A. Peace Process with Armed Groups

The administration that took power on 30 March 2016 inherited a peace process that had been in stasis during the lame-duck period leading up to the November 2015 elections and the lengthy handover period afterwards.[fn]For recent Crisis Group reporting on Myanmar, see Asia Briefings N°s 147, The Myanmar Elections: Results and Implications, 9 December 2015; 146, Myanmar’s Peace Process: A Nationwide Ceasefire Remains Elusive, 16 September 2015; 144, Counting the Costs: Myanmar’s Problematic Census, 15 May 2014; 143, Myanmar’s Military: Back to the Barracks?, 22 April 2014; also Reports N°s 282, Myanmar’s New Government: Finding Its Feet?, 29 July 2016; 266 Myanmar’s Electoral Landscape, 28 April 2015; 261, Myanmar: The Politics of Rakhine State, 22 October 2014; and, for more detailed historical background on the armed conflict, 214, Myanmar: A New Peace Initiative, 30 November 2011.Hide Footnote  The previous government had had considerable early success, agreeing bilateral ceasefires with fifteen armed groups between 2011 and 2013 (see Appendix B and the acronyms in Appendix C). There was much optimism on 31 March 2015, when the government and armed group negotiating teams initialled the NCA. However, concerns over the lack of inclusivity (the government did not allow the three groups without bilateral ceasefires – AA, TNLA and MNDAA – to sign) as well as about giving the government of then-President Thein Sein a major victory just ahead of elections, stalled the process. Eventually, eight armed groups signed the NCA at a ceremony on 15 October 2015; the remaining ten involved in the formal peace process did not. This led to some tensions between signatory and non-signatory groups.[fn]For all armed group acronyms, see Appendix B.Hide Footnote

The NCA contains basic principles recognising the territorial integrity of the state (making clear that separatism or irredentism is unacceptable), committing to “principles of democracy and federalism” and embracing the diversity of the peoples and cultures in “a secular state”. A military code of conduct prohibits certain conduct by all parties in ceasefire areas (attacks, reinforcement, recruitment, new bases, laying landmines, etc.) and sets out troop deployment provisions to avoid clashes. There is provision for a joint ceasefire monitoring body, and “interim arrangements” endorse armed groups’ de facto authority in their areas of control for a transitional period. The NCA is to be followed by a “political dialogue”, consisting of a Union Peace Conference to reach a comprehensive peace agreement that would be “the basis for amending, repealing and adding provisions to the constitution and laws, in line with agreed procedures” – that is, through the legislature – along with armed group disarmament and security sector reform.[fn]For a detailed summary of the NCA, see Crisis Group Briefing, Myanmar’s Peace Process, op. cit., Section IV.Hide Footnote

Finalisation of the NCA was thus only the first step in a long, difficult process needed to reach a comprehensive peace agreement. Many of the most challenging issues, including a possible form of federalism, how revenue would be shared, future status of the armed groups and their possible integration into the military, were deferred to the political dialogue, as were some technical military issues on ceasefire monitoring and code of conduct. It is thus neither a classic ceasefire agreement – many military issues, such as force separation, demarcation and verification, are vague, not included or need further agreement to come into force – nor a full political agreement, as it references many political issues but defers detailed discussion. This hybrid status reflects its genesis, the diverse actors and priorities around the table and political constraints.

Following the partial signing, the previous government took formal steps to implement the NCA, specifically:

  • A first session of the Joint Implementation Coordination Meeting, the body mandated to oversee NCA implementation, was held 15-17 October 2015. It established the committees set out in the NCA to take the process forward: the Joint Monitoring Committee (JMC) for military and ceasefire matters and Union Peace Dialogue Joint Committee (UPDJC) for political dialogue. The JMC contains ten representatives of NCA-signatory armed groups, ten of government (including military), and four independent civilians; there are also subnational committees. The UPDJC initially had sixteen representatives each of NCA-signatory armed groups, government (including military and legislature) and political parties and was chaired by then-Vice-President Sai Mauk Kham.
     
  • A joint legislative session ratified the NCA on 8 December, giving it legal status.
     
  • A Framework for Political Dialogue was agreed on 15 December, including the mandate, agenda, working methods and proportions of representatives to be included in the dialogue.
     
  • The first Union Peace Conference was held 12 to 16 January 2016, with opening addresses by the president, commander-in-chief, Aung San Suu Kyi and Mutu Say Poe, the head of the Karen National Union armed group. The conference had 700 participants but, occurring in the lame-duck period after the elections, was largely symbolic, intended only to launch the process and keep to the NCA’s ambitious political roadmap. Armed groups that did not sign the NCA were invited to observe, but nearly all declined.[fn]In accordance with the Framework for Political Dialogue, the 700 seats were divided 75 each for government and legislature, 150 for military, 150 each for ethnic armed groups and registered political parties, 50 each for ethnic representatives and other relevant persons. The roadmap required the Framework for Political Dialogue to be agreed within 60 days of the NCA signing and the dialogue to commence within 90 days. One non-signatory group, the National Socialist Council of Nagaland-Khaplang, did accept the invitation. Three non-signatory armed groups without bilateral ceasefires (Arakan Army, Ta’ang National Liberation Army, Myanmar National Democratic Alliance Army) were not invited.Hide Footnote

B. Armed Conflict

Notwithstanding these important procedural developments, the peace process essentially was in stasis between the NCA signing and the new government taking up the issue in April 2016. Meanwhile, the situation on the ground remained volatile, with fighting continuing to break out sporadically, and often unexpectedly, in many different parts of the country.

Most groups that signed the NCA are based near the Thai border in southern Shan State and the south-east. Their signing consolidated a fragile local peace, or at least absence of war, that had prevailed for some time. Groups based near the Chinese border did not sign, and the situation in many of those areas continued to be unstable, with regular, sometimes intense fighting, including between ethnic armed groups. The geographic split reflects very different political-economic realities between the areas, including access to funding and weapons and the distinct policies and approaches of China and Thailand.

Serious bouts of conflict since early 2015 include:

  • in Shan State, resumed major fighting between Myanmar National Democratic Alliance Army (MNDAA) troops and government forces in the Kokang Self-Administered Zone since February 2015, which was particularly intense from February to June that year and again in October 2015. Elsewhere in Shan State, there have been sporadic clashes between government forces and the Ta’ang National Liberation Army (TNLA) and between that group and the Shan State Army-South (SSA-South). There have also been clashes between government forces and the SSA-North, of particular intensity from October to November 2015 and in August 2016;
     
  • in Kachin State, between government forces and the Kachin Independence Organisation (KIO) throughout the period, and in particular from July to November 2015, and again from April to August 2016;
     
  • in Rakhine State and southern Chin State, occasional, sometimes heavy clashes between government forces and the Arakan Army, in particular in April 2015, January 2016 and from April to June 2016; and
     
  • in Kayin State, clashes in July 2015 and again from August to September 2016 between a renegade faction of the Democratic Karen Buddhist Army (DKBA) and government troops together with Border Guard Force soldiers.

Such conflicts are usually accompanied by grave violations of human rights by all belligerents.[fn]See, for example, “Report of the Special Rapporteur on the situation of human rights in Myanmar”, UN OHCHR A/HRC/31/71, 18 March 2016.Hide Footnote  They undermine stability and trust in the peace process and severely impact lives and livelihoods – particularly of those most at risk, including women and children – often causing internal displacements.[fn]For a detailed risk analysis, see “Kachin and northern Shan protection concerns and risk analysis”, Protection Sector, October 2015.Hide Footnote  Some 100,000 people remain displaced in Kachin and northern Shan states as a result of fighting following the 2011 breakdown of the KIO ceasefire. Fighting in the Kokang Self-Administered Zone displaced around 80,000 in February 2015, the majority to China, though most have now returned. At least 12,000 were displaced in northern Shan State in the first half of 2016 in the complex conflicts that included government forces, the TNLA and the SSA-South; most have returned home, but some 3,000 remain displaced. The fighting in Rakhine State in March-April 2016 displaced approximately 1,900, who have yet to return home. Most recently, fighting in Kayin State displaced some 4,000 in September 2016.[fn]Figures from UN Office for the Coordination of Humanitarian Affairs, except Kayin State displacements, from “Tatmadaw launch operations against KKO splinter group in Wah Boh Taung-Kyonhtaw, Methawaw regions”, Global New Light of Myanmar, 17 September 2016.Hide Footnote

III. The New Government’s Approach

A. First Steps

During the previous government’s tenure, the National League for Democracy (NLD) was invited, with other political parties, to participate in the peace process. Though it sent representatives, their engagement was limited. Aung San Suu Kyi kept her distance and was at times critical of the process. Her speech to the inaugural Union Peace Conference in January 2016 (above) was thus significant.

Suu Kyi had indicated that achieving peace would be a top priority for her government, and the NLD’s election manifesto addressed this as its first item, promising to “hold political dialogue based on the Panglong spirit in order to address the roots of internal armed conflict” – referring to the pre-independence Panglong Conference, convened by her father in 1947.[fn]“2015 Election Manifesto”, NLD, official translation, p. 5. For details on the 1947 Panglong Conference, see Crisis Group Report, Myanmar: A New Peace Initiative, op. cit., Section I. The 1947 Panglong Agreement was not a peace deal – there was then no insurgency – but an agreement by some ethnic areas (Shan, Kachin and Chin) to join an independent Burma in return for promises of full autonomy in internal administration and an equal share in national wealth.Hide Footnote  In her first major speech after the transfer of power, a Myanmar New Year’s message to the nation on 18 April, Suu Kyi stated that the government would aim to bring remaining organisations into the NCA, and “through peace conferences, we’ll continue to be able to build up a genuine, federal democratic union”.[fn]“State Counsellor offers New Year message”, Global New Light of Myanmar, 18 April 2016.Hide Footnote  She indicated that she would personally lead the process.

She gave the first concrete indication of her plans at a 27 April JMC meeting, announcing that a new 21st Century Panglong (Panglong-21) peace conference would be held within two months. This caused consternation among ethnic leaders due to both form and substance. There had been no prior consultation with ethnic armed groups or political leaders; and no details were provided on the initiative, which was seen as potentially signalling a unilateral shift in approach in a process with a legally-binding framework that had required months of detailed negotiation. The venue for the announcement compounded these concerns, as the JMC is tasked with military or ceasefire matters, not the political dialogue, for which the UPDJC is the mandated body.[fn]Crisis Group interviews, ethnic party and armed group leaders, Yangon, May-July 2016. For example, a month later the leader of the Shan State Army-South, a major armed group that signed the NCA, expressed concern on both aspects. “Lt-Gen Yawd Serk: If this conference is wrong, it will affect the future of the union”, Shan Herald Agency for News, 26 May 2016.Hide Footnote

In a 26-28 May meeting of the UPDJC, which she chairs, Suu Kyi sought to allay some concerns. She confirmed she would continue to follow the NCA framework, and Panglong-21 was only a different name for the Union Peace Conferences that framework envisaged. While this reassured ethnic leaders, other comments raised new concerns, notably her stated intention to narrow the scope of discussions in the political dialogue from the five thematic areas agreed in the UPDJC to federalism and security.[fn]Ibid. “NCA to guide 21st Century Panglong Conference”, Global New Light of Myanmar, 28 May 2016. The previously-agreed five areas are set out in the Framework for Political Dialogue, which is being amended. The three thematic areas proposed to be dropped were: social issues (including culture, language, gender, resettlement, human rights, drugs), economic issues (including foreign investment, tax and revenue distribution and regional development) and issues around land and natural resources (including resource management and revenue sharing).Hide Footnote  This would leave out some key areas of concern and missed an opportunity to build confidence by addressing easier issues, such as language policy. With armed group leaders strongly opposed, the matter was not settled before the Panglong-21 conference, and discussions are ongoing. It is likely armed group concerns will be accommodated, and the dialogue’s scope will remain unchanged, though with some effort to focus on priority issues.[fn]Crisis Group interview, member of UPDJC, Yangon, September 2016.Hide Footnote  There has to date been little outreach to civil society, and few efforts to engage a wider range of voices in the peace process, particularly women and youth.

The government also announced a new peace architecture on 31 May, with three sets of structures:

  • the NCA-mandated JMC and UPDJC, the latter now chaired by Suu Kyi and with party membership limited to those that won seats in the last elections;
     
  • a committee to transform the previous government’s Myanmar Peace Centre into a National Reconciliation and Peace Centre (NRPC). This new centre, launched on 11 July, is headed by Suu Kyi. Under it is a new Peace Commission, chaired by Dr Tin Myo Win, her personal physician and newly-designated chief peace negotiator.[fn]Established by President Office Orders 50/2016 and 51/2016, 11 July 2016.Hide Footnote  Unlike its predecessor, a semi-government body staffed mainly by non-government experts, it is a government institution under Suu Kyi’s State Counsellor Office, staffed by civil servants and governed by civil service laws and financial rules; and
     
  • a Panglong-21 preparatory committee also chaired by Dr Tin Myo Win and sub-committees to liaise respectively with NCA-signatories and non-signatories.

B. Peace Conference Preparations

Though the date for Panglong-21 slipped from her initial late-June proposal, Suu Kyi appeared determined to avoid major delays. This seems to stem from two considerations: not wanting to repeat the experience of the previous government, when negotiations bogged down over process, particularly which armed groups would be included; and a sense that her leverage would be at its greatest early in her term, due to the election landslide. Some observers also believed she wanted the conference before her September meetings with President Obama in Washington DC and at the UN General Assembly. Thus, at her urging, there was agreement with the NCA signatories for Panglong-21 to begin no later than 31 August, a very ambitious timeframe both logistically and for obtaining buy-in of non-signatory armed groups.[fn]Crisis Group interviews, armed group leaders and international peace-process adviser, Yangon, July-August 2016. “Gov’t, NCA signatories agree to hold UPC no later than 31 August”, Global New Light of Myanmar, 29 June 2016.Hide Footnote

The intention to make Panglong-21 inclusive of all armed groups, stated from the outset, was positively received. This has long been a demand of the non-signatories. On 3 June, as a first step to secure their participation, Dr Tin Myo Win met the United Nationalities Federal Council (UNFC), the main umbrella organisation of non-signatories. He then met separately on 17-19 June with the UWSP and NDAA, non-signatories that are not UNFC members. Under the previous government, non-signatories were only invited as observers; the new government got around this by indicating that since the first Panglong-21 conference would be symbolic, with presentations but no negotiations or decisions, all armed groups would be “attendees” (tet-yauk-thu). The government position remained, however, that only signatories could participate in the future political dialogue.[fn]Crisis Group interview, member of the Peace Commission, Yangon, August 2016.Hide Footnote

There were also negotiations with the three previously-excluded groups: AA, TNLA and MNDAA. Since these lack bilateral ceasefires, they are not eligible to sign the NCA, and the military previously insisted they must disarm, something the groups equated with surrender. The commander-in-chief subsequently proposed that it would be sufficient to put their arms beyond use in some verifiable way, along the lines of formulas used in Aceh, Nepal and Northern Ireland, but this was rejected.[fn]Ibid. Also, commander-in-chief meeting with press, 13 May 2016, reported in “Tatmadaw sets out peace conference conditions”, Myanmar Times, 16 May 2016.Hide Footnote  Negotiations then focused on a statement committing the groups to renounce armed struggle in principle. Considerable progress was made, with the only sticking point being the Burmese-language term for “armed struggle” versus “violence”.[fn]Crisis Group interview, member of the Peace Commission, Yangon, August 2016.Hide Footnote  However, no agreement was reached, the three issued no statement, and they were not invited to Panglong-21. Crucially, however, that did not lead to the UNFC and other non-signatories boycotting, though lack of inclusion had been a key reason cited by groups for not signing the NCA.[fn]See Crisis Group Briefing, Myanmar’s Peace Process op. cit., Section III.B.Hide Footnote  

In the lead-up to Panglong-21, representatives of seventeen armed groups held a major strategy meeting in the KIO-controlled town of Maijayang, 26-30 July, to coordinate positions on key issues; the UN and China attended as international observers. Four armed groups did not attend (UWSP, MNDAA, TNLA and NSCN-Khaplang). The UWSP, together with its NDAA ally, went to Naypyitaw to meet on 29 July with Suu Kyi and then the commander-in-chief.[fn]The NDAA participated in both the Maijayang meeting and the Naypyitaw visit.Hide Footnote

C. The Panglong-21 Conference

The conference, officially the “Union Peace Conference – 21st Century Panglong”, was held in Naypyitaw from 31 August to 3 September. Suu Kyi’s opening address was followed by plenary speeches from the lower and upper house speakers, the commander-in-chief, the KNU chairman, NLD patron Tin Oo (an ex-commander-in-chief), the KIO vice chairman and UN Secretary-General Ban Ki-moon.[fn]The KIO vice chairman’s talk was a last-minute concession; there was initially no speaking slot for the non-signatory groups (Major-General N’Ban La also chairs the UNFC).Hide Footnote

Representatives of nearly all armed groups attended, except the AA, TNLA, MNDAA and NSCN-Khaplang.[fn]The first three were not invited; the NSCN-Khaplang, though invited, had long made clear it would not attend, as it is committed to the creation of an independent Naga homeland out of parts of Myanmar and India, which is politically inconsistent with the NCA and the peace process.Hide Footnote  Some 850 attendees participated over the four days. In a move armed group representatives welcomed for its transparency, the 72 ten-minute speeches were carried live on national television, “the first time in more than 50 years that they [were] able to express their desires and pent up aspirations to a national audience without fear of being arrested and put in prison”.[fn]Crisis Group interviews, armed group representatives, Yangon, September 2016. Quote from “Political Monitor No. 20”, Euro-Burma Office, 20 August-2 September 2016.Hide Footnote

The attendance of most non-signatories was an important step forward. However, it does not necessarily indicate significantly greater trust in the new government on the part of armed group leaders. It more reflects the very different political landscape – in particular, the domestic and international legitimacy of Suu Kyi. Many armed group leaders felt they had little alternative but to participate, despite reservations or concerns; some came under pressure from China to attend (see below).[fn]Ibid.Hide Footnote  A prominent ethnic politician, Khun Tun Oo, who chairs the Shan Nationalities League for Democracy, did boycott on the basis that the conference was not fully inclusive of armed groups (though the decision was undoubtedly influenced by political tensions between his party and the NLD).[fn]Crisis Group interviews, armed group representatives and analysts, Yangon, September 2016. “Khun Tun Oo absent from peace talks”, Shan Herald Agency for News, 31 August 2016. For details on the tensions, see Crisis Group Report, Myanmar’s New Government, op. cit., Section III.C.Hide Footnote

Several groups felt the conference had been hastily convened, and there was considerable unhappiness at flawed arrangements. Armed group delegations were not met at the Naypyitaw airport and had to find their own way to their accommodation; delegations, including some senior leaders, were housed dormitory-style by the government; written documents and nameplates did not give military ranks of armed group representatives or other honorifics (failure to use the equivalent of “Mr” or “Ms” before a name is culturally very impolite in Myanmar). A major group, the UWSP, walked out after the first day, saying it felt discriminated against, though this was at least as much a reflection of its ambivalence about the NCA as it was over a specific issue; it had sent only a low-level delegation.[fn]The UWSP delegation had booked itself into a prominent hotel, rather than stay at the government-assigned accommodation. Since groups were not met at the airport, the delegation did not collect its conference passes, and on the opening day a government organiser arranged temporary “observer” badges so the delegation could attend the plenary. Since these were not valid for the following day session, when the UWSP was to give its presentation, security barred the delegation, which then walked out in protest before organisers could remedy the problem. Crisis Group interview, organising committee member, Yangon, September 2016.Hide Footnote  Some of these issues arose from the tight timeframe for convening the conference, but others appear to have been the deliberate result of the government’s approach to organising it.

IV. Huge Challenges Remain

The government has indicated that it plans to hold such Panglong-21 peace conferences every six months.[fn]“Union Peace Conference to be held every six months”, State Counsellor Office statement, 15 August 2016.Hide Footnote  This would impose an artificially rigid set of deadlines on a process that must achieve the buy-in of diverse stakeholders on very contentious issues. Challenges lie in the preliminary matters that must be settled before the next session, the content of future political discussions and the political and security context.

A. Preparations for the Next Conference

Achieving broad participation by armed groups at the recent conference hinged on three things:

  • Suu Kyi, who won an electoral landslide, including in many ethnic areas, and enjoys strong international support as well, has great political capital and legitimacy. Most armed group leaders accordingly felt politically compelled to attend, unlike in the past. This was reinforced by the military’s support for the conference and the clear convergence of views between the soldiers and government on the peace process. China’s backing was also critical. The combination gave Suu Kyi a large advantage over the previous government, which had military support but far less legitimacy and no backing – indeed, sometimes obstruction – from China. (It also amplified the power asymmetry between the government/military and the armed groups, making the latter nervous.)
     
  • Decisions on difficult issues were postponed until after the conference. In particular, discussions on a revised Framework for Political Dialogue continue, and there is not yet agreement on topics to be included and how a series of “national dialogues” to feed into the next Panglong-21 will be conducted. Non-signatory groups declined to attend a September framework review meeting.[fn]Crisis Group interview, UPDJC member, Yangon, September 2016.Hide Footnote
     
  • Perhaps most importantly, the requirement that armed groups must sign the NCA to participate was not enforced. This was possible because the conference was billed as a symbolic launch, without discussions or decisions. But it remains firm government policy and a red line for the military that armed groups wishing to participate in the political dialogue must first sign the NCA. This message was reinforced by Suu Kyi and Commander-in-Chief Min Aung Hlaing, who made the NCA a key focus of their opening speeches.[fn]Reproduced in Global New Light of Myanmar, 1 and 2 September 2016, respectively.Hide Footnote

The timeframe is extremely tight. The next conference is due in February and may be timed to coincide with the 70th anniversary on 12 February of the 1947 Panglong agreement, celebrated annually as Union Day. Before this, there is need for negotiations to secure signing of the NCA by non-signatories and agreement on a revised Framework for Political Dialogue (targeted for end of October), followed by national dialogues in each state and region. All these steps are difficult, time-consuming or both, particularly getting more groups to sign the NCA. The largest armed group, the UWSP, is very reluctant to sign, because it is a de facto mini-state with far more autonomy than anything the NCA offers. The closely-allied NDAA is likely to follow its lead.

The seven UNFC groups (see Appendix B), particularly the larger ones, desire to reach a political settlement on the grievances driving decades of conflict – fundamentally, lack of autonomy and equality. They recognise the current moment may be the best opportunity they will ever get, but exclusion of the AA, TNLA and MNDAA makes the NCA politically problematic for them and a ceasefire militarily unfeasible. They also have not yet been offered any concessions – not even of the face-saving kind – for signing,[fn]In particular, the UNFC has put forward an eight-point proposal for amending/supplementing the NCA. It will be very difficult for the government to accept any changes now that it is signed by the former president, commander-in-chief and legislative speakers, as well as eight armed groups, and been ratified by the legislature. Some of the specific proposals are also quite difficult, but a compromise must be found. See also, Sai Wansai, “Framework for Political Dialogue: UNFC’s boycott leads to peace process deterioration”, Shan Herald Agency for News, 21 September 2016.Hide Footnote  and will be reluctant to do so if the only reason is to gain access to a process they view as driven unilaterally by the government and insufficiently sensitive to their concerns. They worry that conforming to an artificial, government-imposed timeframe would set a precedent for unilateral imposition of any subsequent political solutions.

Some UNFC members may also want to delay major decisions until the KNU holds its congress in November.[fn]Crisis Group interview, senior armed group representative, Yangon, September 2016.Hide Footnote  If a more hardline leadership results, they believe it could pave the way for this influential armed group to rejoin the alliance, enhancing its power and bargaining position. However, if the UNFC tries to prolong the process too much, it risks being marginalised, for example not being eligible to participate in the national dialogues, thereby giving government and political parties a stronger role in defining the peace process agenda.[fn]Crisis Group interviews, armed group leaders, members of government peace bodies and analysts, Yangon, July-October 2016.Hide Footnote

The issue of the three groups, AA, TNLA and MNDAA, without bilateral ceasefires is even more difficult. Including them in the next conference requires, at a minimum, agreement on a statement renouncing violence in principle; even then, they could likely attend only as observers. Having declined that for the last conference, it is far from clear whether they will do so ahead of the next; the TNLA sent an open letter to Panglong-21 stating it would “never lay down arms or renounce arms, at any time or under any circumstance”.[fn]TNLA open letter to the Panglong-21 conference, 31 August 2016.Hide Footnote  This not only matters for inclusivity, but also has on-the-ground consequences. These groups are to various degrees allied with or supported by the UWSP and KIO, and they fight together in joint patrols and in some cases together with the KIO and SSPP. All operate in adjacent or overlapping territory, and it is hard to imagine any ceasefire being sustainable without the three non-ceasefire groups.[fn]See “Military confrontation or political dialogue: Consequences of the Kokang crisis for peace and democracy in Myanmar”, Transnational Institute, July 2015.Hide Footnote  

A huge amount of procedural work and negotiation is required before the next peace conference. In addition to the inherent challenges, the peace architecture has quite limited capacity. Lead negotiator Dr Tin Myo Win works extremely hard but has no chief of staff for the process and continues his medical work for Suu Kyi and as a surgeon at a philanthropic hospital. The NRPC, tasked with the day-to-day work, has only a handful of staff, compared with 120 under its predecessor. Because Suu Kyi decided to establish it as a fully government entity under her office (its predecessor was semi-independent, at least administratively), it must follow civil service staffing and budgeting regulations. Scaling up will take considerable time, and it will be difficult to draw on outside expertise.[fn]Crisis Group interviews, individuals with direct knowledge, Yangon, June-September 2016. The new multi-donor Joint Peace Fund is an initiative that can provide significant resources, but it cannot necessarily overcome the regulatory restrictions the NRPC operates under.Hide Footnote  There is thus a worrying lack of institutional capacity to support peace-process mechanics, and the armed groups also have little support capacity.

B. Questions of Content

Now that the peace process set out in the NCA has been launched symbolically on two occasions – the Union Peace Conference in January 2016 and Panglong-21 in August – the next conference will have to start addressing the substantive issues. Assuming that a revised Framework for Political Dialogue can be agreed and reasonable inclusivity of armed groups can be achieved through an expansion in NCA signatories, participants will then need to start grappling with the substance. All agree this will be very challenging, and it will likely be many years before a comprehensive peace agreement can be reached. Three key questions arise:

  • Is a negotiated federal solution possible? This is the main demand of armed groups and ethnic leaders, and Suu Kyi has strongly committed to achieving “the democratic federal union of our dreams”. The military is far more cautious. The commander-in-chief did not use the term “federal” in his opening speech at Panglong-21, emphasising “peace and unity” and that armed struggle is inconsistent with democracy. However, the military is not rejecting federalism; the commander-in-chief signed the NCA, whose first point is to “establish a union based on the principles of democracy and federalism”, and a senior military officer used the term at Panglong-21.[fn]Aung San Suu Kyi, opening speech, Panglong-21, Naypyitaw, 31 August 2016. NCA Section 1(a); speech of Lt. General Yar Pyae, JMC chair, at Panglong-21, reported in “21st Century Panglong commences in Nay Pyi Taw”, Global New Light of Myanmar, 1 September 2016.Hide Footnote  The potential deal is federalism in return for disarmament of armed groups. However, this will be complicated given the number of armed groups and their divergent interests, and the extent of federal powers that military and government are ready to devolve is not yet clear. There are also hundreds of armed militias, some of which have ethno-nationalist positions, but most are primarily economic actors.[fn]For details, see John Buchanan, “Militias in Myanmar”, The Asia Foundation, July 2016.Hide Footnote

  • Can the concerns of sub-minorities be accommodated? One of the more intractable issues is likely to be their status. Federalism has tended to be conceived, in geographic terms, as devolution of powers to the existing seven ethnic states.[fn]Chin, Kachin, Kayah, Kayin, Mon, Rakhine and Shan.Hide Footnote  This alarms smaller minority groups within these states, who fear that political domination at the state level will replace domination by Naypyitaw. This was already clear from the speeches at Panglong-21, where specific claims for new states were made by the Wa, Ta’ang and Pao (all currently having self-administered areas within Shan State) and the Red Shan (in Kachin State and Sagaing Region, where they have no territorial designation). Many other potential claims can be anticipated.[fn]See comments of Sai Htay Aung (Red Shan), Khun Myint Tun (Pao) and U Yan Kyaw (Wa), Global New Light of Myanmar, 3 September 2016; and TNLA open letter, op. cit., which specifically calls for creation of a Ta’ang (Palaung) State.Hide Footnote  Shan and Kachin political and armed group leaders in general oppose these proposals.
     
  • Will any negotiated solution be regarded as legitimate and be implemented? Even if a reasonably inclusive process can be achieved and consensus reached on the complex substantive issues, many constituencies may feel marginalised by the process. Minority ethnic representation is limited to those that have armed groups or political parties that won seats (in a recent change Suu Kyi initiated, those that did not win legislative seats in 2015 have only a token number at the peace conference and no UPDJC representation).[fn]See “Kayah political parties boycott Panglong Conference”, Myanmar Times, 22 August 2016.Hide Footnote  Many influential ethnic parties won nothing in the NLD landslide and will have a minimal voice in the process; some minority groups are not represented by an armed group; and questions can be asked about how representative armed groups are of communities in their areas.

There is a fundamental doubt about whether state-based federal solutions can appropriately be negotiated between armed groups and government, in particular when civil society voices, women and youth feel marginalised in the process.[fn]“CSOs pine for seat at table”, Myanmar Times, 26 August 2016; statement by Alliance for Gender Inclusion in the Peace Process on Panglong-21, September 2016; “No women, no peace: Gender equality, conflict and peace in Myanmar”, Transnational Institute, 13 January 2016; “Youth ethnic alliance emerges after summit”, Myanmar Times, 3 August 2016.Hide Footnote  That process should be adjusted to ensure that it has broader legitimacy. Even where representation has strong legitimacy – for example, the NLD government’s support from the majority Burman group (and many others) – the population at large has had little engagement with the peace process and may oppose solutions that devolve too much political authority and economic control to minority areas. Minority communities will not necessarily see the NLD as representing their interests, even if they voted for it, because that vote was in many ways a referendum on military rule, reflecting determination to vote out the military-backed party.[fn]For discussion of the election outcome in ethnic areas and its interpretation, see Crisis Group Briefing, The Myanmar Elections, op. cit., Section IV.C; and “The 2015 general election in Myanmar: What now for ethnic politics?”, Transnational Institute, December 2015.Hide Footnote

While Suu Kyi’s focus has been on federalism and security – she initially proposed that the political dialogue deal directly with only those issues – minority communities have many other concerns. These include rights and discrimination, revenue sharing, natural resource management and language policy.[fn]For detailed discussion, see Crisis Group Report, Myanmar: A New Peace Initiative, op. cit., Section IV.Hide Footnote  Whether these are dealt with up-front as potentially more tractable confidence-building measures or sidelined by more fundamental issues can have a big impact on the dynamics of the peace process. Overlooking them would likely be a mistake.

C. The Political and Security Environment

Since the peace process was launched in 2011, it has had to face significant external and domestic challenges. Serious armed conflict on the ground and China’s role have been particularly important and are to some degree interlinked.

The most significant outbreak of conflict in recent years was the collapse of the KIO ceasefire in 2011, the seeds of which were sown prior to the 2010 election. Fighting resumed ahead of the formal launch of the peace process in August 2011, and a serious escalation in December 2012 threatened to derail it, but China’s intervention, prompted in part by fighting spilling over its border, pushed the sides back to the negotiating table.[fn]See Crisis Group Asia Briefing N°140, A Tentative Peace in Myanmar’s Kachin Conflict, 12 June 2013.Hide Footnote  Another major test came in April 2014, when serious clashes displaced some 5,000 civilians and eroded the trust of all parties in the NCA negotiations. The crisis deepened in November 2014, when an army mortar attack on a military training centre at KIO headquarters almost caused the talks to collapse. Serious fighting in the Kokang Self-Administered Zone between government forces and the MNDAA from February 2015 hardened opposing positions of the military and several armed groups over inclusivity, part of the reason why a number of groups were unwilling to sign the NCA that year.[fn]Crisis Group Briefing, Myanmar’s Peace Process, op. cit., Section II.D.Hide Footnote

With a fragile peace holding in parts of the borderlands and clashes ongoing in many others (Section II.B above), the peace process is likely to continue to be buffeted. Rigid timelines for Panglong-21 conferences risk becoming an obvious target for spoilers and an unsatisfactory framework for adjusting to unpredictable but inevitable escalations in the conflict. The military may feel less constrained by the peace process than under the previous government; given the power asymmetries, it is likely to continue pressing its ground advantage, especially with NCA non-signatories and in particular if the peace process moves slowly or it feels that armed groups are being obstructive.

China’s influence can have a big impact on ground dynamics and the peace process, given its considerable leverage over the groups on its border. It has regularly intervened, positively and negatively. Relations with the Thein Sein administration were often strained, starting with suspension of the Myitsone dam project in 2011 and difficulties with the Letpadaung copper mine – both major China-backed projects – and long delays in announcing that a Chinese company had won the tender for the Kyaukpyu deep-sea port and special economic zone, a major Chinese strategic interest.[fn]See Yun Sun, “Aung San Suu Kyi’s visit to Beijing: Recalibrating Myanmar’s China policy”, Transnational Institute, 16 August 2016.Hide Footnote  Myanmar’s markedly improved relations with the U.S. intensified China’s angst that it had lost its “traditional advantage”.[fn]“China’s engagement in Myanmar: From Malacca Dilemma to Transition Dilemma”, Transnational Institute, July 2016.Hide Footnote  The poor relations, combined with specific irritants such as Myanmar’s intrusion into Chinese airspace in 2015 to attack the MNDAA, a flood of refugees into China and Naypyitaw’s invitation to Japan and the West to become involved in the peace process, produced a negative stance toward the NCA, to the point that persistent allegations emerged that China was lobbying armed groups in 2015 not to sign.[fn]China has denied the allegations, which were made publicly by a member of the Myanmar Peace Centre and subsequently retracted, and privately to Crisis Group and others by a wide range of people connected to the peace process. Whether true or not, it is clear from talk with armed groups leaders at the time that there was no Chinese pressure to sign the NCA and massive private financial support from China that the authorities must have been aware of. See “Fraud probe alleges Chinese firm sent money to Myanmar insurgents”, Frontier Myanmar, 3 February 2016.Hide Footnote  

The situation has shifted significantly under the new government. China feels Suu Kyi gives more priority to the bilateral relationship, and it supports her peace overtures. At the July summit of armed group leaders hosted by the KIO, the Chinese special envoy publicly called on all groups to attend Panglong-21, and Beijing successfully put considerable pressure on several to do so. China has also given several million dollars to fund the JMC but remains uncertain about the trajectory of relations, the chances for success in the peace process and how many years that would take; it is thus likely to continue to balance support for Naypyitaw and maintaining ties with armed groups along its border.[fn]Yun Sun, “Aung San Suu Kyi visit to Beijing”, op. cit. Crisis Group interview, Myanmar expert on China, Yangon, September 2016.Hide Footnote

V. Conclusion

The Panglong-21 conference encapsulated both the significant advantages Suu Kyi has for forging peace and the enormous challenges she must surmount. The broad attendance of armed groups gives hope of a more inclusive, successful peace process, but it would be a mistake to think that the fundamental problems have become easier to solve. It will take difficult negotiations to convince most groups to sign the NCA, a sine qua non the government and military have each expressed. The announced scheduling of Panglong-21 conferences every six months artificially limits the flexibility required to secure signatures. Weak capacity in the government’s NRPC peace secretariat makes the job more difficult.

The government should consider adopting a less rigid timeframe and less unilateral approach and take steps to ensure it has the necessary support capacity in place. Armed groups need to recognise that, though they have legitimate concerns about the process, they may never get a better chance to negotiate a settlement. Aung San Suu Kyi has expressed firm support for a federal, democratic solution and has the political authority to deliver. Now is the time to start discussing the contours of that deal, rather than continuing to focus on preliminaries.

Yangon/Brussels, 19 October 2016

Appendix A: Map of Myanmar

Map of Myanmar. CRISIS GROUP

Appendix B: The Main Ethnic Armed Groups and their Ceasefire Status

  1. United Wa State Party (UWSP)
    Bilateral ceasefire: 6 September 2011. NCA-signatory: No
     
  2. National Democratic Alliance Army (NDAA, “Mongla group”)
    Bilateral ceasefire: 7 September 2011. NCA-signatory: No
     
  3. Democratic Kayin Benevolent Army (DKBA)
    Bilateral ceasefire: 3 November 2011. NCA-signatory: Yes
     
  4. Restoration Council of Shan State/Shan State Army-South (RCSS/SSA-South)
    Bilateral ceasefire: 2 December 2011. NCA-signatory: Yes
     
  5. Chin National Front (CNF)  
    Bilateral ceasefire: 6 January 2012. NCA-signatory: Yes
     
  6. Karen National Union (KNU)
    Bilateral ceasefire: 12 January 2012. NCA-signatory: Yes
     
  7. Shan State Progress Party/Shan State Army-North (SSPP/SSA-North)
    Bilateral ceasefire: 28 January 2012. NCA-signatory: No
     
  8. New Mon State Party (NMSP)
    Bilateral ceasefire: 1 February 2012. NCA-signatory: No
     
  9. Karen National Liberation Army Peace Council
    Bilateral ceasefire: 7 February 2012. NCA-signatory: Yes
     
  10. Karenni National Progressive Party (KNPP)
    Bilateral ceasefire: 7 March 2012. NCA-signatory: No
     
  11. Arakan Liberation Party (ALP)
    Bilateral ceasefire: 5 April 2012. NCA-signatory: Yes
     
  12. National Socialist Council of Nagaland-Khaplang
    Bilateral ceasefire: 9 April 2012. NCA-signatory: No
     
  13. Pao National Liberation Organisation (PNLO)
    Bilateral ceasefire: 25 August 2012. NCA-signatory: Yes
     
  14. All Burma Students Democratic Front (ABSDF)
    Bilateral ceasefire: 5 August 2013. NCA-signatory: Yes
     
  15. Kachin Independence Organisation (KIO)
    Bilateral ceasefire: (30 May 2012)*. NCA-signatory: No
     
  16. Ta’ang National Liberation Army (TNLA)
    Bilateral ceasefire: No. NCA-signatory: No
     
  17. Myanmar National Democratic Alliance Army (MNDAA, “Kokang group”)         
    Bilateral ceasefire: No†. NCA-signatory: No
     
  18. Arakan Army (AA)
    Bilateral ceasefire: No. NCA-signatory: No

* An agreement was signed on 30 May 2012. It was not a formal ceasefire, but contained inter alia a commitment to “efforts to achieve de-escalation and cessation of hostilities”.

† The MNDAA’s 1989 ceasefire ended after an army attack in 2009, with one faction being routed (and its leaders fleeing to China) and the other agreeing to become a Border Guard Force unit under partial army control. The routed faction subsequently reactivated, with support from other groups.

The United Nationalities Federal Council (UNFC) is an armed group umbrella organisation, whose seven members have not signed the NCA: SSPP/SSA-North, NMSP, KNPP, KIO, Lahu Democratic Union, Arakan National Council, Wa National Organisation. The last three do not have significant armed forces, so have not been directly included in the ceasefire process.

Appendix C: List of Acronyms

AA: Arakan Army

ABSDF: All Burma Students Democratic Front

ALP: Arakan Liberation Party

CNF: Chin National Front

DKBA: Democratic Kayin Benevolent Army, Democratic Kayin Buddhist Army

JMC: Joint Monitoring Committee

KIO: Kachin Independence Organisation

KNPP: Karenni National Progressive Party

KNU: Karen National Union

MNDAA: Myanmar National Democratic Alliance Army (Kokang)

NCA: Nationwide Ceasefire Agreement

NDAA: National Democratic Alliance Army (“Mongla group”)

NMSP: New Mon State Party

NRPC: National Reconciliation and Peace Centre

NSCN-Khaplang: National Socialist Council of Nagaland-Khaplang

PNLO: Pao National Liberation Organisation

RCSS: Restoration Council of Shan State

SSA-North: Shan State Army-North

SSA-South: Shan State Army-South

SSPP: Shan State Progress Party

TNLA: Ta’ang National Liberation Army

UNFC : United Nationalities Federal Council

UPDJC: Union Peace Dialogue Joint Committee

UWSP: United Wa State Party

A Rohingya refugee looks at the full moon with a child in tow at Balukhali refugee camp near Cox's Bazar, Bangladesh, December 3, 2017. REUTERS/Susana Vera TPX IMAGES OF THE DAY
Report 292 / Asia

Myanmar’s Rohingya Crisis Enters a Dangerous New Phase

The mass flight of Rohingya Muslims from Myanmar’s Rakhine State has created a humanitarian catastrophe and serious security risks, including potential cross-border militant attacks. The international community should press the Myanmar government to urgently implement the Annan commission’s proposals, including as regards discrimination, segregation and citizenship.

  • Share
  • Save
  • Print
  • Download PDF Full Report
  • What’s the issue?  The response of Myanmar’s military to militant group Arakan Rohingya Salvation Army’s (ARSA) August attacks has led to one of the most catastrophically fast refugee exoduses in modern times. More than 624,000 Rohingya Muslims have fled Myanmar to Bangladesh, creating the world’s largest refugee camp.
  • Why does it matter?  The eviction of the Rohingya community from Myanmar is far from the end of the crisis. The situation is transforming Myanmar’s domestic politics and international relations, and potential future cross-border attacks by ARSA militants could increase tensions between Myanmar and Bangladesh.
  • What should be done?  Imposing targeted sanctions can send an important message and potentially deter others from similar actions against minority communities. But they are unlikely to produce positive change in Myanmar. Even as they impose targeted sanctions, the international community should continue to provide humanitarian support for Rohingya refugees and resist pressure to disengage from the country.

Executive Summary

Three months after militant attacks triggered a brutal army operation targeting Rohingya Muslim communities in Myanmar’s northern Rakhine State, more than 624,000 have fled to Bangladesh, one of the fastest refugee exoduses in modern times. In addition to unimaginable human suffering, the crisis has transformed Myanmar’s domestic politics and international relations and will have a huge impact on the regional security landscape.

Myanmar is rapidly losing what remains of the enormous international good-will that its political transition had generated. State Counsellor Aung San Suu Kyi in particular has been widely criticised for failing to use her moral authority and domestic legitimacy to shift anti-Rohingya sentiment in Myanmar and the government’s current course. Meanwhile, the exodus continues and will likely soon reach its tragic end point: the almost complete depopulation of Rohingya from northern Rakhine State.

As the world struggles to define a response, and as the crisis enters a new, fraught and highly uncertain phase, several important elements need to be borne in mind. First, there needs to be continued insistence on the right of refugees to return in a voluntary, safe and dignified manner. At the same time, the grim reality is that the vast majority of the Rohingya in Bangladesh will not be going home any time soon. This presents the enormous humanitarian challenge of sustaining lives and dignity in the largest refugee camp in the world. It also presents grave political and security risks that need to be addressed, including potential cross-border attacks by the Arakan Rohingya Salvation Army (ARSA) militant group and possible transnational terrorism.

Second, it is important to recognise that Myanmar’s political direction has been set and will be extremely difficult to change. The strength of the national consensus is hard to overstate: the government, military and almost the entire population of the country are united on this issue as on no other in its modern history. This will make it extraordinarily difficult to move official policy. Any imposition of sanctions thus requires careful deliberation: they can help send a welcome signal that might deter others around the world contemplating similar actions, but they are unlikely to produce positive change in Myanmar and, depending on what precisely is done, could make the situation worse.

This report examines the lead-up to the ARSA attacks on 25 August 2017, revealing new and significant details about the group’s preparations, and the attacks themselves. This is based on research in Myanmar and Bangladesh since October 2016, including interviews with members of ARSA, analysis of WhatsApp messages sent by the group and its supporters, publicly-posted videos and interviews with villagers in Rakhine State and recently-arrived refugees in Bangladesh. Much of the research has been done by experienced personnel fluent in the Rohingya language.  The report also assesses the impact the crisis will have on Myanmar. Finally, it discusses some possible international policy responses.

Brussels, 7 December 2017

I. Background to the Crisis

While the current crisis is rooted in longstanding discrimination and denial of human rights, the immediate trigger was the emergence of a militant group within the Rohingya population in the north of Rakhine State.[fn]For detailed background on the situation in Rakhine State, see Crisis Group Asia Reports N°s 283, Myanmar: A New Muslim Insurgency in Rakhine State, 15 December 2016; 261, Myanmar: The Politics of Rakhine State, 22 October 2014; and 251, The Dark Side of Transition: Violence Against Muslims in Myanmar, 1 October 2013. For other recent Crisis Group reporting on Myanmar, see Asia Briefings N°s 149, Myanmar’s Peace Process: Getting to a Political Dialogue, 19 October 2016; 147, The Myanmar Elections: Results and Implications, 9 December 2015; also Asia Reports N°s 290, Buddhism and State Power in Myanmar, 5 September 2017; 287, Building Critical Mass for Peace in Myanmar, 29 June 2017; 283, Myanmar: A New Muslim Insurgency in Rakhine State, 15 December 2016; 282, Myanmar’s New Government: Finding Its Feet?, 29 July 2016.Hide Footnote This hardened national sentiment toward the Rohingya and shifted the calculus of the security forces.

Harakah al-Yaqin, subsequently rebranded in English as the Arakan Rohingya Salvation Army (ARSA), first began organising itself after deadly communal violence in 2012. It launched its initial attacks – coordinated assaults on the Border Guard Police (BGP) headquarters and two other bases – on 9 October 2016. Previous armed militant groups had been based in the hills (the Arakan mujahidin in the 1950s), or launched hit-and-run attacks from across the border in Bangladesh (for example the Rohingya Solidarity Organisation in the 1990s). In contrast, ARSA operates from within Rohingya villages, using cells of villagers who have been given some basic training but most of whom do not have access to firearms, only bladed weapons and some improvised explosive devices (IEDs).[fn]See Crisis Group Report, A New Muslim Insurgency in Rakhine State, op. cit.Hide Footnote

In response to the October 2016 attacks, the military deployed overwhelming retaliatory force against nearby villages, followed by extensive “clearance operations” – brutal counter-insurgency operations that the military has used for decades in other parts of the country – with the stated purpose of recapturing the dozens of small arms and thousands of rounds of ammunition looted by ARSA.[fn]For details on the Myanmar military’s counterinsurgency approach, known as the “four cuts”, see Martin Smith, Burma: Insurgency and the Politics of Ethnicity, 2nd ed. (London, 1999), p. 288 ff.; Andrew Selth, Burma’s Armed Forces (Norwalk, 2001), pp. 91-92; and Maung Aung Myoe, “Military Doctrine and Strategy in Myanmar”, Strategic and Defence Studies Centre, working paper 339, 1999, p. 10.Hide Footnote When troops came under attack from militants and villagers and a senior officer was killed, the military further escalated, including the use of helicopter gunships in civilian areas. Over the following weeks, tens of thousands of Rohingya fled to Bangladesh and security forces burned down several thousand homes.[fn]Ibid.Hide Footnote A United Nations (UN) human rights office report found the “very likely commission of crimes against humanity”.[fn]“Interviews with Rohingyas fleeing from Myanmar since 9 October 2016”, Flash Report, UN Office of the High Commissioner for Human Rights (OHCHR) mission to Bangladesh, 3 February 2017.Hide Footnote A retired senior army officer noted that it would have been more effective to use the police to achieve the operation’s stated purpose of recovering the looted weapons and ammunition (most were not found).[fn]Crisis Group interview, Yangon, March 2017.Hide Footnote

In the months following the October 2016 attacks, ARSA set about consolidating its authority in Rohingya villages in northern Rakhine and preparing for the next round of attacks. It did this through the targeted killings of dozens of Rohingya men with links to the authorities (such as village heads, other local administrators and suspected informers), ramped up training in the hills as well as IED production in safe houses. The authorities were aware of these developments, with the state media reporting many of the killings as well as the discovery of IED factories. For them, the next ARSA attacks were seen as a matter of when, not if.[fn]Crisis Group interviews, government officials and military officers, Yangon, Naypyitaw and Sittwe, November 2016-August 2017; “Rakhine slayings by insurgents”, Global New Light of Myanmar (GNLM), 22 July 2017. GNLM is the state-owned English-language daily. For details on discovery of IED production sites, see section II below.Hide Footnote  

II. Build-up to the Crisis

In the months before the August 2017 ARSA attacks, a series of incidents suggested an uptick in ARSA training and preparation, putting Rakhine Buddhist villagers and the security forces on edge:

  • On 4 May, the accidental detonation of an IED during an ARSA explosives training course in Kyaung Taung village tract (north Buthidaung) killed seven men including the instructor, and injured at least five others. According to a reliable source close to the events, the instructor was Pakistani, not Rohingya. He was badly injured and died in Padakar Ywar Thit village tract (Maungdaw) while being carried to Bangladesh for treatment.[fn]Crisis Group interviews, local villagers with direct knowledge of the events, Rakhine State, May 2017. It appears there was a subsequent – possibly related – mass killing by the army of “at least scores” of Rohingya in an adjacent village (Min Gyi, or Tula Toli) on 30 August. “‘My World Is Finished’: Rohingya Targeted in Crimes Against Humanity in Myanmar”, Amnesty International, 18 October 2017, p. 21.Hide Footnote The people carrying him asked a village head to arrange his burial in a local cemetery but after being informed of the situation, security officials arrested the village head and took the body to Buthidaung hospital. These officials were the source of domestic Myanmar media reports some days later about the death of a foreign militant.
  • On 7 May, security forces investigating the IED detonation discovered the training camp and bomb-making materials. Six days later, the government announced it had found the bodies of five victims buried nearby, which they said included two foreigners. This prompted security forces to undertake violent evictions and clearance operations in the area (particularly around adjacent Tin May village tract), killing several people and prompting some families to flee to Bangladesh in May and June.[fn]While the government says two foreigners were killed in the 4 May incident, ARSA sources say there was only one, the Pakistani trainer who died. Six other people died, four on the spot and one later at a medical facility in Bangladesh. Five injured people received treatment at different medical facilities in Bangladesh; three were reportedly arrested by the Bangladeshi authorities. Crisis Group interviews, medical staff, Bangladesh, May 2017; refugees from Tin May, Bangladesh, May-July 2017. See also “Five Bodies Found in Buthidaung”, The Irrawaddy, 15 May 2017; “Five bodies unearthed near 5 May explosion site in Buthidaung”, GNLM, 16 May 2017.Hide Footnote
  • On June 20-21, the government reported that security forces had killed three men while clearing a likely ARSA training camp in the mountains near Sein Hnyin Pyar village tract (south Buthidaung).[fn]Crisis Group interviews, local villagers, June 2017; “Terrorist training camps, guns uncovered in Mayu Mountains”, GNLM, 22 June 2017.Hide Footnote
  • On 24 June, four Rakhine Buddhist villagers came across bomb-making material while foraging in Kyun Pauk Pyu Su village tract (north Maungdaw). ARSA members shot two of them dead; the two others, one of whom was injured, fled and alerted authorities. However, ARSA members apparently removed the incriminating material before the security forces reached the spot. This was the first known case of ARSA killing non-Rohingya civilians, and significantly increased anxiety among Rakhine Buddhist villagers; some 200 fled to Maungdaw town, fearing ARSA attacks. On 27 June, security forces in the area were placed on high alert; on 30 June, senior government officials in Naypyitaw discussed the situation at a “special meeting on Rakhine State”.[fn]Crisis Group interview, ARSA member with knowledge of the events, June 2017. See also, “Four local ethnic people were attacked by swords and killed two”, GNLM, 26 June 2017; “Troops in Myanmar’s Rakhine on high alert after killings of Rohingya”, Reuters, 27 June 2017; “Special meeting on Rakhine issue held”, GNLM, 1 July 2017. The “special meeting” comprised the president, State Counsellor Aung San Suu Kyi, vice presidents one and two, the legislative speakers, deputy commander-in-chief, relevant ministers, and national security adviser.Hide Footnote
  • On 1 August, authorities reported that an IED accidentally exploded at an ARSA safe house in Pan Taw Pyin village tract (Maungdaw) and that they found explosives and other bomb-making material at the house.[fn]Crisis Group interviews, local Rohingya villagers, August 2017; “IED explodes in Maungdaw”, GNLM, 2 August 2017.Hide Footnote Two days later, eight members of the Mro ethnic group, both men and women, were killed in the hills of Maungdaw township. The government immediately blamed ARSA, although some local villagers say the killings were related to the illicit methamphetamine trade.[fn]According to local sources, the area is on a methamphetamine smuggling route from Buthidaung to Bangladesh, and there had been previous tensions between the Mro village and a nearby NaTaLa (Buddhist resettlement) village, but generally good relations with nearby Rohingya villages; the method of killing of the Mro was not consistent with ARSA assassinations, which normally involve a machete cut to the neck. Crisis Group interviews, Rohingya villagers in the area, August 2017.Hide Footnote
  • On 4 August, BGP clashed with a group of villagers in Auk Nan Yar village tract, Rathedaung township, firing a dozen or more shots while trying to disperse a 300-strong crowd angry over the arrest of villagers suspected of being associated with ARSA, including a prominent local imam. During the clash, one of the suspected militants escaped; local villagers reported several injuries from gunshots, including four people taken to Bangladesh for treatment.


There were already significant tensions in the area. On 27 July 2017, a Rakhine villager had gone missing while foraging in nearby Chut Pyin village tract. Three days later, while searching in the surrounding hills, security forces and villagers discovered a stash of tarpaulins and food, including World Food Programme (WPF)-branded energy biscuits (see section V.A below), which they took to be an ARSA camp. Believing militants killed the missing person, Rakhine villagers declared a boycott of Muslims in the area. In the nearby village of Zay Di Pyin, Buddhist villagers blocked all access roads with barbed wire and prevented residents from going to work or accessing the mosque, food markets and water sources.[fn]Crisis Group interviews, local villagers, August 2017. See also “Tents of violent attackers discovered in Mayu Mountain”, GNLM, 1 August 2017; “Attack on police force arresting financial supporter of violent attackers in Yathedaung”, GNLM, 5 August 2017; “Rohingya villagers blockaded amid fresh tensions in Myanmar’s Rakhine – residents”, Reuters, 22 August 2017.Hide Footnote According to various sources, on 27 August, security forces and local vigilantes perpetrated a mass killing of “at least scores” of Rohingya villagers in Chut Pyin.[fn]Amnesty International, op. cit., p. 13.Hide Footnote

These events provoked heightened nervousness. On 9 August 2017, the commander-in-chief and other senior military officers met with leaders of the Arakan National Party, the largest party in Rakhine State – a rare meeting between the top brass and a political party. The party expressed concerns about the security situation in northern Rakhine and requested the arming of local Rakhine Buddhist militias. That same day, State Counsellor Aung San Suu Kyi convened a ministerial meeting on the security situation in Rakhine to discuss the recent killings and rising tensions. The following day, the government highlighted its deployment of some 500 troops to northern Rakhine to reassure local non-Muslim villagers and conduct patrols in the mountains between Maungdaw and Buthidaung where militants were suspected of having established training camps.[fn]See Senior General Min Aung Hlaing, Facebook post, 10 August 2017, http://bit.ly/2yqQYSA; “State Counsellor, Union Ministers hold talks on security in Rakhine State”, GNLM, 10 August 2017; “Myanmar Army Deployed in Maungdaw”, The Irrawaddy, 11 August 2017.Hide Footnote

The escalatory dynamic was well under way. On 16 August, ARSA uploaded a video of its commander, Ata Ullah, flanked by armed fighters and warning the Myanmar military to demilitarise northern Rakhine State and end abuses of Rohingya; he specifically cited the blockade of Rohingya villagers in Zay Di Pyin. He reiterated that the group had no relation with international jihadist groups and said that, contrary to government assertions, it did not target Rakhine civilians.[fn]ARSA Commander Addresses Rohingya diaspora & the world; Warns Myanmar military”, video, YouTube, 16 August 2017.Hide Footnote  

III. ARSA Attacks and Military Response

A. ARSA Attacks

In the early hours of 25 August 2017, from 1am until dawn, ARSA launched attacks on some 30 BGP posts and an army base.[fn]The information in this sub-section comes from Crisis Group interviews with ARSA members, Rohingya in Rakhine State and refugees in Bangladesh, August-October 2017; and from analysis of WhatsApp audio messages sent by Ata Ullah and others.Hide Footnote Their human wave attacks in some cases involved hundreds of people, mostly untrained local villagers armed with farm tools as well as some hand-held and remote-detonated IEDs. A small number of further clashes occurred over the next several days. The official death toll was fourteen members of the security forces, one government official and 371 people the government characterised as militants.[fn]Death toll listed in “Humanitarian aid provided to displaced people without segregation”, GNLM, 6 September 2017.Hide Footnote

ARSA initiated the attacks via a WhatsApp audio message delivered shortly after 8pm on 24 August. It instructed cell leaders to mobilise all male villagers over the age of fifteen, assemble in pre-planned locations with whatever sharp objects were available and attack designated targets. Many ordinary villagers apparently responded to the call, which was often conveyed by respected local Islamic clerics (known as “Mullahs” or “Maulvis”) or scholars (“Hafiz”) who seemingly made up most cell leaders and who enjoy considerable religious and community authority. Many untrained villagers were provided with IEDs for use in the attacks.

The targets were mostly small police posts and checkpoints, except for the army base in Chin Tha Mar village (near Nga Yant Chaung or Taung Bazar), Buthidaung township, though not many villagers appear to have joined this attack, which was quickly overpowered. ARSA members claim they planned to attack additional targets but that some police posts were deserted when militants reached them. Other targets were more heavily defended than expected and the attackers suffered heavy casualties. The security forces assert that they had several hours advance warning; whether accurate or not, they clearly were expecting attacks at some point.

On 25 August, ARSA issued a series of messages apparently intended both to instil confidence and resolve among its members and followers and to promote and glorify martyrdom, the goal being to encourage lightly armed male villagers to participate in highly risky attacks. Some messages falsely claimed that ARSA was taking control of the areas it attacked. Members were also reassured that armed reinforcements had been dispatched; they never arrived.

On 28 August, Ata Ullah issued WhatsApp audio messages instructing his followers to burn down Rakhine Buddhist villages with Molotov cocktails. This was in direct contradiction to the group’s repeatedly stated policy and prior approach, which was to refrain from attacking non-security targets. The reason for this change is not clear, though it may have been because non-Rohingya vigilantes from nearby villages were helping the military burn Rohingya villages during clearance operations. ARSA might have concluded that Rakhine and other non-Rohingya villagers therefore were a fair target.[fn]The Rakhine, a predominantly Buddhist ethnic group, make up the majority of the non-Rohingya population in northern Rakhine State, but numerous other ethnic groups live in the area and some have also reportedly been involved in vigilante attacks. See Amnesty International, op. cit.Hide Footnote In the event, the order does not appear to have been widely acted upon as only three non-Rohingya villages are known to have been attacked or burned down by Rohingya.[fn]On 28 August 2017, there were deadly attacks on the Rakhine Buddhist village of Auk Pyu Ma and the Mro village of Khon Taing (Pa Da Kar Ywar Thit village tract), as well as an earlier attack on the Daingnet village of Aung Zan (all in Maungdaw township). ARSA’s involvement in attacks on two Hindu villages (Myo Thu Gyi and Kha Maung Seik) is alleged, but not confirmed.Hide Footnote

One particularly high-profile case is the alleged massacre by ARSA of dozens of Hindu men and women in Kha Maung Seik (also known as Fakira Bazar) in Maungdaw township. Conflicting accounts of the incident and of who was responsible have surfaced. Survivors who fled to Bangladesh initially told Bangladeshi journalists in late-August that the killers were Rakhine militants; others said later that they wore masks, preventing identification. The first report of the incident by Myanmar media on 5 September 2017 attributed the killings to ARSA, based on interviews with survivors in Myanmar. A more detailed account reaching the same conclusion was posted on Facebook on 13 September by a Rakhine nationalist parliament member who investigated the incident. The security forces reported finding and exhuming a mass grave containing the victims’ bodies on 24 September; these subsequently were cremated. It is not clear what forensic evidence remains.[fn]“Hindus too fleeing persecution in Myanmar”, The Daily Star (Bangladesh), 31 August 2017; “Mystery surrounds deaths of Hindu villagers in Myanmar mass graves”, The Guardian, 12 October 2017; “Dozens of Hindus Killed in Maungdaw: Relatives”, The Irrawaddy, 5 September 2017; Kyaw Zaw Oo (Arakan National Party, Sittwe-2 constituency), Facebook post, 12 September 2017, http://bit.ly/2ApcmZ0; See “45 Hindu corpses cremated”, GNLM, 29 September 2017.Hide Footnote

B. Catastrophic Military Response

A brutal military response that failed to discriminate between militants and the general population, followed by continued insecurity and restrictions that have imperilled livelihoods, has driven more than 624,000 Rohingya into Bangladesh. This is one of the fastest refugee exoduses in modern times and has created the largest refugee camp in the world. A large proportion of Rohingya villages in the area have been systematically reduced to ashes by both troops and local Rakhine vigilante groups that were equipped and supported by the military following the 25 August ARSA attacks.

This [Rohingya crisis] is one of the fastest refugee exoduses in modern times and has created the largest refugee camp in the world.

Grim details of the military and local vigilante campaign of violence, described by the UN as “a textbook example of ethnic cleansing” (a characterisation that has now been echoed by the United States) and by human rights groups as crimes against humanity, have been set out in a series of detailed reports by these organisations. They document widespread, unlawful killings by the security forces and vigilantes, including several massacres; rape and other forms of sexual violence against women and children; the widespread, systematic, pre-planned burning of tens of thousands of Rohingya homes and other structures by the military, BGP and vigilantes across northern Rakhine State from 25 August until at least October 2017; and severe, ongoing restrictions on humanitarian assistance for remaining Rohingya villagers.[fn]See, in particular, Amnesty International, op. cit., as well as “Destroyed areas in Buthidaung, Maungdaw, and Rathedaung Townships of Rakhine State”, UN Institute for Training and Research (UNITAR)/UNITAR’s Operational Satellite Applications Programme (UNOSAT) imagery analysis, 16 November 2017; “Burma: New Satellite Images Confirm Mass Destruction”, Human Rights Watch, 17 October 2017; “Mission report of OHCHR rapid response mission to Cox’s Bazar, Bangladesh, 13-24 September 2017”, OHCHR, October 2017; “U.N. sees ‘textbook example of ethnic cleansing’ in Myanmar”, Reuters, 11 September 2017.Hide Footnote

Crisis Group’s analysis of population data for northern Rakhine State from various sources suggests that around 85 per cent of the Rohingya population in these three townships has fled to Bangladesh over the last twelve months, leaving behind only 100,000-150,000. There are also some 320,000 Muslims in central Rakhine State, many but not all of whom identify as Rohingya; 120,000 of these have been confined to displacement camps since communal violence in 2012.[fn]Analysis based on 2014 census estimates of non-enumerated (Rohingya) population; government 2016 General Administration Department figures; UN figures for camp populations; and community estimates of Rohingya population by township, all broadly consistent. There are 20,000-plus Muslims in southern Rakhine, where communal relations tend to be better.Hide Footnote

The three northern townships were impacted in somewhat different ways:

  • Maungdaw township was the focus of ARSA attacks on 25 August 2017 and in October 2016. It had the largest Rohingya population and shares the longest border with Bangladesh (river and land, as well as adjacent seaboard). It bore the brunt of the military response and it appears that almost the entire township has been depopulated of Rohingya, apart from some parts of Maungdaw town and a small number of villages.[fn]UNITAR/UNOSAT imagery analysis, op. cit.Hide Footnote
  • Buthidaung township has historically been less affected by violence and displacement than Maungdaw. It also shares a land border with Bangladesh, along the hilly and hard to access northern part of the township; most of the population lives in the south. There were no ARSA attacks here in October 2016, only a small number in August 2017, to which the initial military response appears to have been more localised and limited. Far fewer Rohingya villages were initially burned here compared to Maungdaw. While the military response and burnings triggered some immediate departures to Bangladesh, the vast majority left later to escape untenable living conditions: continued burning of villages and attacks or threats by Rakhine vigilantes plus new, severe movement restrictions that deprived people of their normal means of survival from farming, fishing, foraging and trading. With humanitarian assistance also heavily restricted, communities came to the decision in late September 2017 that they had no choice but to make the long and dangerous journey in large groups, over the mountains to Maungdaw and on to Bangladesh.[fn]Crisis Group interviews, villagers, Buthidaung and Bangladesh, August-November 2017.Hide Footnote
  • Rathedaung township, unlike Maungdaw and Buthidaung, is a Rakhine Buddhist-majority area that does not share a border with Bangladesh. One of the three October 2016 ARSA attacks was here, in Koe Tan Kauk (close to the boundary with Maungdaw); the government claimed an ARSA attack in this area on 25 August 2017. Subsequent anti-Rohingya violence and threats had a much greater communal component. Nearly all Rohingya in the township have now fled to Bangladesh, apart from five villages with no viable escape route and only very limited access to food or humanitarian support.[fn]“‘We will kill you all’ – Rohingya villagers in Myanmar beg for safe passage”, Reuters, 17 September 2017.Hide Footnote


In addition to the massive Rohingya exodus, the crisis also led to the displacement of some 27,000 non-Rohingya villagers and government employees in northern Rakhine, most of whom fled the initial ARSA attacks and subsequent clashes. Nearly all moved or were evacuated inland, to the main towns of Maungdaw, Buthidaung and Sittwe. The government is now strongly encouraging them to return and begin rebuilding their damaged or destroyed houses.[fn]“Ethnic IDPs who fled homes due to terrorist attacks”, GNLM, 6 September 2017; “Rakhine State Govt to Close Hindu, Ethnic Arakanese Displaced Person Camps”, The Irrawaddy, 30 October 2017.Hide Footnote

Since 25 August 2017, the government has blocked access to northern Rakhine State by the UN and most other humanitarian actors. The Red Cross movement (the International Committee, International Federation, and Myanmar Red Cross Society) have been permitted to work, although they face delays and restrictions as well as enormous logistical challenges in reaching populations in need; they have called for other humanitarian actors to be granted access. On 6 November, the World Food Programme was able to resume food aid to Rohingya and non-Rohingya communities through the government but with no staff access to monitor distribution directly.[fn]Crisis Group interviews, international humanitarian staff, Yangon, September-November 2017.Hide Footnote

IV. Dangers Ahead

A. Repatriation Remains a Distant Hope

More than 624,000 Rohingya have fled to Bangladesh in the last three months. Myanmar’s neighbours and other members of the international community must insist on their right of return and press the Myanmar authorities to create conditions conducive to a voluntary and safe repatriation. At the same time, prospects are extremely dim for the return of any significant number of Rohingya refugees to their home areas in Myanmar in the short or medium term.

Myanmar and Bangladesh signed a repatriation agreement on 23 November 2017 in Naypyitaw.[fn]“Arrangement on Return of Displaced Persons from Rakhine State”, 23 November 2017.Hide Footnote While it was politically expedient for both sides – Bangladesh to signal that it will not host the refugees indefinitely, and Myanmar to respond to charges of ethnic cleansing and ease pressure for action – it should be seen as a statement of intent rather than a sign that return is imminent. On paper, the criteria for returnees to be accepted by Myanmar are not too onerous: they need to have left Myanmar after 9 October 2016 (ruling out historical caseloads) and to provide evidence of bona fide residence in Myanmar, with no need for any particular documentation (an address should be sufficient).

But the main obstacle to repatriation is that most [refugees] are very unlikely to want to do so.

But the main obstacle to repatriation is that most are very unlikely to want to do so (according to the agreement, returns must be voluntary). The conditions on the ground in northern Rakhine are far from conducive, and the exodus of deeply traumatised refugees continues. There is lack of clarity from Myanmar on whether they would be allowed to return to their villages of origin and reclaim their farmland. The agreement also provides for the issuance of National Verification Cards at the point of return – a document most Rohingya reject out of fear that it will codify second-class citizenship status. The government and security forces have expressed concern about the presence of “terrorists” (that is, ARSA) or their supporters among the refugees, warning they would arrest such individuals upon return, which suggests returnees will be subject to extreme scrutiny or vetting. Another major obstacle is that Rakhine Buddhist leaders and communities are strongly opposed to the return of any Rohingya refugees.

A repatriation effort on this scale would overwhelm Myanmar’s capacity and resources.

Even if these obstacles could be overcome, a repatriation effort on this scale would overwhelm Myanmar’s capacity and resources; a senior official asserted that only 300 could be processed per day. Myanmar has consistently declined any role for the UN Refugee Agency, which could mobilise the necessary support as well as credibility in the eyes of the Rohingya and internationally; the bilateral agreement does not require it.[fn]Crisis Group interviews, Rohingya refugees, Bangladesh, September-November 2017; “Sales from Maungtaw paddy kept as national budget”, GNLM, 12 November 2017; “Govt Suggests Possible Daily Repatriation of 300 Rohingya Refugees”, The Irrawaddy, 30 October 2017; “Tensions over Rohingya return highlight donor dilemmas”, Nikkei Asian Review, 27 October 2017; “Returning Rohingya may lose land, crops under Myanmar plans”, Reuters, 22 October 2017; “‘Caged Without a Roof’: Apartheid in Myanmar’s Rakhine State”, Amnesty International, 21 November 2017.Hide Footnote

Fundamentally, neither the government nor security forces possess the political will to create conditions for voluntary return and implement a credible and effective process to that end. This raises the prospect of a long-term concentration of hundreds of thousands of traumatised Rohingya confined to squalid camps in Bangladesh, with no obvious way out or hope for the future. That would not only be a human tragedy, but also a grave security threat. Such a context would be ripe for mobilising further violent responses and potential transnational jihadist recruitment.

B. Security Risks

ARSA may still be reeling from the enormity of the crisis that its attacks triggered; tellingly, no videos of Ata Ullah have been released since 28 August 2017. Still, it appears determined to regroup and remain relevant. A Twitter account that likely represents the group remains active. It issued a statement on 7 October 2017 announcing the end of its unilateral ceasefire two days later, putting pressure on the group to demonstrate its continued capabilities. ARSA has not launched any new attack since then, but will undoubtedly strive to do so.[fn]“Assessment of the humanitarian pause”, ARSA press statement, 7 October 2017. The Twitter handle is @ARSA_Official; the 28 August video is available at http://bit.ly/2hn2V5a.Hide Footnote

Given how ARSA is organised, this will require a significant departure from its previous way of operating. Rather than basing uniformed, armed militants in camps, ARSA has, to date, organised cells within hundreds of villages, led by a network of respected local leaders, including young Mullahs. It attempted to incite a general uprising among the population, overrunning police posts using overwhelming numbers of ordinary villagers with farm tools, rather than military might. Yet operating under cover of the civilian population is no longer possible given that few Rohingya villages remain. Most of the group’s organisers and fighters are now in the Bangladesh camps, having fled along with the rest of the population.[fn]Crisis Group interviews, ARSA members and well-placed individuals in the camps, Bangladesh, September-November 2017. For details on ARSA organisation, see Crisis Group Report, Myanmar: A New Muslim Insurgency in Rakhine State, op. cit.Hide Footnote

The group may thus shift to cross-border attacks, which would require different training, access to weapons as well as operating space in Bangladesh. Acquiring that space might now be more realistic given Bangladesh’s anger and frustration toward Myanmar. If ARSA launches cross-border attacks, it could aim at opportunistic security targets in northern Rakhine or turn to attacking any non-Muslim villagers resettled on Rohingya lands, an easier target.

Inevitably, such attacks would have profoundly negative consequences. They would escalate tensions between Bangladesh and Myanmar and could potentially lead to clashes between the two countries’ militaries. New ARSA attacks would reinforce anti-Rohingya sentiment within Myanmar and prompt heightened security measures that would further diminish prospects for an eventual refugee return. Moreover, attacks against Rakhine Buddhist villagers would inflame anti-Muslim sentiment in general and could tip central Rakhine State, so far untouched by the recent violence, into crisis. Intercommunal relations are now on a knife-edge, which further constrains the ability of Muslims in the area to move freely and access services and livelihoods. Communal attacks there are a very real threat, and unlike their coreligionists in northern Rakhine, these communities have no viable escape routes.

While new ARSA attacks could provoke further violence, international jihadist groups represent a far bigger security threat to Myanmar.

Finally, while new ARSA attacks could provoke further violence, international jihadist groups represent a far bigger security threat to Myanmar. The country has justified what it calls clearance operations by arguing the nation faces a terrorist threat. This could be a self-fulfilling prophecy. The plight of the Rohingya has captured the attention of the Muslim world, becoming a cause célèbre like perhaps no other since Kosovo.

Al-Qaeda, Islamic State and other jihadist groups, which have long issued statements of solidarity with the Rohingya for propaganda purposes, are now calling directly for attacks on Myanmar and its leaders. Most recently, on 27 October 2017, the media arm of al-Qaeda in the Indian subcontinent released a video message from the group’s leader, Abu Syed al-Ansari, repeating calls for a jihad against Myanmar in support of the Rohingya. Myanmar is not prepared to prevent or deal with such an attack, which could be directed or merely inspired by these jihadist groups. Any attack, particularly on a religious target in a major city, would shred the fraught relations between Buddhists and Muslims across Myanmar, potentially sparking widespread communal violence; there are Muslim communities in most cities and many rural areas in Myanmar.[fn]See “Bangladesh dragging feet over repatriating Rohingya refugees, says Myanmar”, Reuters, 1 November 2017. For examples of earlier propaganda statements see Crisis Group Report, Myanmar: A New Muslim Insurgency in Rakhine State, op. cit., section V.E.Hide Footnote

C. Impact within Myanmar

Extreme Buddhist nationalist sentiment, a growing concern in Myanmar in recent years, has contributed to – and been reinforced by – the current crisis. This has included anti-Rohingya hate speech in state media under the civilian government’s editorial control and in sermons by prominent Buddhist monks.[fn]For background, see Crisis Group Report, Buddhism and State Power in Myanmar, op. cit. See also a 2016 editorial (in Burmese and English) referring to the Rakhine State violence as caused by “detestable human fleas” that “we greatly loathe for their stench”; “A flea cannot make a whirl of dust, but …”, GNLM, 27 November 2016.Hide Footnote

A sermon by Sitagu Sayadaw, one of Myanmar’s most revered monks and a leading doctrinal authority, is particularly alarming. Preaching to military officers at a garrison and training college in Kayin State on 30 October 2017, he urged unity between the military and monkhood, then appeared to provide a religious justification for the mass killing of non-Buddhists. He recounted a well-known fifth century legend from Sri Lanka commonly used in Myanmar to justify violence in defence of the faith, telling the soldiers that no matter how much they had to fight, they should remember that non-Buddhists killed were “not fully human”. The sermon and local media reporting of it have been widely shared on social media, with many Myanmar people expressing support, though some have voiced unease or opposition.[fn]See Matthew J. Walton, “Religion and Violence in Myanmar: Sitagu Sayadaw’s Case for Mass Killing”, Foreign Affairs, 6 November 2017.Hide Footnote

The government and military’s repeated, blanket denials of wrongdoing, widely disseminated in English and Burmese via state media, further reinforce a climate of impunity. This is particularly dangerous given that negative sentiments toward the Rohingya population are widespread at all levels of the military and in society as a whole. A recent editorial in the state paper dismissed “baseless accusations against the Myanmar Armed Forces” and stated that “it certainly does not take a legal expert to come to the conclusion that all those village[r]s who took part in the raids are also punishable under the anti-terrorism law. This fact may perhaps explain why nearly half-a-million people decided to cross over to … Bangladesh”.

A detailed internal investigation by the military concluded that troops fired “not a single shot” on civilians and that “all security members … strictly abided by the orders”, a further signal of impunity. In a 21 September speech to northern Rakhine State troops, Commander-in-Chief Min Aung Hlaing honoured “brilliant efforts to restore regional peace, security” and warned that a “race cannot be swallowed by the ground but only by another race” (a well-known Burmese saying that is also the motto of the immigration department). In a 15 November meeting with U.S. Secretary of State Tillerson, he stated that those who had fled to Bangladesh were ARSA terrorists and their families.[fn]“Rakhine State affair and cooperation”, GNLM, 2 November 2017; “Troops did not commit sexual violence nor killed civilians: Investigation Team”, GNLM, 14 November 2017, p. 10; Senior General Min Aung Hlaing, Facebook posts, 21 September 2017, http://bit.ly/2hCQq9o and 16 November 2017, http://bit.ly/2muFJY4. See also the dated immigration ministry website at http://bit.ly/259WAfy.Hide Footnote

Beyond the risk of further abuses against the Rohingya, the authorities have reinforced an ugly strand of nationalism that will outlast the current crisis and could be channelled to target other minorities.

Beyond the risk of further abuses against the Rohingya, the authorities have reinforced an ugly strand of nationalism that will outlast the current crisis and could be channelled to target other minorities. At a minimum, it will be more difficult for national leaders to make the necessary concessions in the peace process of greater minority rights and political and economic devolution.[fn]For details on the peace process with ethnic armed groups, see Crisis Group Report, Building Critical Mass for Peace in Myanmar, op. cit. ARSA is not (and likely never will be) part of the peace process, given that the Rohingya are not a recognised ethnic group.Hide Footnote This could undermine prospects for a stable, peaceful and more prosperous future, and thus imperil the country’s political transition or significantly shift its landing spot.

The crisis also will define the country in the eyes of much of the world for years to come. This will have a negative impact on trade, investment, tourism and global good-will, at a time when Myanmar is emerging from decades of isolation from the West. This is in turn likely to feed anti-Western sentiment, leading to greater estrangement and potentially cementing the country’s status as a pariah. The government’s priority long-term aims – balancing China’s geostrategic influence, integrating into the global economy and rehabilitating the military’s international image – may now be all but impossible to achieve.

V. Government and International Response

A. Government Position

On the day of the attacks, the government declared ARSA a terrorist group under domestic law. It issued a warning to the media to refer to ARSA as “extremist terrorists” rather than use terms such as “insurgents”. It claimed that international NGOs may have been collaborating with ARSA and that World Food Programme (WFP) and United States Agency for International Development (USAID) food aid had been diverted to the group. The government also stated that ammonia and tubes provided by development agencies for construction had been turned into IEDs. These statements set the tone for Myanmar’s escalatory response to the attacks and uncompromising attitude toward the UN and humanitarian agencies.[fn]“Anti-Terrorism Central Committee Order No. 1/2017”, 25 August 2017, under 2014 Anti-Terrorism Law, §72(B); “Warning in relation with extremist terrorists”, GNLM, 28 August 2017; “Terrorist hideouts discovered, items provided by int’l organisations found”, GNLM, 30 August 2017.Hide Footnote

Allegations of aid agency collusion were condemned by the U.S. ambassador to Myanmar as “absurd” and by the UN Human Rights chief as “irresponsible”, as they placed humanitarian staff “in danger and may make it impossible for them to deliver essential aid”. The accusations resulted in a boycott of aid agencies by their local contractors in Rakhine State and shipments came under mob attack. The government blocked access to northern Rakhine for all organisations (except the Red Cross) and most media.[fn]“US Ambassador Rejects Govt Implication of Aid Agencies in Rakhine Attacks”, The Irrawaddy, 31 August 2017; “‘Humanitarian catastrophe’ unfolding as Myanmar takes over aid efforts in Rakhine state”, The Guardian, 15 September 2017; “Myanmar police fire warning shots in Rakhine as mob attacks aid boat”, Agence France-Presse, 21 September 2017.Hide Footnote

The government blocked access to northern Rakhine for all organisations (except the Red Cross) and most media.

On 19 September and 12 October 2017, Aung San Suu Kyi addressed the Rakhine crisis in speeches that were criticised internationally, but gained strong local support. She questioned why Rohingya were fleeing, saying there were “allegations and counter-allegations” and claiming many Muslim villages were untouched and peaceful. She also announced the creation of a national fund for Rakhine State under her direction – the Union Enterprise for Humanitarian Assistance, Resettlement and Development – and lobbied for Myanmar conglomerates and the general population to contribute cash; it has so far received some $20 million. Nine taskforces were established, all related to development.

The risk is that if, as seems likely, repatriation does not proceed quickly or at scale, and there is no dramatic progress on desegregation or citizenship for Muslim communities across Rakhine State, this fund will end up supporting development initiatives that increase inequality and exacerbate conflict.[fn]“State Counsellor: ‘Myanmar does not fear world scrutiny’”, GNLM, 20 September 2017; “Join hands for peace in Rakhine”, GNLM, 13 October 2017. The taskforces are: infrastructure, agriculture and livestock, economic zone development, information and public relations, job creation and vocational training, healthcare, microfinance, crowdfunding, tourism promotion; “Nine private sector task forces formed to participate in UEHRD programme”, GNLM, 22 October 2017.Hide Footnote As we have noted in prior reports and briefings, development interventions must be properly sequenced with political steps to address discrimination, segregation and citizenship status.[fn]See, for example, Crisis Group Report, Myanmar: The Politics of Rakhine State, op. cit.Hide Footnote

B. International Response

The crisis has prompted significant international scrutiny and criticism. UN Secretary-General Guterres sent an official letter to the Security Council on 2 September 2017 – the first time a Secretary-General has done so on any issue since 1989 – saying that “the international community has a responsibility to undertake concerted efforts to prevent further escalation of the crisis”. The Council met five times in August-October on the issue – including a briefing by Guterres on 28 September and a 13 October closed-door “Arria Formula” briefing with Kofi Annan, who was appointed by Suu Kyi in 2016 as chair of an advisory commission on Rakhine State, which completed its work in August. Guterres called on Myanmar to end the violence, allow unfettered humanitarian access, ensure the safe, voluntary, dignified and sustainable return of the refugees to their areas of origin, and prioritise implementation of the Annan commission recommendations – points echoed by several Council members.

On 6 November, given Chinese and Russian opposition to a resolution, the Council instead unanimously agreed on a presidential statement that “strongly condemns the widespread violence that has taken place in Rakhine State since 25 August, which has led to the mass displacement” of Rohingya communities; “expresses alarm at the significantly and rapidly deteriorating humanitarian situation”; and “demands the Government of Myanmar grant immediate, safe and unhindered access to United Nations agencies and their partners”. Myanmar expressed “deep concern” at the adoption of the statement and its use of the term “Rohingya”. The UN General Assembly approved a human rights resolution on Myanmar on 16 November, reviving annual resolutions dropped in 2016 in recognition of the country’s progress.[fn]See “Briefing under ‘any other business’”, Whatsinblue.org, 12 September 2017; “Public Briefing by the Secretary-General”, 27 September 2017; “Adoption of a Presidential Statement”, 6 November 2017; UNSC Presidential Statement S/PRST/2017/22, UN Security Council, 6 November; Statement of Myanmar Permanent Representative, GNLM, 8 November 2017; “Situation of human rights in Myanmar”, UN doc A/C.3/72/L.48, UN General Assembly, 31 October 2017.Hide Footnote

Some countries also raised concerns bilaterally in a series of phone calls and meetings with Suu Kyi and the Commander-in-Chief. On 19 September, the UK announced it was suspending training programs for the Myanmar military and Prime Minister Theresa May signalled her willingness to support further action. The European Union Council of Foreign Ministers decided on 16 October to suspend visits of Myanmar military officers to Europe and review all defence cooperation, while also flagging the possibility of more formal sanctions.

On 23 October, the U.S. issued a statement outlining its own steps, including restrictions on travel of current and former senior military leaders to the U.S., cancelling military-to-military engagements and exploring options for visa bans and asset freezes under the Global Magnitsky Act. On 22 November, Secretary Tillerson declared that the situation in northern Rakhine constituted ethnic cleansing and that accountability would be pursued through U.S. law, including possible targeted sanctions. Congress is currently vetting draft legislation that would re-impose some of the sanctions lifted in 2016.

Canadian Prime Minister Justin Trudeau appointed a Special Envoy to spearhead diplomatic efforts to address the crisis, but the envoy, Bob Rae, was unable to secure any meetings with government officials during his visit to Myanmar in early November 2017.[fn]“UK suspends aid for Myanmar military”, BBC News, 19 September 2017; “Myanmar/Burma: Council adopts conclusions”, European Council Press Release, 16 October 2017; “Accountability for Human Rights Abuses in Rakhine State, Burma”, U.S. Department of State Press Statement, 23 October 2017; “Efforts To Address Burma’s Rakhine State Crisis”, U.S. Secretary of State Press Statement, 22 November 2017; Crisis Group interviews, diplomats, Yangon, November 2017.Hide Footnote

Myanmar set its political direction early in the crisis, and, so far, international scrutiny, pressure and diplomatic engagement has brought about no meaningful change – not even seemingly minor concessions such as allowing UN humanitarian access to the area or signalling openness to international support or advice. Extremely strong domestic political consensus on this issue has united the government, military and vast majority of the population as never before in Myanmar’s modern history.

The huge reservoir of international good-will for Myanmar and for Suu Kyi personally that existed prior to the crisis is rapidly drying up.

The international community thus faces a major challenge. In the face of ethnic cleansing and crimes against humanity, the political and moral imperative to take action has become overwhelming. The huge reservoir of international good-will for Myanmar and for Suu Kyi personally that existed prior to the crisis is rapidly drying up. Many countries wish to support Myanmar’s transition away from military rule, and have no desire to undermine its first democratically elected government in more than 50 years. But given the strong perception that the diplomatic channel is not producing results, and with public views hardening in many countries in the West and the Muslim world, the imposition of sanctions by Europe and the U.S. seems inevitable. Over time, the drumbeat for holding those most responsible criminally accountable will also likely increase.

Sanctions are very unlikely to prompt positive change in Myanmar.

Yet policymakers should be under no illusions: sanctions are very unlikely to prompt positive change in Myanmar. Indeed, – depending on specifics – they could make matters worse. Unlike in the past, there is no domestic debate on different policy approaches that sanctions might be thought to influence. Their most likely effect will thus be to push the government, military and population even closer together and to reinforce current narratives in Myanmar that the West is a fickle friend and unreliable partner. Government leaders have explicitly warned that criticism and punitive actions from the West will only push them closer to China.[fn]Crisis Group interviews, Western diplomats, Yangon, September-November 2017. See also “U.S. Pressure on Aung San Suu Kyi Only Helps China, Aides Warn”, Wall Street Journal, 13 November 2017.Hide Footnote

History also is a guide. Until 2012, Myanmar was under some of the most stringent bilateral sanctions of any country; contemporaneous Crisis Group research indicates that these did almost nothing to influence the military regime and had very little tangible impact on it. Although termed “targeted”, they had little impact on the regime and its leaders, but caused significant damage to the general economy and the fortunes of ordinary people – something acknowledged for example by then-Secretary-of-State Hillary Clinton when she initiated a review of U.S. policy in 2009.[fn]See Crisis Group Asia Report N°78, Myanmar: Sanctions, Engagement or Another Way Forward?, 26 April 2004; Asia Briefing N°118, Myanmar’s Post-Election Landscape, 7 March 2011. “Shift Possible on Burma Policy”, Washington Post, 19 February 2009.Hide Footnote There are few new options on the table, and any return to sanctions will inevitably involve some of the same basic elements. For Myanmar, these do not represent ominous new threats but rather the prospect of return to a very familiar status quo ante.

Policymakers nevertheless feel they should act, not only in response to political pressure from their constituents but also to send an important broader signal to would-be perpetrators that such abuses will not go unpunished. There are ways policymakers can limit potential negative impact on the Myanmar people, who should not pay the price for the actions of a military that is constitutionally outside of democratic control.

  • First, resist the urge to disengage. Policymakers should not lose sight of the distinction between government and people. Myanmar is home to millions of the poorest people in the region, and their aspirations for a better economic future must not be forgotten. The urge to disengage from the country, therefore, should be resisted. People-to-people exchanges with the West through academic, cultural and commercial interactions and tourism are crucial for a country that was isolated for so many decades.
  • Second, maintain development assistance and non-military engagement. This will be easy for Western countries to commit to in theory, but hard to deliver in practice now that Myanmar is no longer a global good news story and its government is showing little flexibility on aid modalities. Trade preferences recently reinstated by the EU and U.S. are critical in supporting manufacturing jobs in Myanmar and should not be revoked.
  • Third, work carefully to minimise the collateral impact of any targeted sanctions. Targeted sanctions on specific individuals and entities against whom there is evidence of wrongdoing, can help to promote accountability. Recent experience in Myanmar shows, however, that ostensibly targeted sanctions can have broader systemic impact on the economy that should be avoided.
  • Fourth, engage with the military and government prior to imposing any sanctions. The goal should be to maximise any leverage that is available (even if minimal) at the critical moment of opportunity, by raising the prospect of any new sanctions and pushing for progress on the key objectives before these measures are imposed.


Given the limited utility of sanctions, the international community should do all it can to mitigate the humanitarian disaster and influence the situation in other ways. This could include:

  • Provide substantial ongoing humanitarian support to the Rohingya refugees, to reduce the risks of a further humanitarian catastrophe and alleviate the enormous burden on Bangladesh and local communities. This can help also mitigate the risk of refoulement.
  • Assist Myanmar to define a pathway out of the current crisis, on the understanding that at least part of the challenge relates to management and implementation ability, in addition to political will. In particular, since the development-first approach being pursued by the government will be neither credible nor effective, pushing for political decisions to implement key recommendations of the Annan commission, including as regards discrimination, segregation and citizenship. Meaningful progress on these issues is vital to creating an environment conducive to voluntary repatriation, and giving international credibility to the Myanmar’s efforts.
  • Begin contingency planning for the humanitarian, security and political consequences of a scenario where the Myanmar-Bangladesh bilateral process does not lead to significant numbers of refugees returning home. This will be discussed in detail in forthcoming Crisis Group reporting.


China is particularly well-placed to promote positive outcomes should it decide to prioritise these. While in recent decades it has always supported Myanmar governments politically, and continues to be sceptical of international pressure, its blanket support cannot be taken for granted by Myanmar. China does not want this to come at the cost of its important relations with Bangladesh and the wider Muslim world, which is part of the reason why it allowed the recent UN Security Council presidential statement to be issued. China also has significant economic and strategic interests in Rakhine State that could be impacted by the crisis. So far, however, it has focused on allowing Myanmar and Bangladesh to work out the issue bilaterally.

VI. Conclusion

The actions of the Myanmar military in northern Rakhine State have created a major humanitarian catastrophe, a crisis for the country and a security threat to the region. It has strengthened an ugly strand of nationalism that will be long-lasting and could lead to the targeting of other minorities in the future. The crisis will define Myanmar in the eyes of much of the world for years to come, with hugely negative consequences across the board on trade, investment, tourism. The country has squandered its considerable reserves of global good-will just when it needed them most, as it was emerging from decades of isolation from the West. Myanmar has also put itself at much greater risk of attack by transnational jihadist groups. Priority long-term aims of balancing China’s geostrategic influence and economic dominance in the country and rehabilitating the military’s international image have been significantly set back.

The abuses against the Rohingya minority have captured global public opinion, and the uncompromising posture of the government has exacerbated the situation. Western countries almost certainly will re-impose some of the sanctions that had been lifted in recent years. As they do so, they should acknowledge their inherent limitations and approach them in a manner that can maximise leverage while minimising collateral damage on Myanmar’s long-suffering population.

Brussels, 7 December 2017

Appendix A: Map of Myanmar

Map of Myanmar Mike Shand/International Crisis Group, 2017.